RESEARCH ARTICLE www.advenergymat.de Functionalized MXene/Halide Perovskite Heterojunctions for Perovskite Solar Cells Stable Under Real Outdoor Conditions Masoud Karimipour,* Ashitha Paingott Parambil, Kenedy Tabah Tanko, Tiankai Zhang, Feng Gao, and Monica Lira-Cantu* Despite the performance improvement in perovskite solar cells (PSCs) when 1. Introduction MXenes are employed as transport layers, device stability studies are still 2D MXene nanomaterials have been ap- missing. Especially under real outdoor conditions where devices are subjected plied successfully in many fields of re- to the synergy of multiple stressors. In this work, functionalized 2D titanium search such as sensors,[1] batteries,[2] supercapacitors,[3] water purification,[4] carbide (Ti3 C2 ) MXene is employed in normal PSC configuration, at the or energy harvesters[5–7] including many interface between the halide perovskite and the hole transport layer. The others. Among all types of 2D MXenes, functionalization of the Ti3 C2 MXene is made utilizing the same organic Ti3 C2 -Tx nanosheets are characterized by additive passivating the halide perovskite layer. The functionalizing strategy their ease of chemical exfoliation, low creates a continuous link between the MXene and the halide perovskite layer. cost, high conductivity, and abundant ter- minal groups for surface chemistry that Champion MXene-based PSCs revealed a ≈22% efficiency, in comparison permit the manipulation of its proper- with the control device showing 20.56%. Stability analyses under ISOS ties for the targeted applications.[8,9] The protocols under different conditions (dark, continuous light irradiation and availability of abundant terminal groups real outdoor analysis) reveal that the enhancement of the PSCs lifespan is on the surface of Ti3 C2 nanosheets pro- always observed when the MXene layer is employed. Analysis under vides not only tremendous possibilities continuous light irradaition (ISOS-L) reveal an almost 100% retention of the to chemically manipulate and harness its surface chemistry,[10] but also the feasi- efficiency for the MXene-modified device, and outdoor testing (ISOS-O) bility of tuning its work function from carried out for > 600 h reveals a T80 of ≈600 h, while the control device metallic to semiconductor behavior.[11,12] degrades completely. To the best of the authors’ knowledge, this is the first The benefits of these properties have report of the stability assesment of MXene-based PSCs carried out under real been also implemented in perovskite outdoor (ISOS-O) conditions. solar cells as selective or intermediate layers.[13–18] The main idea of apply- ing 2D MXene nanosheets in perovskite solar cell has been mostly focused on addressing the issue of in- creasing the power conversion efficiency (PCE)[16,19,20] by means M. Karimipour, A. Paingott Parambil, K. Tabah Tanko, M. Lira-Cantu Catalan Institute of Nanoscience and Nanotechnology (ICN2) of defect passivation,[21] work function alignment[17,19,22] and CSIC and the Barcelona Institute of Science and Technology (BIST) efficient charge transfer.[5,13] For example, Bati et al. has em- Building ICN2 ployed a mixture of MXene nanosheets and single walled carbon Campus UAB nanotubes to passivate the electron transport layer (ETL :SnO2 Bellaterra, Barcelona E-08193, Spain E-mail: masoud.karimipour@icn2.cat; monica.lira@icn2.cat nanoparticle layer) at the perovskite interface with a demon- T. Zhang, F. Gao strated PCE over 20%.[23] Similarly, Wang et al., have also used Department of Physics, Chemistry and Biology (IFM) MXene nanosheets as electron extracting booster on the SnO2 Linköping University NPs layer and observed a significant increase of the short cur- Linköping 58183, Sweden rent density (JSC ) due to the alignment of the work function stud- ied by Ultraviolet photoelectron spectroscopy (UPS) and time- The ORCID identification number(s) for the author(s) of this article resolved photoluminescence (TRPL) analyses.[16] In a similar ap- can be found under https://doi.org/10.1002/aenm.202301959 proach, MXene oxidation has been used as a novel strategy to © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH. This is an open access article under the terms of the enhance ETL/perovskite interface passivation by Yang et al.[24] Creative Commons Attribution-NonCommercial-NoDerivs License, and Niu et al.[25] Yang et al. found that the oxidation of MX- which permits use and distribution in any medium, provided the original ene Ti3 C2 can generate Ti─O bonds that lead to better attach- work is properly cited, the use is non-commercial and no modifications ment to perovskite film, thus the oxidized MXene can be used or adaptations are made. as ETL layer.[5] Niu et al. found that Nb2 C MXenes can be DOI: 10.1002/aenm.202301959 incorporated in the ilattice facets spacing of SnO2 layer and Adv. Energy Mater. 2023, 13, 2301959 2301959 (1 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de lead to the increase of SnO2 grains and therefore better en- In this work, functionalized 2D Ti3 C2 MXene is employed ergy band alignment with the perovskite layer, increasing charge in a normal PSC configuration and placed at the interface be- transfer.[25] In another report, MXene nanosheets have been tween the halide perovskite (HP) and the hole transport layer added to the perovskite precursor solution and spin coated on (HTL). The functionalization of the Ti3 C2 MXene was made em- substrates to obtain large grains after annealing, obtaining an ploying the same organic additive used for the halide perovskite enhanced PCE of the devices by increasing JSC [13] or by pas- film. Our strategy permits us to create an improved and contin- sivating the Pb2+ and PbI antisite defects by extra etching of uous nexus between the MXene and the halide perovskite layer. XMenes in HF-Dimethyformamide (DMF) solution, obtaining Thermal admittance spectroscopy (TAS) analyses demonstrates F-rich Ti3 C2 Tx MXene nanosheets.[26] Recently, Agresti et al. that shalow defect passivation takes place when the functional- have also investigated that the addition of MXene nanosheets ization is made only at the bulk of the materials, while both, in the bulk of the perovskite film together with the interface shallow and deep defects, are passivated when an interface is in- engineering of TiO2 /perovskite, would allow the fabrication of volved. In addition, stability analyses under any analyzed condi- hysteresis-free devices with over 26% enhancement in PCE in tion (dark, continuous light irradiation and real outdoor analysis comparison with the reference PCE.[19] Taotao Chen et al. also (ISOS-L and ISOS-O),[29] demonstrate that enhancement of the introduced Ti3 C2 MXene nanosheets as accelerating hole extract- PSCs lifespan is always observed when the MXene layer is em- ing layer for all inorganic CsPbBr3 perovskite solar cells (PCE ployed. Analysis under continuos light irradaition (ISOS-L) re- 9.01%) together with a carbon contact. They found that MXene vealed an almost 100% retention of device performance, while nanosheets can suppress electron recombination and enhance outdoor testing (ISOS-O) for > 600 h revealed a T80 at ≈600 h, hole extraction at the interface of perovskite/carbon layers.[27] In an oustanding response if compared to the control PSC with T80 another recent work, Yao et al. have incorporated graphene ox- at 250 h. To our knowledge, this is the first report of the stabil- ide (GO) into MXene nanosheet synthesis, to passivate further ity analysis of MXene-based PSCs carried out under real outdoor the oxygen and OH bonding of termination groups of MXene (ISOS-O) conditions and indicates that MXene can pave the way nanoshets and induce their hygroscopic behavior at the inter- to a highly stable PSC required for the commercialization of the face of carbon/perovskite layer to enhance charge extraction and technology. achieved a significant PCE of 15.04% for CsPbI2 Br carbon based solar cells.