Halogen Bonding in Perovskite Solar Cells: A New Tool for Improving Solar Energy Conversion

Abstract Hybrid organic–inorganic halide perovskites (HOIHPs) have recently emerged as a flourishing area of research. Their easy and low‐cost production and their unique optoelectronic properties make them promising materials for many applications. In particular, HOIHPs hold great potential for next‐generation solar cells. However, their practical implementation is still hindered by their poor stability in air and moisture, which is responsible for their short lifetime. Optimizing the chemical composition of materials and exploiting non‐covalent interactions for interfacial and defects engineering, as well as defect passivation, are efficient routes towards enhancing the overall efficiency and stability of perovskite solar cells (PSCs). Due to the rich halogen chemistry of HOIHPs, exploiting halogen bonding, in particular, may pave the way towards the development of highly stable PSCs. Improved crystallization and stability, reduction of the surface trap states, and the possibility of forming ordered structures have already been preliminarily demonstrated.


Introduction
"The fun in science lies not in discovering facts,b ut in discovering new ways of thinking about them." [1] All of the modern advancements in science and technology are never stand-alone.I nm ost cases,t hey have been made possible thanks to fundamental discoveries,which often are hundreds of years old. Theessence of scientific progressquoting Sir Lawrence Bragg,1915 Nobel laureate in physicsis trying to find new ways of thinking about well-known fundamental facts. [1] In this respect, neither halogen bonding (XB) nor perovskites are new in science.T hey are consolidated tools in materials science and engineering,a nd their innovative applications opened up new possibilities for the design of advanced functional materials. [2][3][4][5][6] In this Minireview we will first introduce an overview of the unique features of XB-directionality,t unability,h ydrophobicity,a nd donor atom size-that have driven its exponential growth in the last few years.T his proves the great impact of XB in several fields,opening up new opportunities for the design of supramolecular functional systems with awide range of applications.Then, we will see how the recent combination between XB and perovskites may help overcome some of the critical issues that still prevent the practical application of perovskite solar cells.P reliminary data in this field have,infact, already demonstrated the efficacy of XB in improving crystallization and stability,a nd reducing surface trap states.Abright future can therefore be foreseen for XB as anew tool for improving solar energy conversion.