[28] 2. Results and Discussion Reports on the application of MXenes for the stability enhance- ment of PSC have not been investigated profoundly. Most stud- 2.1. MXene Delamination (d-MXene) and Functionalization with ies are limited to the analysis of devices under storage conditions H3pp (MXene:H3pp) in the dark. However, the study of accelerating aging tests un- der continuous light irradiation and especially under real out- The 2D MXene Ti3 C2 was obtained commercially and used door operational conditions is currently in its infancy. Studies as received. The bulk MXene powder was delaminated us- are reported mostly for inorganic perovskites which are known ing a method based on a multi-sequential solution stirring for their superior stability in comparison to the hybrid organic- on Tetrapropylammonium hydroxide (TPAOH), Tetrabutylmmo- inorganic perovskites. Among the few examples is the work de- nium hydroxide (TBAOH), and Tetramethylammonium hydrox- veloped by Chen et al. who has exposed inorganic perovskite solar ide (TMAOH) at room temperature for 24 h. The as-prepared cells to thermal and humidity conditions in the dark (ISOS-D-2 dispersion was then washed three times in isopropanol follow- protocol).[29] In a similar way, Wang et al. have set encapsulated ing by centrifugation[7] (see Figure S1, Supporting Information). MXene-based PSCs devices under ambient air and ambient hu- A green-colored dispersion was obtained as the characteristic of midity conditions and tracked their shelf life in the dark (ISOS- the delaminated MXene (d-MXene). D-2).[16] Analysis under continuous light irradiation (ISOS-L pro- We have selected the additive 3-phosphonopropionic acid tocol) has been carried out by Song et al, who reported a 50% (H3pp) due to its two functional groups, phosphonate and car- performance loss after 500 h in comparison with the control de- boxylate, and the outstanding device stability observed when ap- vice that loses almost the 80% of its initial efficiency.[30] Hyuk plied within the halide perovskite layer.[32] In this case, we want to Im et al., have recently reported the application of oxidized MX- explore a possible beneficial effect of this type of additives when enes (OMXene) applied as a barrier layer for oxygen and mois- intercalated into the MXene layer. Thus, the functionalization of ture, in CsPbI3 based PSC minimodules. The devices were stored the MXene (MXene:H3pp) was obtained by the direct addition of in a chamber at 85°C/85 RH under light irradiation (no electri- the additive H3pp into the d-MXene dispersion and was sonicat- cal connections) with Xenon lamp (modified ISOS-L3 protocol), ing for 20 min. The H3pp was employed using three different resulting in device stability for more than 1000 h under con- concentrations, 0,25 mg, 0.5 mg, and 1 mg, per 2 ml of green tinuous light irradiation.[31] Reports on the outdoor stability as- d-MXene. Figure 1a–d shows a schematic representation of the sessment of PSC can be found in the literature for 2D materi- delamination (Figure 1b) and functionalization (Figure 1c) pro- als. For example, the groups of Kymakis and Di Carlo, et al.,[11] cess followed by its deposition on a halide perovskite (HP) thin reported on the outdoor analysis of PSCs modules fabricated film (Figure 1d). with 2D graphene and 2D MoS2 carried out over months-long After synthesis, the structural, morphological, and elec- timescale.[11] However, to our knowledge, this is the first report of tronic properties of the materials were verified comprehensively. the stability assessment of 2D materials such as MXenes applied Figure 1e shows the XRD of the d-MXene (red) and its compar- in PSCs and carried out under real outdoor (ISOS-O protocol) ison to the bulk MXene (blue) before delamination. The peaks conditions. indicated as (002), (006), (008), (103), and (110) correspond to Adv. Energy Mater. 2023, 13, 2301959 2301959 (2 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 1. Schematic represenation of a) bulk MXene, b) delaminated MXene (d-MXene), c) MXene functionalized with H3pp (MXene:H3pp) and d) thre heterojunction made of the MXene:H3pp thin film and the Halide Perovskite thin film (MXene:H3pp/HP); (e) wide range X-ray diffraction (XRD) patterns of bulk MXene (blue), d-MXene and the pattern of the Ti3 AlC2 ICSD No. 153 266[33–35]. as reference (black); f) zoomed area of e); g) wide range X-ray diffraction patterns of Bulk MXene (blue), d-MXene (red) and MXene:H3pp (green); (h) zoomed area of g); i) Raman spectra of d-MXene (red) and MXene:H3pp (green); j) Fourier-Transform inferared (FTIR) spectra of d-MXene (red) and MXene:H3pp functionalized with H3pp 0.5 mg (light green) and 1 mg (dark green). Adv. Energy Mater. 2023, 13, 2301959 2301959 (3 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Ti3 AlC2 MAX phase, ICSD No. 153 266 (Figure 1e, black).[33–35] 2.2. The Halide Perovskite (HP) and MXene:H3pp As it is inferred, the MXene structure undergoes a signifi- Heterojunction (HP/MXene:H3pp) cant shift of the main diffraction peak (002) after delamination corresponding to an increase of lattice space in the (002) atomic The fabrication of the HP/MXene:H3pp heterojunction was plane direction. According to Figure 1e,f and the Bragg diffrac- made by spin coating a solution of the MXene:H3pp on top of the tion equation, this shift corresponds the expansion of the atomic Rb0.05 Cs0.05 MA0.15 FA0.75 Pb(I0.95 Br0.05 )3 quadruple perovskite thin plane from 10.01 Å to 14.14 Å. To scrutinize the effect of func- film layer. The as-prepared HP/MXene:H3pp heterojunction was tionalization of MXene with the ligand H3pp, the XRD diffrac- then annealed at 70°C for 3–5 min (see Figure S4, Supporting tion pattern of the MXene:H3pp is shown in Figure 1g and com- Information). Figure S5 (Supporting Information) shows the top pared with the d-MXene before functionalization. Figure 1g,h view SEM images of the HP/MXene:H3pp heterojunction. For show that the MXene functionalization causes a slight but ob- comparison purposes we included a HP/d-MXene thin film het- servable shift of the (002) Bragg peak to lower angles, an in- erojunction (without the H3pp) and the control thin film made dication of a slight increase of the atomic plane distance due of the bare HP thin film. Figure S6 (Supporting Information) to the intercalation of H3PP ligand (≈7.0 Å) within the MX- shows the EDX analyses of the d-MXene modified HP thin film ene nanosheets (see schematic representation in Figure 1a–d). corresponding to the SEM image of Figure S5 (Supporting In- To examine, the ligand H3pp intercalation within the MXene formation). The samples show uniform coverage of the HP film nanosheets, Raman spectra of MXene:H3pp (MXene:H3pp) are with d-MXene nanosheets, and the observed MXene nanosheets shown in Figure 1i and compared with the d-MXene without are so thin that the HP grains are observable even thorough the ligand. Peaks at 127 cm−1 , 148 cm−1 , 245 cm−1 , 439 cm−1 , and MXene nanosheets. It also indicates the presence of large area 613 cm−1 correspond to different vibration modes of the Ti3 C2 MXene nanosheets for the Ti and Carbon elemental mapping, nanosheets in well agreement with the literature for the delami- with distinctive contrast from the background, an indication of nated MXene (d-MXene).[36,37] The peak ≈245 cm−1 represents the presence of the MXene nanosheets on the surface of the per- out-of-plane stretching vibration of surface Ti with ─OH ter- ovskite film. To confirm the successful deposition of the MXene mination groups. After functionalization of the d-MXene with nanosheets without altering the HP thin film chemical composi- the H3pp molecule, the peaks at 148 cm−1 , 613 cm−1 , and tion, we carried out X-ray Photoelectron spectroscopy (XPS) anal- 245 cm−1 have been quenched and a new peak at 301 cm−1 is yses of the HP thin film with and without d-MXene on top (HP/d- observed which is attributed to the in-plane oxygen vibration MXene heterojunction) (Figure S7, Supporting Information). We modes from the termination groups (Tx ) of MXene naosheets[38] observed that the deposition of d-MXene on the HP film, results while preserving the peak at 963 cm−1 .[37,39] These results demon- in a slight increase of binding energy for Pb 4f, I 3d, and more strate that after addition of the H3pp molecule, the Ti and C importantly N1s orbitals which can alter the work function and vibration modes are strongly weakened and the oxygen termi- effective passivation of the surface of the HP thin film.[42–44] nation group are strengthened due to the presence of a P─O bonding in H3pp by its attachment to the MXene nanosheets surfaces. 2.2.1. PSCs Applying the HP/MXene:H3pp Heterojunction as HTL We employed FTIR analyses to understand the type of bond- ing between the H3pp and the MXene nanosheets. Figure 1j PSCs were fabricated in a normal configuration of shows the FTIR spectra of d-MXene and the MXene:H3pp func- the type: FTO/c-TiO2 /m-TiO2 /HP/MXene:H3pp/Spiro- tionalized with different amounts of H3pp (0.5 and 1 mg). The OMeTAD/Au. Where the HP is the quadruple perovskite MXene:H3pp spectra shows representative peaks at 980 and Rb0.05 Cs0.05 MA0.15 FA0.75 Pb(I0.95 Br0.05 )3 absorber, and the MX- 1216 cm−1 that are attributed to P─O vibration modes of the ene:H3pp acts as the hole transport conductor. To understand H3pp molecule.[40] With the increase of H3pp concentration the effect of the MXene nanosheets and its functionalization during synthesis (from zero to 1 mg), the characteristic MX- on the PV properties of the final device, we fabricated PSCs ene vibration modes (Ti-O)[34] at 631 and 1317 cm−1 shifts to- where the MXene:H3pp was replaced by the bare d-MXene ward higher wavenumbers (as shown with arrows in Figure 1j). and compared to the PSC without any interface modification This is an indication of the binding of the H3pp molecule to (control). Figure 2a shows the schematic representation of a the MXene surface which provokes an increase of rigidity of the perovskite cell applying the functionalized MXene with H3pp structure and vibration frequencies.[41] To further confirm the (MXene:H3pp), at the interface between the SpiroOMeTAD morphological properties of the materials, scanning electron mi- and the HP layers. The photovoltaic (PV) response of the PSCs croscopy (SEM) and transTEM images of the powdered sam- are depicted in Figure 2b–d, where Figure 2b shows the statis- ples are depicted in Figure S2 and S3, respectively (Supporting tics observed for the PCE with the following values: control Information). We can observe the morphology evolution of the 19.45(±0.12) %, d-MXene 19.69(±0.11) %, and MXene:H3pp Ti3 C2 -Tx MXene from the bulk MXene to the delaminated MX- 19.82(±0.07) %. Figure 2c shows the reverse scan of the JV ene (d-MXene) and the functionalized material (MXene:H3pp). curves of the champion devices with highest PCE of 19.57 %, Once it evolves from the bulk to the delaminated MXene, a thin 19.80 %, and 19.89 %, for the control, the d-MXene and the nanosheets with deflected surface can be observed attributed to MXene:H3pp devices, respectively. In general, the PV response structural distortion. Moreover, functionalization of d-MXenes is very similar for all devices, with a very slight increase in with H3pp molecule, does not affects the morphology of the PCE observed for the solar cells applying the MXene:H3pp thin d-MXenes significantly by just slightly folding the edges of the film. This enhancement mostly comes from a small increase nanosheets. in FF, VOC and JSC .[45,46] Incident photon to charge conversion Adv. Energy Mater. 2023, 13, 2301959 2301959 (4 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 2. a) Schematic representation of a perovskite solar cell applying the MXene functionalized with H3pp (MXene:H3pp); b) PCE performance statistics of control PSC (black), PSC modified with d-MXene (red) and MXene:H3pp (green); c) JV curves of champion devices; d) IPCE of the champion devices; e) photoluminescence spectra (PL) of the HP film with/without a coated thin film of the d-MXene (red) and MXene:H3pp (green). efficiency analyses (IPCE), Figure 2d, also agree with these could be due to bonding interaction of the MXene nanosheets results showing a slightly higher IPCE values for the solar cells with the HP surface as XPS analyses indicate. In summary, employing the MXene:H3pp modified device. In addition, the interface modification using both d-MXene and MXene:H3pp calculated integrated JSC values observed from IPCE analyses show enhancement in device performance parameters and are in good agreement with those obtained from the JV curves. effective charge extraction as well as initial indication of effective Figure 2e depicts the photoluminescence (PL) spectra of the passivation of HP film. different samples, where the HP/MXene:H3pp HP/d-MXene films (Figure 2e-green and red) changes both in terms of peak intensity and peak position with respect to the control sample 2.2.2. The Interaction between the HP and the MXene:H3pp (Figure 2e-black). Thus, the functionalization of the MXene with the H3pp ligand results in a slight but observable shift of the The photovoltaic results and PL analyses described above indi- PL peak in ≈10 nm toward higher wavelengths (lower ener- cate that the MXene functionalization may affect carrier redis- gies) together with an increase in the peak intensity. This shift tribution in a PSC.[47] To prove this claim, we constructed the Adv. Energy Mater. 2023, 13, 2301959 2301959 (5 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 3. a) Band diagram of PSCs with/without interface modification using d-MXene (red) and MXene:H3pp (green) obtained from UPS analyses, b) TRPL spectra of the MHP films, c) Mott-Schottkey plots of the PSC devices with/without interface modification using d-MXene (red) and MXene:H3pp (green), d) Reverse-Forward JV scans of the devices, and e) trap density versus energy depth profile plots obtained from TAS analyses of the corresponding devices at 320 K. energy band diagrams. For this, ultraviolet photoelectron spec- general, all the modifications indicate a shift of the VB from troscopy (UPS) was performed (Figure S8, Supporting Informa- −5.29 eV for the control perovskite, to −5.67 eV for the d-MXene tion) to measure the work function and valence band position and −5.89 eV for the MXene:H3pp, an indication of an enhanced of the bare HP film (control) and the corresponding heterojunc- hole-extraction properties of the HP to the HTL, in agreement tions with the MXene thin films. The corresponding band dia- also with the IPCE analyses. grams are shown in Figure 3a (bandgap was calculated as 1.48 To further analyze the interaction between the HP and the MX- (±0.01) eV from IPCE measurements) for the corresponding de- ene:H3pp, TRPL spectroscopy was performed on the devices and vice configuration with cross-SEM image in Figure 3b. The com- the resultant spectra are shown in Figure 3c. The calculated life- parison of the band diagrams for the different samples indicates time values, obtained by fitting with the three exponential de- that the surface of the HP film, with an original work function cay function with the R2 >0.98 using pre-scripted PicoQuant soft- (WF) of −4.51 eV for the non-modified film, experiences a shift ware, show significantly larger values for samples employing de- downward to lower energies, −4.82 eV. Respectively, when the d- laminated MXene nanosheets in comparison with the control de- MXene is employed, a further decrease of the WF to −5.09 eV vice. Moreover, PSCs with the MXene:H3pp thin film showed the is observed when the MXene:H3pp (1 mg of H3pp) is used. In longest lifetimes. Adv. Energy Mater. 