The Halogen Bond
Within the panel of chemical interactions available to chemists,t he XB,i .e., the non-covalent interaction involving halogen atoms as electrophilic sites, [7] has experienced an explosive growth in the last years.Essentially unknown before X-ray crystallographic studies on the Br 2 ···O(CH 2 CH 2 ) 2 O adduct reported by O. Hassel in 1954, [8] XB has quickly become aunique tool for molecular recognition processes,as shown by the increasing number of papers published per year with the topic "halogen bonding" ( Figure 1A).
Ac areful look at these papers shows the great evolution of XB research. At the beginning, it was just as cientific curiosity,e xploited to construct beautiful supramolecular adducts. [9,10] However,o ver time the attention has shifted towards the new functionalities induced in the final adducts. [11,12] Nowadays,XBisexploited for various applications at the boundary between chemistry,nanotechnology,physics, medicinal chemistry,and molecular biology. [5] This interdisciplinary character suggests that there is still much to explore about XB and research on new applications and advanced functionalities induced by XB will surely continue to grow in the next years.
Thereasons for such an explosive growth lie in the unique features of XB,w hich cannot be easily met by other noncovalent interactions.F irst of all, the anisotropic distribution of electron density in covalently bound halogen atoms allows the formation of highly directional interactions with nucleophiles (XB through the s-hole) and electrophiles (at the negative belt perpendicular to the covalent bond formed by the halogen) as shown in Figure 1B. [13] This high directionality,t ogether with the tunable interaction strength enabled by the halogen atom selection, makes the XB an effective tool Hybrid organic-inorganic halide perovskites (HOIHPs) have recently emerged as aflourishing area of research. Their easy and low-cost production and their unique optoelectronic properties make them promising materials for many applications.I np articular,H OIHPs hold great potential for nextgeneration solar cells.However,their practical implementation is still hindered by their poor stability in air and moisture,whichisresponsible for their short lifetime.Optimizing the chemical composition of materials and exploiting non-covalent interactions for interfacial and defects engineering,a swell as defect passivation, are efficient routes towards enhancing the overall efficiency and stability of perovskite solar cells (PSCs). Due to the richh alogen chemistry of HOIHPs,e xploiting halogen bonding,i np articular,m ay pave the waytowards the development of highly stable PSCs.Improved crystallization and stability,reduction of the surface trap states,and the possibility of forming ordered structures have already been preliminarily demonstrated.
for controlling the self-assembly of molecular building blocks and fine-tuning their functional properties.M oreover,t he frequent presence of fluorinated segments in the XB-donor not only boosts interaction strength, but also increases the hydrophobicity of the final supramolecular adducts,g iving protection against humidity and boosting material stability. [12] Finally,the high polarizability of the heaviest halogen atoms, although it may pose some steric limitations for some applications,onthe other hand, is beneficial for constructing efficient all-organic solid-state electronic materials.I nf act, the presence of halogen atoms simultaneously allows to control the molecular packing through XB and modulate the LUMO level, lowering the HOMO-LUMO gap,t hus promoting charge transport. [14][15][16][17][18] Further,t he bare size of halogen atoms may significantly alter the light-emitting properties of halogenated dyes by promoting singlet-to-triplet intersystem crossing [19] and affording high phosphorescence quantum yields. [20] Last, the high directionality of XB may determine the obtainment of non-centrosymmetric structures promoting the emergence of second-order nonlinear optical (NLO) properties. [21] More recently,X Bh as shown its great potential in designing supramolecular photovoltaic materials with optimized morphology and charge transport and promoting efficient dye regeneration in dye-sensitized solar cells. [22][23][24] Latest research in the field proved that XB can be successfully applied to ameliorate the performances of perovskite solar cells (PSCs), bringing innovative photovoltaic systems able to fulfil the demand for clean and sustainable energy. [25] 3. Hybrid Organic-Inorganic Halide Perovskites Similar to XB,perovskites have been known since the 19 th century.H owever,t he first evidence of their use as semiconductors for optoelectronic applications dates back only to the late 20 th century. [26,27] In particular, hybrid organicinorganic halide perovskites (HOIHPs) have emerged as promising candidates for application as lasers,l ight-emitting diodes,field-effect transistors,photodetectors,and solar cells, thanks to their easy and cheap production, high photoluminescence quantum yields,m ulticolor emission, and excellent excitonic and charge carrier properties ( Figure 1A). [28][29][30][31][32][33][34] In particular,H OIHPs hold great potential for next-generation solar cells.T he first application of perovskites as al ight harvester in solar cells was reported in 2009, [35] and since then, PSC performance has improved dramatically,r eaching efficiencies higher than 25 %. [36] Further enhancements of cell performances can be achieved by controlling crystal growth and surface interactions of the perovskite layer. [37][38][39] In this respect, due to the role of halide chemistry in HOIHPs, exploiting the unique features of XB in this field may be particularly useful.

XB Passivation of Halide Perovskites
Undercoordinated halide anions and migratory species may exist on the surface of perovskite grains.T hese superficial defects can trap positive charges and holes,p romoting non-radiative charge recombination. XB may provide passivation of these undercoordinated halide anions and, at the same time,i mprove material crystallinity and enhance the efficiency and stability of the devices.T he first example of perovskite surface passivation through XB was reported by Abate et al. [40] back at the dawn of the PSC era. Their work used iodopentafluorobenzene (IPFB) to passivate the undercoordinated halide ions,w hich acted as hole traps on the perovskite surface (Figure 2A). This treatment led to ad efinite performance improvement in devices.IPFB,indeed, can bind, through XB,t he I À ions shielding their electrostatic charge and reducing the accumulated charge at the interface, therefore suppressing hole recombination and promoting charge transfer. These results,l ater on, motivated Zhang et al. [41] to study the effect of XB passivation in PSCs from atheoretical point of view.T heir work confirmed the role of XB in passivating the CH 3 NH 3 PbI 3 perovskite surface and demonstrated its effectiveness in modulating the crystallinity of perovskite films.Upon adsorption of IPFB,while the inner layers of the perovskite see negligible changes,o nt he top layer, the Pb-I-Pb angle increases and the Pb À Ibond lengths  A) Schematic view of the XB between iodopentafluorobenzene (IPFB, XB-donor) and ageneric halide anion (X À = I À ,Br À ,Cl À , XB-acceptor). Adapted with permission from ref. [40].B)Optimized geometries of CH 3 NH 3 PbI 3 perovskites urface with (right) and without (left) IPFB functionalization, and the related effect on the lattice structure. Adapted with permission from ref. [41].C )The perovskite/ bromoacetate/TiO 2 tri-layer structure highlights( red dashed line) the XB between the bromoacetate molecule and the halide perovskite layer. Adapted with permissionf rom ref. [42].
are reduced, as shown in Figure 2B.I nt his way,t he PbI 6 surface octahedra, which generally adopt ad istorted geometry in the bare perovskite,u pon interaction with IPFB, become less distorted and more similar to the bulk structure.

XB-Donors as Organic Interfacial Modifiers
Recently,t he same group investigated the effects of haloacetate molecules (chloro-, bromo-, and iodoacetates) as bifunctional interfacial modifiers between TiO 2 and perovskite layers. [42] Haloacetates can act as bifunctional ligands, forming XBs with the halides of the perovskite through the halogen atom and covalent bonds with the TiO 2 substrate through the carboxylate group ( Figure 2C).
This interaction enhances the interfacial contact between the two materials,t hus reducing lattice mismatch and improving the interfacial charge transfer properties.S imilar results were also found by Dai et al., [43] who inserted abifunctional interfacial modifier between SnO 2 and perovskite.Such modifier has Si(OCH 3 ) 3 as anchoring group to the SnO 2 and an iodine-terminated alkyl chain for the interaction with the perovskite.Also in this case,the presence of XB improved the interfacial adhesion affording higher efficiency and improved operational stability.

XB-Donors as Crystallization Modulators
Experimental evidence of the role of XB-donors in modulating the crystallization of perovskite films and improving the resulting photovoltaic performances has been provided by Bi et al. [44] They added either diiodoperfluorobutane (I(CF 2 ) 4 I) or diiodobutane (I(CH 2 ) 4 I) as XB-donors in the CH 3 NH 3 PbI 3 perovskite precursor solution (Figure 3A). Their results revealed that upon XB with these additives,the concentration of free I À anions in the precursor solution is lowered, thus suppressing the formation of iodoplumbate complexes (e.g., PbI 3 À ,P bI 4 2À ,a nd highercoordination compounds), which generally hinder the crystallization of perovskites.I mproved crystallinities and morphologies were obtained for the perovskite film, which showed am ore compact and uniform surface ( Figure 3B). This is reflected in longer photoluminescence (PL) decay and ahigher steady-state PL peak, suggesting further suppression of the surface traps.S pace-charge-limited current measurements indicated better charge transport and suppressed recombination. Moreover,asubstantial reduction of hysteresis was noticed, which agrees with limited ion mobility under electric field due to suppressed non-stoichiometry and surface charge trapping. Finally,t he improvements mentioned above resulted in enhanced environmental stability for the devices prepared with XB additives.
Theg eneral applicability of this concept was further demonstrated by Ruiz-Preciado et al. [45] by introducing 1,2,4,5-tetrafluoro-3,6-diiodobenzene (TFDIB) as bifunctional supramolecular modulator (Figure 4). TFDIB was chosen for its hydrophobicity and capability to strongly bind perovskite halides through XB.Ifadded to the precursor solution or used as an interlayer between perovskite and charge transport materials,itresulted in improved V oc and enhanced  long-term operational stability,w ith 80 %e fficiency retained after 500 hofcontinuous illumination. Theimproved stability was ascribed to the hydrophobic nature of the molecule and to its effect on the stabilization of the FAPbI 3 a-phase (FA = formamidinium), as shown by X-ray diffraction measurements.T his work also provided af ascinating insight into the atomic-level mode of interaction of TFDIB and perovskite, thanks to detailed NMR analyses and DFT calculations, which evidenced the role of crosslinking XB.T heir results showed that the modulator preferably coordinates PbI 2 or MAI (MA = methylammonium) through both symmetric and asymmetric coordination binding modes.A tt he same time, there was no evidence of the formation of the FAI-TFDIB complex. As imilar crosslinking mechanism has also been proposed for 2-bromo-6-fluoronaphthalene introduced as an interfacial modulator for printable hole-conductor-free mesoscopic PSCs through post-treatment. [46]

XB Self-Assembled Monolayers
Molecules capable of coordinating though XB are also excellent candidates to form self-assembling monolayers (SAMs), which are of particular interest because of their stability and ordered distribution. An example of the advantages following the employment of XB-driven SAMs on perovskites is given by the work of Wolff et al., [47] where iodoperfluoroalkanes of different chain lengths (IPFC n with n = 8, 10, 12) were used to functionalize the surface of perovskites in inverted PSCs.The results showed asignificant V oc improvement, which the authors ascribed to reducing nonradiative recombination and gain in charge separation by looking at absolute PL, time resolved-PL, and surface photovoltage measurements.Moreover,increased stability was also reported for the treated devices that were able to withstand harsh stress conditions like 250 ho fc ontinuous illumination at 85 8 8C, and retain 95 %efficiency after 4months of storage in ambient conditions. IPFC 10 has also been employed by Canil et al. in am ore recent work showing that it is possible to exploit SAMs to tune the perovskite energy levels continuously.T he authors demonstrated that it was possible to control the deposition kinetics by managing the molecules deposition parameters, i.e., solution concentration or dipping time.T hus the magnitude and direction of the energy levels shift. XB allowed obtaining ordered and stable monolayers,proving very useful for this kind of application. It enhances the ability to control the deposition and fine-tuning the perovskite energy level alignment. [48] This concept was later on further developed by the same group,w ho obtained enhanced stability and performance of the devices by directly incorporating in the hole transport layer (HTL) XB-donor groups able to interact with the perovskite. [49] Thanks to the presence of XB,t he HTL can form amore ordered and compact layer, resulting in astronger interfacial dipole,reduced energetic offset for hole transport, and suppression of recombination processes.T he improved interface also increased the resilience against moisture and solvent, affording outstanding operational stability with ap rojected lifetime more than two-fold the standard devices. [50]