2023, 13, 2301959 2301959 (6 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de To analyze the possible defect passivation and non-radiative re- 25th of analysis) as observed in the ISOS-D-1 stability analysis combination on the HP surface by the d-MXene or MXene:H3pp shown in Figure 4e. We can observe in Figure 4a that at the 25th thin films placed at the interface of the perovskite/HTL, we car- day of stability analyses, the PSC that maintain the best PV re- ried out density of trap states analyses. We employed capaci- sponse is the one applying the MXene:H3pp sample (PCE of tance measurements at different DC voltages at 20 kHz. Mott- 19.85%), followed by the d-MXene (PCE of 19.59%) and finally Schottky analyses were performed, and the experimental details the control device (PCE of 19.10%). At the end of the stability anal- are described in Supporting Information (Section 4). The re- ysis (Figure 4e) the PSC applying the MXene:H3pp showed the sulting Mott-Schottky plots are shown in Figure 3d and the cal- highest stability (green), while the PSCs with the d-MXene (red) culated trap densities values are 3.49 × 1016 , 3.19 × 1016 , and and the control device (black) were observed to maintain a simi- 2.32 × 1016 cm1 3 for the control, d-MXene, and MXene:H3pp de- lar but lower performance if compared to the former. This result vices, respectively.[48] The reduction of trap states observed for the is in good agreement with the interaction between the Ti3 C2 -Tx MXene:H3pp device is more evident, while the d-MXene show MXene and ligand molecules containing phosphonate groups, as a slight decrease in trap density value if compared to the con- in H3pp, which is a well know method to stabilize MXenes. The trol device. This is in well agreement with the results of TRPL formation of a -Ti-O-P- bond (demonstrated by our Raman anal- analyses.[49] To shed more light on the trap densities and defect yses) prevents the oxidation of MXene, improving the stability energy levels, a characterization of the distinct type of defects PSC of our devices. (deep or shallow trap states) was made. Thus, thermal admittance Stability under ISOS-L Protocol: To further study the effect spectroscopy (TAS) was employed to study the significance of trap of MXene:H3pp and d-MXene nanosheets on the PSC stability states in energy space for the modified and non-modified devices. under light irradiation conditions, the devices were set under Figure 3e shows the trap densities of the devices in energy space simulated 1 Sun (calibrated LED lamp) and continuous N2 gas below conduction band that are calculated from TAS analyses at flow at maximum power point tracking (MPPT) for 1000 h ac- 320 K. The application of the d-MXene results in a 10-fold and cording to ISOS-L-1 protocol (Figure 4f). We observed that under 1000-fold decrease of the shallow and deep traps state, respec- these conditions, the most stable devices are those PSCs where tively, if compared to the control device. Interestingly, addition of MXene have been applied (d-MXene and MXene:H3pp). In this H3pp as functionalized for d-MXene (MXene:H3pp) results in a case, the d-MXene and the MXene:H3pp devices both show simi- 10-fold increase of shallow trap state if compared to the control lar stability under continuous illumination and MPP conditions, device. This shift observed in the Fermi level toward the conduc- always higher if compared to the control PSC. The d-MXene, tion band for the MXene-H3pp devices is in well agreement with MXene:H3pp, and control devices have reached T80 under MPP the UPS results. Thus, the use of the Ti3 C2 d-MXene as the in- tracking at 256 h, 140 h, and 84 h, respectively.[50] terface between the HP and the Spiro-OMeTAD, works in favor To investigate the effect of interface modification on the pho- of the passivation of both deep and shallow trap states of HP sur- tostability of the MHP layer, the cross-section SEM images of face, while the introduction of the H3pp ligand in MXene:H3pp, the devices after 1000 h of MPPT under continuous light illu- results in the sole passivation of the deep trap states. These re- mination (ISOS-L) were analyzed and are shown in Figure S9 sults are also in excellent agreement with the Voc and hysteresis (Supporting Information). The SEM images demonstrate that of the samples shown in Figure 3f, where an increase of the Voc the control device has undergone perovskite degradation as in- and the reduction of the hysteresis index is observed from the dicated by the formation of large crystals (indicated as dotted cir- control device (bare HP), to the d-MXene and MXene:H3pp de- cles) that have been attributed to the formation of PbI2 ,[45] while vices, respectively. d-MXene and MXene:H3pp devices do not show any sign of PbI2 In summary, MXene nanosheets have been delaminated (d- formation.[45] A clear formation of similar crystals as shown in MXene) and functionalized with the organic ligand H3pp (MX- Figure S9a (Supporting Information), for the control device can ene:H3pp). The application of the functionalized MXene passi- be seen in Figure S9b (Supporting Information) for the PSC ap- vates both shallow and deep trap states. As a result, band align- plying only the H3pp at the interface between the HP and the ment and charge extraction are improved. Spiro-OMeTAD. The latter demonstrates that the application of MXenes (bare or functionalized) are beneficial for the stability of the PSC. 2.2.3. Stability Assessment of PSCs Applying the HP/MXene:H3pp Stability under ISOS-O Protocol: Outdoor analysis is consid- Heterojunction ered the best method to analyze the stability of PSCs due to the synergy of multiple stressors that vary in intensity and strength In order to analyze the effect of the d-MXene and MXene:H3pp during the analysis of the PSCs under continous real conditions. interface modifications on the lifetime stability of PSCs, we car- Variations not only in light intensity but also in temperature, hu- ried out stability assessment of compete solar cells following the midity, bias voltage, including the cycling of these factors, are ob- ISOS-D, ISOS-L, and ISOS-O protocols.[29] served during outdoor testing. In these conditions, the encapsu- Stability under ISOS-D Protocol: The devices underwent an lation made to the devices plays a critical role in the overall device initial lifetime check under ISOS-D-1 protocol conditions by performance. We have developed and protected a methodology keeping the un-encapsulated samples in the dark and N2 at- for the encapsulation of our PSCs for outdoor analysis (Patent ap- mosphere (inside the glove box) for more than 3000 h, while plication submitted. Internal Reference Number ICN2IND_23- JV curves were regularly measured during the whole test 22) ensuring not only a good encapsulation but also the retention (Figure 4a–e). Figure 4a-d show the statistics of the PV param- of the PV parameters before and after encapsulation (before out- eters corresponding to the devices at their maximum PCE (day door analysis). Thus, PSCs with the highest PCE performances Adv. Energy Mater. 2023, 13, 2301959 2301959 (7 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 4. a–d) Summary of performance statistics of the devices at 25th day of life time tracking using ISOS-D-1 protocol (un-encapsulated, kept in dark, room temperature and N2 atmosphere), e) life time tracking of the PCE of the devices for 2880 h (120 days) using ISOS-D-1 protocol, f) Normalized MPP tracking of the devices for 1000 h following ISOS-L-1 protocol (un-encapsulated devices, N2, under continuous simulated 1-sun light (A.M. 1.5G), at 20 °C). observed before and after encapsulation (Figure S10, Supporting ments and are shown in Figure S11 (Supporting Information). Information) were selected to be studied under outdoor condi- We observed that the control PSC device (black) degraded com- tions and MPP tracking (ISOS-O-1). Figure 5a,b shows the first pletely after 275 h, while the PSCs applying the MXene:H3pp 350 h of the outdoor stability analysis where only the data ob- (green) together with d-MXene (red) have the best stability af- tained at maximum PCE are plotted (Figure 5a) and all the data ter 350 h. We selected four specific days at different times dur- (day and night) are shown (Figure 5b) for the same test. In both ing the outdoor stability analysis to analyze in more detail the cases, the temperature, humidity, power input (sun intensity il- PV response. Figure 5c–f shows the evolution of the MPP for lumination power) were also recorded alongside the measure- the whole day and night for each sample at ≈24 h, 90 h, 172 h, Adv. Energy Mater. 2023, 13, 2301959 2301959 (8 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 5. a) Obtained outdoor MPP values of the encapsulated devices (ISOS-O-1 at Barcelona, 41.5021° N, 2.1039° E, Spain) for 350 h fabricated by interface modification using d-MXene (red) and MXene:H3pp (green), b) The observed highest daily outdoor MPP values corresponding to highest irradiance for each day in (a) , c–f) selected days of MPP tracking of the devices for the comparison of performance evolution of the devices marked with numbers (1-4) in b). Adv. Energy Mater. 2023, 13, 2301959 2301959 (9 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de and 218 h. We have observed that despite of the slightly better and the HP thin films, to form a HP:H3pp/MXene:H3pp initial performance of the control device (day 1-Figure 5c), the heterojunction. The PSC configuration is the follow- PSC with the MXene:H3pp sample has superior performance ing: FTO/c-TiO2 /m-TiO2 /HP:H3pp/MXene:H3pp/Spiro- under high humidity (60-85%) and low light intensities (90 h- OMeTAD/Au. Where the HP is the quadruple perovskite 172 h, Figure 5d,e), and it is able to maintain and even improve Rb0.05 Cs0.05 MA0.15 FA0.75 Pb(I0.95 Br0.05 )3 absorber. The HP:H3pp its performance after a low light irradiation and high humid- precursor solution was prepared as detailed in Supporting ity raining day (218 h, Figure 5f). An interesting observation is Information (Section 1.5) and deposited by spin coating on top the recovery observed for the PSCs after the samples were ex- of the m-TiO2 thin film. After drying, the MXene:H3pp was posed to a low-light irradiation and rainy day (after ≈115 h). The deposited on top of the HP:H3pp thin film following the method PSCs with MXenes recover and maintain their stability during employed in the previous sections. For comparison purposes the following days, while the control sample, although it also re- we also fabricated PSCs with only the HP/MXene:H3pp, and covers initially, decays faster until complete degradation before control PSC without any thin film modification. 300 h. These results indicate that, in general, the use of a MXene as heterojunction with the HP enhances device stability, show- ing an outstanding response under continuous light irradiation 2.3.1. Interaction with H3pp (ISOS-L) when the H3pp functionalized MXene, MXene:H3pp, is employed. Thus, it can be concluded that the application of the Figure 6a shows a schematic representation of the PSC where functionalized Ti3 C2 MXene with H3pp permits the effective de- the H3pp is part of the MXene layer and also part of the bulk fect passivation and the enhancement of the PSC device oper- of the HP. XRD measurements were carried out for the dif- ational stability.[17,51] It is worth mentioning that although this ferent samples (Figure 6b). We can observe that the thin films interface modification does not increase the device efficiency sig- with the incorporation of MXene nanosheets show clearly less nificantly, it results in very stable devices, observed under both PbI2 content and higher Bragg intensities for (002) preferen- ISOS-L and ISOS-O protocols. Such a similar behavior has been tial growth plane which indicates better crystallinity and pos- reported also by Yakusheva et al.[22] sibly is due repair of edges at the crystal lattices explained by Summarizing the PSC stability analysis of the devices under Hong et al.[47] Figure 6c,d show the TRPL and PL spectra of the ISOS-D (Dark), ISOS-L (continuous light irradiation) and ISOS- different thin film samples. It can clearly be observed that the O (light-irradiation during day and recovery at night), we have ob- films employing the H3pp in both HP and MXene layers, have served that the best PSCs stability after 1000 h corresponds to the longer electron-hole recombination lifetime (1.180 μs). The PL PSCs applying the MXene:H3pp in the dark, to the sole d-MXene spectra also indicates a red shift for both, the HP/MXene:H3pp under ISOS-L and to the MXene:H3pp under ISOS-O conditions. thin film and the HP:H3pp/MXene:H3pp samples in compari- We attribute these different responses to several factors. Under son to the control HP thin film. According to Hong et al. this dark conditions the H3pp additive prevents the degradation of can be due to the effect of crystal change at the edges of the the MXene through the formation of a strong Ti-O-P covalent grains.[47] PSCs were fabricated with both the HP/MXene:H3pp bond, stabilizing the PSC. Under constant light irradiation we and the HP:H3pp/MXene:H3pp heterojunctions and, for com- speculate that this response can be due to a photocatalytic effect of parison purposes, we also included control PSCs devices where the MXene[52] which, under continuous light irradiation (ISOS- no modifications were made. Figure 6e shows statistics of the L, Figure 4f), the H3pp could be degraded by the MXene itself device PCEs, and Figure S12 (Supporting Information) contains with time. However, under outdoor (ISOS-O, Figure 6a) the so- the relevant statistical comparison of all the cells parameters lar cells are subjected to variable stress factors such as light and for the different categories. Figure 6f shows the JV curves of temperature cycling or different light irradiation intensity. Some the corresponding champion devices where the PCEs obtained of these stressors are detrimental for the devices (like high light are 21.21%, 20.9%, and 20.29% for HP:H3pp/MXene:H3pp, intensity) but others, like light cycling observed during day/night HP:H3pp/d-MXene, and the MXene:H3pp thin film as the in- shifts, could be beneficial to the stability of perovskite solar cells terface, respectively. All the solar cells showed greater PCE than due to the mechanism of recovery in the dark, well-known for the control PSC with 20.28%. Figure S13 (Supporting Informa- PSCs. The multiple conditions observed under real outdoor con- tion) shows the IPCE of the champion devices for each sample ditions requires in-deep and careful experimental analysis and in good correspondence with the JSC values obtained by the cor- are currently being carried out in our laboratory. responding integration. 2.3. The HP:H3pp/MXene:H3pp Heterojunction 2.3.2. Stability of the PSCs Applying the HP:H3pp/MXene:H3pp Heterojunction Given that the use of MXenes and MXene:H3pp in PSCs resulted in an enhanced PV performance and device stability, we want to Stability under Constant Light Irradiation Following ISOS-L Pro- further improve these results by applying the organic additive tocol: We carried out stability analysis of the different PSCs H3pp also in the bulk of the HP. This methodology, the addition applying the ISOS-L-1 protocol under MPP tracking. Figure 7a of additives with phosphonate and carboxylate groups within the shows the change observed in maximum power point (MPP) halide perovskite layer, was developed by our group with excel- for the first 1000 h under continuous simulated light irradia- lent results in terms of device stability.