2D-HOIHPs
Finally,X Bh as also been exploited to modulate the crystal structure of 2D-HOIHPs,w hich consist of single perovskite sheets separated by organic cations. [51] According to their crystal structure,2 D-HOIHPs can be classified as either Dion-Jacobson (DJ) or Ruddlesden-Popper (RP) structures ( Figure 5A). RP-HOIHPs feature monovalent organic cations interdigitating between adjacent inorganic layers of MX 6 octahedra with as taggered arrangement. Differently,the DJ-HOIHPs feature divalent organic cations between the layers,w hich adopt an eclipsed arrangement. Tr emblay et al. [51] reported the crystal structures of afamily of 2D hybrid structures (4-Y-C 6 H 4 CH 2 NH 3 ) 2 PbI 4 (where Y = F, Cl, Br,I ), and demonstrated that despite the use of am onovalent organic cation, for Y = F, Cl, and Br the resulting 2D perovskites gave aDJstructure.Conversely,when iodine was used, i.e., Y= I, for which stronger XB is expected, an RP-like structure was obtained.
This study suggests that tuning the interaction strength between organic and inorganic layers allows controlling the perovskite stacking pattern in HOIHPs.Asimilar approach has been adopted by Fu et al., [52] who recently used am onovalent tetrafluoro iodoaromatic synthon to increase the XBdonor ability of the cation in a2D/3D hybrid perovskite.The interfacial XB between 2D and 3D perovskites anchors the iodide anions at the grain boundaries,which suppresses phase separation and significantly improves the long-term operational stability of the photovoltaic devices.  Chemie

Conclusion
As is often the case for rapidly emerging and growing fields,m any researchers from different backgrounds join in, attracted by the rapid pace of citations growth. While we recognize that some multidisciplinary areas,s uch as PSCs, may benefit from inputs from different backgrounds,itisour opinion it has always to be done in keeping with the established understanding of phenomena and correct use of related terminology.O ur critical review of current literature about the use of XB in HOIHPs has revealed that phenomena misinterpretation and term misuses may occur.
As an example,T ang et al. reported that N···Br halogen bonds,b etween N-doped-C heterostructures and the Br À atoms from CsPbBr 3 (CPB), chemically immobilize CPB on N-C layers,e nhancing the aqueous stability of CPB@Co 3 O 4 / N-C by prohibiting CPB decomposition. [53] In our opinion, it is more likely that this interaction can be ascribed to agenuine hydrogen bond, where the nucleophilic Br À ion is behaving as the electron donor towards the pyrrolic Hatom. Paek et al., instead, reported that fluoro-substituted 2D materials bearing perfluoroaromatic synthons are more effective than monofluorinated ones,r esulting in strong interaction with the 3D perovskite,w hich induces ah ighly in-plane oriented growth of the crystals while preserving excellent hole transfer. [54] We found these results extremely interesting. However, they may be more likely related to the peculiar electron distribution in perfluorophenyl rings,w hich usually behave as electronacceptors in anion···p interactions. [55] Similar interactions were also wrongly attributed to XB when using 2,3,5,6tetrafluoro-7,7,8,8-tetracyano-quinodimethane (F4TCNQ) as interlayer between the HTL and perovskite active layer in PSCs. [56,57] In summary,w eh ave highlighted the potential that XB has to bring great and reproducible advantages for developing highly stable PSCs.C ommon features among most of the reported examples are improved crystallization and stability, reduction of the surface trap states,a nd the possibility of forming ordered structures and layers.T hese results are of utmost interest considering the variety of perovskite compositions and device structures and the sensitivity of the current materials to external factors.Nevertheless,there is still much research needed in both XB and HOIHPs.I np articular, an atomic/molecular understanding is highly needed for fully exploiting the advantages of the combination of these two rapidly growing research fields.However, the future of XB in HOIHPs looks bright.