[32] Thus, in this work, tion of 1 sun at RT and under N2 atmosphere. We observed that the H3pp molecule was added to the bulk of both, the MXene the best stability corresponds to the PSC employing the H3pp Adv. Energy Mater. 2023, 13, 2301959 2301959 (10 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 6. PSC employing the H3pp ligand in both the HTL and the absorber in PSC of the type: FTO/c-TiO2 /m-TiO2 /HP:H3pp/MXene:H3pp/Spiro- OMeTAD/Au. a) schematic representation of the PSC with the HP:H3pp/MXene:H3pp heterojunction. Characterization of the MXene:H3pp (green), HP:H3pp+d-MXene (red), the HP:H3pp/MXene:H3pp (blue) and the control PSC (black): b–d) XRD, TRPL and PL; e) PCE statistics, f) JV curves of the champion devices for each category with their relevant device parameters (inset table). Adv. Energy Mater. 2023, 13, 2301959 2301959 (11 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 7. a) Stability analysis under MPP tracking of un-encapsulated devices following the ISOS-L-1 protocol (1sun, N2 atmosphere and RT), and b) trap density versus energy depth profile from conduction band plots obtained from TAS analyses of the corresponding devices at 320 K, c) JV curves of champion devices with bulk and heterojunction passivation: Control with no passivation (black), Interface passivation with MXene:H3pp(green), bulk passivation with H3pp+ interface passivation with d-MXene (red) and, bulk passivation with H3pp+ interface passivation with MXene:H3pp abbreviated as HP:H3pp+MXene:H3pp (blue) including the table of device performance parameters under simulated 1-sun irradiance. at both layers the perovskite (HP:H3pp) and the MXene (MX- In summary, the outdoor stability results indicate that the ene:H3pp), showing almost no degradation for more than 1000 h simultaneously bulk and interface functionalization can lead at MPP (Figure 7a, red). The PSC applying the MXene:H3pp to highly stable, durable devices. The use of organic addi- (Figure 7a, green) retained 75% of its initial MPP, in well repro- tive molecules containing the phosphnate and carboxilate func- ducibility agreement with the results shown in Figure 4c. Both tional groups can be extended from the bulk to the interface PSCs showed better stability than the control PSC (Figure 7a, and in combination with MXene nanosheets can act as ion black). migration inhibitor and moisture barrier, as well as effective In view of the afformentioned findings, we conducted a TAS trap density passivating agents. Figure 7c demonstrates the JV measurements for the selected high perfomance devices as curves of the champion devices together with device performance shown in Figure 7b,c. Figure 7b depicts the trap density Vs. en- parameters in Figure 7c-table inset. All PSC applying the H3pp ergy depth profile with respect to the conduction band edge. Our molecule show higher PCE than the control device with 20.57% results clearly indicates that all the passivation works in favor of and 1.11 V, for PCE and Voc , respectively. The corresponding val- the reduction of trap state densities, from a 10 order of magni- ues for HP:H3pp/MXene:H3pp device are 21.95%, and 1.14 V, tude up to more than 100 times. The most significant trap reduc- respectively. tion corresponds to the PSC where the additive is part of the bulk To demonstrate the effect of the H3pp ligand in the band of both the MXene layer and the perovskite absorber which in- diagram alignment of the PSC, UPS analyses of the different dicates a significant reduction of both shallow trap density and samples were carried out (see also Figure S14a, Supporting In- deep trap density of the order of 100 times for each one. Thus, formation) and the corresponding band alignment is shown in it can be concluded that the passivation of only the bulk of the Figure S14b (Supporting Information). The application of the MXene or the perovskite, independently, passivates shallow trap H3pp in the MXene:H3pp and the HP:H3pp thin films is ob- densities[32] while the application of the H3pp in both layers and served to decrease the VB position from −5.62 eV to −5.87 eV in at the same time, passivates also the deep trap densities at the comparison to the use of the H3pp only in the MXene:H3pp thin interface. film. This can enhance the hole extraction and reduced electron Adv. Energy Mater. 2023, 13, 2301959 2301959 (12 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Figure 8. ISOS-O protocol tracking for PSCs employing bulk and interface passivation; a) obtained averaged MPP values for 3–4 devices for each category during 600 h by holding the encapsulated devices at their open circuit voltage and recording their JV by 20 min intervals in outdoor at Barcelona, 41.5021° N, 2.1039° E, Spain, b–e) Daily evolution of averaged MPP values for the different categories: Control with no passivation (black), Interface passivation with MXene:H3pp (green), bulk passivation with H3pp+interface passivation with d-MXene (red) and, bulk passivation with H3pp+ interface passivation with MXene:H3pp (blue) from 1–4, in different days of ISOS-O tracking. recombination in the PSCs.[19] Chen et al. also reported the ac- HP:H3pp/MXene:H3pp heterojunction in a PSC, we carried celerated charge extraction via application of MXene nanosheets out an outdoor stability analyses for > 600 h by holding the in all inorganic perovskite solar cells.[27] They concluded that the devices at their open circuit voltage[53] and measuring their strong binding between = O surface bonds of Ti3 C2 and under- current-voltage (JV) every 20 min, see Figure 8. To ensure re- cordinated Pb atoms lead to passivation of interface and mit- producibility, between 3 and 4 devices for each category were igation of work function alignment.[27] In summary, we have analyzed simultaneously, including the control PSC with no demonstrated that the simultaneous functionalization of the MX- passivation (Figure 8a, black), the PSCs with the MXene:H3pp ene transport layer and the HP absorber results in the best pho- (Figure 8a, blue), the HP:H3pp/d-MXene (Figure 8a, red) and tovoltaic performance and excellent device stability under contin- the HP:H3pp/MXene:H3pp (Figure 8a, green). Before the out- uous light irradiation, ISOS-L protocol. door analysis, we verified that all PSCs efficiencies were be- Stability under Outdoor Conditions Following ISOS-O Proto- tween 18–21% under 1 sun simulated light intensity. Figure S15 col: To scrutinize in more detail on the performance of the (Supporting Information) shows the recorded humidity (black) Adv. Energy Mater. 2023, 13, 2301959 2301959 (13 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de and sun light irradiance (red) for 600 h. Figure S16 (Support- MXene:H3pp respectively, versus 61.93 (±1.08)° value for the ing Information) demonstrates the collective obtained PCEs non-modified (control) perovskite film. These results clearly in- of all the devices per each category and Figure 8a shows the dicate that the superior performance of the interface modi- extracted and averaged (including error bars) MPP of each fied devices are slightly influenced by hydrophobicity from the category. H3pp or the MXene. Thus, it can be inferred that the su- Figure 8b–e show the daily evolution of MPP values for the perior outdoor stability performance of the MXene modified four selected days. It can be elaborated from Figure S16 (Support- devices with/without H3pp functionalization is largely due to ing Information) and Figure 8a (dashed lines are for a guide for interface deffect passivation, and ion migration suppression the eye) that the control PSCs devices possess the most inferior with small contribution of the hydrophobicity properties of the stabilty with T80 between 250–275 h, while all the PSC employing materials.[32] MXenes (both the HP:H3pp/MXene:H3pp and the HP:H3pp/d- In summary, the introduction of the H3pp within the halide MXene, red and green respectively) show the highest stability perovskite absorber and the MXene HTL in the same PSC has with estimated T80 of 500 -600 h. Moreover, the PSCs devices with shown the best stability response under continous light irradia- MXene:H3pp (blue color) show intermediate stability with an es- tion (ISOS-L) and under real oudoor conditions (ISOS-O). This timated T80 of 420–450 h (which is in good agreement with the response is attributed to the stability effect that the H3pp has on results shown previosly in Figure 5a,b). the MXene layer but also to the significant ion-inmobilization ef- It is also worth noticing that daily tracking of the device´s fect of the H3pp molecule when employed witin the perovskie performance show the stability of the devices under different layer. This ion-inmobilization effect has also shown to be ben- light intensity, temperature and humidity conditions as shown eficial for the recovery of the solar cells after outdoor periods in Figure 8b–e. Day (1) (Figure 8b, from 40 h -56 h) represents under low light intensity or high humidity (clouds, rain) con- the MXene-based PSC photovoltaic response in a fully sunny day ditions. The solar cells employing the H3pp ligand are able to and at an early stage of the outdoor analysis. At this stage, the av- recover and stabilize after those conditions, while samples with eraged MPP of the devices are very similar for all the categories, only one of the layers contining the H3pp lighand or the con- which is in good consistency with Figure 6f-table inset. This is trol device, deteriorate steady with time until complete degra- also an indication of the consistency of our encapsulation pro- dation. Contact angle measurements carried out to the samples cess (Internal Reference Number ICN2IND_23-22). supports our observation that the hydrophobic effect of the func- The day labeled (2) (Figure 8c, from 208 h to 224 h) shows a tionalizaed thin films don’t play an important role on the device low-light intensity and partially rainy day (see also Figure S15, stability. Supporting Information). It is interesting to note that after the raining period, during the rise of the sunlight irradiance, all the PSCs employing the MXene material recover their photovot- 3. Conclusion laic performance while the control PSCs showed the worst re- covery. The PSC devices employing the d-MXene nanosheets In summary, we demonstrate the application of functionalized and the perovskite with the H3pp additive, showed the highest 2D Ti3 C2 MXene in Perovskite solar cells employed as the inter- perfromance improvement.[32] These results are in agreement face between the hole transport layer and the perovskite absorber. with the H3pp role of being a good linker to suppress ion mi- The functionalization of the Ti3 C2 MXene was made employing gration in the bulk[32,53] and d-MXene as interface passivstor and the same organic additive chosen for the halide perovskite layer. humidity inhibitor.[30] During day labelled as (3) (Figure 8d, from Our strategy permits us to create an improved and continuous 424 h to 440 h) after more than 400 h of holding the devices at nexus between the MXene and the halide perovskite layers, in- their open circuit voltage, a clear distinction is observed between cluding the interface. Defect passivation analyses by thermal ad- the PSCs applying the MXene and the control device, an indica- mittance spectroscopy (TAS) indicates that shallow defects (re- tion of the beneficial effect of the simultaneously passivating the sponsible for device stability) are passivated at the bulk of the thin bulk of the perovskite with the H3pp ligand and the d-MXene films, while deep defects (responsible for device efficiency) are nanosheets. also passivated at the interfaces. As a result, champion MXene- During day (4) (Figure 8e, from 568 h to 584 h), which was based PSCs with ≈22% efficiency were obtained, in comparisson mostly a rainy and cloudy day (Figure S15), the beginning of the with the 20.56% obtained for the control device. Stability analy- analyses shows low device performance. The later increase of the ses under any analyzed condition (dark, continuous light irradi- sun irradiance intensity, between 574 h -578 h, showed excellent ation and real outdoor analysis) demonstrates that the enhance- recovery for the PSC when the MXene interlayer was employed. ment of the PSCs lifespan is always observed when the MXene In order to demonstrate any effect of hydrophobicity in our layer is employed. Especially relevant is the improvement in the samples that could explain the enhanced stability of the solar device’s lifespan observed under real operational outdoor condi- cells, we carried out contact angle measurements of the the per- tions (ISOS-O) where the solar cells showed a T80 at ≈600 h. To ovskite thin films with different functionalizations employing our knowledge, this is the first report of the stability analysis of the pure H3pp, d-MXene and MXene:H3pp. Figure S17 (Sup- MXene-based PSCs carried out under outdoor (ISOS-O) condi- porting Information) shows the results of contact angle mea- tions. Our results indicate that MXenes can pave the way towards surements. All modified samples show a slight increase in the highly stable perovskite solar cells. We believe our findings can contact angle of the water droplet independently of the pres- have an enormous impact on perovskite solar cell stability, which ence of the H3pp or the MXene, with average values of 65.41 is one of the main issues preventing the deployment of this pho- (±0.65)°, 64.87 (±0.27)°, 64.72°(±0.56)° for H3pp, d-MXene and tovoltaic technology. Adv. Energy Mater. 2023, 13, 2301959 2301959 (14 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Supporting Information [8] L. Yin, C. Liu, C. Ding, C. Zhao, I. Z. Mitrovic, E. G. Lim, H. Wang, Y. Sun, Y. Han, Z. Li, L. Yang, C. Q. Ma, C. Zhao, Cell Reports Phys. Sci. Supporting Information is available from the Wiley Online Library or from 2022, 3, 100905. the author. [9] V. Natu, J. L. Hart, M. Sokol, H. Chiang, M. L. Taheri, M. W. Barsoum, Angew. Chemie – Int. Ed. 2019, 58, 12655. [10] J. Ji, L. Zhao, Y. Shen, S. Liu, Y. Zhang, FlatChem 2019, 17, 100128. Acknowledgements [11] S. Pescetelli, A. Agresti, G. Viskadouros, S. Razza, K. Rogdakis, I. Kalogerakis, E. Spiliarotis, E. Leonardi, P. Mariani, L. Sorbello, M. The authors give thanks to the Spanish State Research Agency for the grant Pierro, C. Cornaro, S. Bellani, L. Najafi, B. Martín-García, A. E. Del Rio, Self-Power (PDI2022-143344OB-100 / AEI / 10.13039/501100011033). R. Castillo, M. Oropesa-Nuñez, S. Prato, M. L. Maranghi, A. Parisi, R. This work was part of the project PCI2020-112185, granted by MCIN/AEI/ Sinicropi, F. Basosi, E. Bonaccorso, A. Kymakis, C. Di, Nat. Energy 10.13039/501100011033 project and the European Union “NextGenera- 2022, 7, 597. tionEU”/PRTR. To the OrgEnergy Excelence Network (CTQ2016-81911- [12] M. Mozafari, M. Soroush, Mater. Adv. 2021, 2, 7277. REDT), to the Agencia de Gestió d’Ajuts Universitaris i de Recerca (AGAUR) for the support to the consolidated Catalonia research group [13] Z. Guo, L. Gao, Z. Xu, S. Teo, C. Zhang, Y. Kamata, S. Hayase, T. Ma, 2021 SGR 01617 and the Xarxa d’R+D+I Energy for Society (XRE4S). Small 2018, 14, 1802738. Part of this work was under Materials Science Ph.D. Degree for A.P. [14] Y. Zhao, X. Zhang, X. Han, C. Hou, H. Wang, J. Qi, Y. Li, Q. Zhang, and K.T. of the Universitat Autonoma de Barcelona (UAB, Spain). Chem. Eng. J. 2021, 417, 127912. ICN2 was supported by the Severo Ochoa program from Spanish [15] Y. Li, Y. Lu, X. Huo, D. Wei, J. Meng, J. Dong, B. Qiao, S. Zhao, Z. Xu, MINECO (grant no. SEV-2017-0706) and was funded by the CERCA D. Song, RSC Adv. 2021, 11, 15688. Programme/Generalitat de Catalunya. The ICN2 was supported by the [16] Y. Wang, P. Xiang, A. Ren, H. Lai, Z. Zhang, Z. Xuan, Z. Wan, J. Severo Ochoa Centres of Excellence programme, Grant CEX2021-001214- Zhang, X. Hao, L. Wu, M. Sugiyama, U. Schwingenschlögl, C. Liu, Z. S, funded by MCIN/AEI/10.13039.501100011033. The authors thank the Tang, J. Wu, Z. Wang, D. Zhao, ACS Appl. Mater. Interfaces 2020, 12, Spanish MINECO through the Severo Ochoa Centers of Excellence Pro- 53973. gram under grant no. SEV-2017-0706 for the postdoctoral contract to M.K. [17] D. Saranin, S. Pescetelli, A. Pazniak, D. Rossi, A. Liedl, A. Yakusheva, F.G. would like also to thank the supporting fund from the Swedish Energy L. Luchnikov, D. Podgorny, P. Gostischev, S. Didenko, A. Tameev, D. Agency (P2022-00756). Lizzit, M. Angelucci, R. Cimino, R. Larciprete, A. Agresti, A. Di Carlo, Nano Energy 2021, 82. [18] J. Ge, W. Li, X. He, H. Chen, W. Fang, X. Du, Y. Li, L. Zhao, Mater. Conflict of Interest Today Energy 2020, 18, 100562. [19] A. Agresti, A. Pazniak, S. Pescetelli, A. Di Vito, D. Rossi, A. Pecchia, The authors declare no conflict of interest. M. Auf der Maur, A. Liedl, R. Larciprete, D. V. Kuznetsov, D. Saranin, A. Di Carlo, Nat. Mater. 2019, 18, 1228. [20] X. Chen, W. Xu, N. Ding, Y. Ji, G. Pan, J. Zhu, D. Zhou, Y. Wu, C. Chen, Data Availability Statement H. Song, Adv. Funct. Mater. 2020, 30, 2003295. [21] V. S. N. Chava, P. S. Chandrasekhar, A. Gomez, L. Echegoyen, S. T. The data that support the findings of this study are available from the cor- Sreenivasan, ACS Appl. Energy Mater. 2021, 4, 12137. responding author upon reasonable request. [22] A. Yakusheva, D. Saranin, D. Muratov, P. Gostishchev, H. Pazniak, A. Di Vito, T. S. Le, L. Luchnikov, A. Vasiliev, D. Podgorny, Small 2022, 18, 2201730. Keywords [23] A. S. R. Bati, M. Hao, T. J. Macdonald, M. Batmunkh, Y. Yamauchi, L. Wang, J. G. Shapter, Small 2021, 17, 2101925. additive engineering, MXenes, passivation, perovskite solar cells, stability [24] L. Yang, D. Kan, C. Dall’Agnese, Y. Dall’Agnese, B. Wang, A. K. Jena, Y. Wei, G. Chen, X. F. Wang, Y. Gogotsi, T. Miyasaka, J. Mater. Chem. Received: June 26, 2023 A 2021, 9, 5016. Revised: September 7, 2023 [25] Y. Niu, C. Tian, J. Gao, F. Fan, Y. Zhang, Y. Mi, X. Ouyang, L. Li, J. Li, Published online: October 15, 2023 S. Chen, Nano Energy 2021, 89, 106455. [26] Y. Zhao, B. Li, C. Tian, X. Han, Y. Qiu, H. Xiong, K. Li, C. Hou, Y. Li, H. Wang, Chem. Eng. J. 2023, 143862. [27] T. Chen, G. Tong, E. Xu, H. Li, P. Li, Z. Zhu, J. Tang, Y. Qi, Y. Jiang, J. [1] G. P. Neupane, T. Yildirim, L. Zhang, Y. Lu, Adv. Funct. Mater. 2020, Mater. Chem. A 2019, 7, 20597. 30, 2005238. [28] X. Yao, Z. Qi, P. Yang, J. Li, W. Yang, Chem. Eng. J. 2023, 144315. [2] C. F. Du, X. Zhao, Z. Wang, H. Yu, Q. Ye, Nanomaterials 2021, 11, 166. [29] M. V. Khenkin, E. A. Katz, A. Abate, G. Bardizza, J. J. Berry, C. Brabec, [3] S. Shah, M. Sayyad, K. Khan, J. Sun, Z. Guo, Nanomaterials 2021, 11, F. Brunetti, V. Bulović, Q. Burlingame, A. Di Carlo, R. Cheacharoen, 2151. Y. B. Cheng, A. Colsmann, S. Cros, K. Domanski, M. Dusza, C. J. Fell, [4] L. Damptey, B. N. Jaato, C. S. Ribeiro, S. Varagnolo, N. P. Power, S. R. Forrest, Y. Galagan, D. Di Girolamo, M. Grätzel, A. Hagfeldt, E. V. Selvaraj, D. Dodoo-Arhin, R. V. Kumar, S. P. Sreenilayam, D. von Hauff, H. Hoppe, J. Kettle, H. Köbler, M. S. Leite, S. (Frank) Liu, Brabazon, V. K. Thakur, S. Krishnamurthy, Glob. Challenges 2022, 6, Y. L. Loo, J. M. Luther, et al., Nat. Energy 2020, 5, 35. 2100120. [30] Y. Zhang, L. Xu, J. Sun, Y. Wu, Z. Kan, H. Zhang, L. Yang, B. Liu, B. [5] L. Yang, C. Dall’Agnese, Y. Dall’Agnese, G. Chen, Y. Gao, Y. Sanehira, Dong, X. Bai, Adv. Energy Mater. 2022, 12, 2201269. A. K. Jena, X. F. Wang, Y. Gogotsi, T. Miyasaka, Adv. Funct. Mater. 2019, [31] J. H. Heo, F. Zhang, J. K. Park, H. J. Lee, D. S. Lee, S. J. Heo, J. M. 29, 1905694. Luther, J. J. Berry, K. Zhu, S. H. Im, Joule 2022, 6, 1672. [6] W. J. Sun, Y. Y. Zhao, X. F. Cheng, J. H. He, J. M. Lu, ACS Appl. Mater. [32] H. Xie, Z. Wang, Z. Chen, C. Pereyra, M. Pols, K. Gałkowski, M. Interfaces 2020, 12, 9865. Anaya, S. Fu, X. Jia, P. Tang, D. J. Kubicki, A. Agarwalla, H. S. Kim, D. [7] S. Jolly, M. P. Paranthaman, M. Naguib, Mater. Today Adv. 2021, 10, Prochowicz, X. Borrisé, M. Bonn, C. Bao, X. Sun, S. M. Zakeeruddin, 100139. L. Emsley, J. Arbiol, F. Gao, F. Fu, H. I. Wang, K. J. Tielrooij, S. D. Adv. Energy Mater. 2023, 13, 2301959 2301959 (15 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH 16146840, 2023, 44, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/aenm.202301959 by Readcube (Labtiva Inc.), Wiley Online Library on [24/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.advenergymat.de Stranks, S. Tao, M. Grätzel, A. Hagfeldt, M. Lira-Cantu, Joule 2021, 5, [44] C. Wu, W. Fang, Q. Cheng, J. Wan, R. Wen, Y. Wang, Y. Song, M. Li, 1246. Angew. Chem., Int. Ed. 2022, 61, e202210970. [33] W. Feng, H. Luo, S. Zeng, C. Chen, L. Deng, Y. Tan, X. Zhou, S. Peng, [45] M. Karimipour, S. Khazraei, B. J. Kim, G. Boschloo, E. M. J. H. Zhang, Mater. Chem. Front. 2018, 2, 2320. Johansson, ACS Appl. Energy Mater. 2021, 4, 14080. [34] Y. Li, X. Zhou, J. Wang, Q. Deng, M. Li, S. Du, Y. H. Han, J. Lee, Q. [46] S. Cacovich, G. Vidon, M. Degani, M. Legrand, L. Gouda, J.-B. Puel, Y. Huang, RSC Adv. 2017, 7, 24698. Vaynzof, J.-F. Guillemoles, D. Ory, G. Grancini, Nat. Commun. 2022, [35] K. Chen, K. Yan, Q. Xie, H. Zhu, X. Li, Z. Dong, G. Yuan, J. Zhang, Y. 13, 2868. Cong, Res. Chem. Intermed. 2022, 48, 4443. [47] D. Hong, J. Li, S. Wan, I. G. Scheblykin, Y. Tian, J. Phys. Chem. C 2019, [36] Y. Cao, Q. Deng, Z. Liu, D. Shen, T. Wang, Q. Huang, S. Du, N. Jiang, 123, 12521. C. Te Lin, J. Yu, RSC Adv. 2017, 7, 20494. [48] H. Yang, Y. Shen, R. Zhang, Y. Wu, W. Chen, F. Yang, Q. Cheng, H. [37] A. Sarycheva, Y. Gogotsi, Chem. Mater. 2020, 32, 3480. Chen, X. Ou, H. Yang, Adv. Energy Mater. 2022, 2202207. [38] E. Berger, Z.-P. Lv, H.-P. Komsa, J. Mater. Chem. C 2023, 11, 1311. [49] E. V. Péan, S. Dimitrov, C. S. De Castro, M. L. Davies, Phys. Chem. [39] R. Botta, P. Chindaudon, P. Eiamchai, M. Horprathum, S. Chem. Phys. 2020, 22, 28345. Limwichean, C. Chananonnawathorn, V. Patthanasettakul, A. [50] G. Yang, Z. Ren, K. Liu, M. Qin, W. Deng, H. Zhang, H. Wang, J. Liang, Jomphoak, N. Nuntawong, J. Raman Spectrosc. 2019, 50, 1817. F. Ye, Q. Liang, Nat. Photonics 2021, 15, 681. [40] M. C. Zenobi, C. V. Luengo, M. J. Avena, E. H. Rueda, Spectrochim. [51] C. Jiang, J. Zhou, H. Li, L. Tan, M. Li, W. Tress, L. Ding, M. Grätzel, C. Acta – Part A Mol. Biomol. Spectrosc. 2008, 70, 270. Yi, Nano-Micro Lett. 2022, 15, 12. [41] A. Barth, Biochim. Biophys. Acta (BBA)-Bioenergetics 2007, 1767, 1073. [52] J. Zhu, E. Ha, G. Zhao, Y. Zhou, D. Huang, G. Yue, L. Hu, N. Sun, Y. [42] X. Jin, L. Yang, X.-F. Wang, Nano-Micro Lett. 2021, 68, 1. Wang, L. Y. S. Lee, Coord. Chem. Rev. 2017, 352, 306. [43] A. S. R. Bati, A. A. Sutanto, M. Hao, M. Batmunkh, Y. Yamauchi, L. [53] M. Prete, M. V. Khenkin, D. Glowienka, B. R. Patil, J. S. Lissau, I. Wang, Y. Wang, M. K. Nazeeruddin, J. G. Shapter, Cell Reports Phys. Dogan, J. L. Hansen, T. Leißner, J. Fiutowski, H.-G. Rubahn, ACS Appl. Sci. 2021, 2, 100598. Energy Mater. 2021, 4, 6562. Adv. Energy Mater. 2023, 13, 2301959 2301959 (16 of 16) © 2023 The Authors. Advanced Energy Materials published by Wiley-VCH GmbH