Revealing the role of Pb 2+ in the stability of organic–inorganic hybrid perovskite CH 3 NH 3 Pb 1− x Cd x I 3: an experimental and theoretical study

This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication. Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available. You can find more information about Accepted Manuscripts in the Information for Authors. Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains. Accepted Manuscript www.rsc.org/pccp PCCP 1 Revealing the role of Pb2+ in the stability of organic-inorganic hybrid perovskite CH3NH3Pb1-xCdxI3. An experimental and theoretical study Javier Navas,* ,a Antonio Sánchez-Coronilla,* ,b Juan Jesús Gallardo, a Elisa I. Martín, c Norge C. Hernández, d Rodrigo Alcántara, a Concha Fernández-Lorenzo, a Joaquín Martín-Calleja a a Departamento de Química Física, Facultad de Ciencias, Universidad de Cádiz, E-11510 Puerto Real (Cádiz), Spain; b Departamento de Química Física, Facultad de Farmacia, Universidad de Sevilla, E-41012 Sevilla, Spain; c Departamento de Ingeniería Química, Facultad de Química, Universidad de Sevilla, E-41012 Sevilla, Spain; and d Departamento de Física Aplicada I, Escuela Técnica Superior de Ingeniería Informática, av. Reina Mercedes, Universidad de Sevilla, E- 41012 Sevilla, Spain. Corresponding Authors: * Javier Navas ([email protected]); Antonio Sánchez-Coronilla ([email protected]). Table of Contents entry Experimental and theoretical analysis revealing the role of Pb 2+ in the stability of organic-inorganic hybrid perovskites Page 1 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 2 Page 2 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 3 Abstract This paper presents the synthesis of organic-inorganic hybrid perovskite CH 3 NH 3 Pb 1- x Cd x I 3 . The effect of incorporating Cd 2+ or Pb 2+ on the stability of the perovskite structure was analysed from a theoretical and experimental viewpoint. The XRD results showed that the tetragonal perovskite structure was formed for values of x of up to 0.5, which seems to indicate that the presence of a considerable amount of Pb 2+ is necessary to stabilise the structure. In turn, UV-Vis spectroscopy showed how the presence of Cd 2+ led to a reduction in the optical band gap of the perovskite structure of up to 9% for CH 3 NH 3 Pb 0.5 Cd 0.5 I 3 with regards to the MAPbI 3 structure. Moreover, periodic-DFT calculations were performed to understand the effect of the increased concentration of Cd on the structural and electronic properties of MAPbI 3 perovskites. The analysis of both the ELF and the non-covalent interaction (NCI) index show the important role played by the Pb 2+ ions in stabilizing this kind of hybrid perovskite structures. Finally, the DOS analysis confirmed the experimental results obtained using UV-Vis spectroscopy. The theoretical band gap values decreased as the concentration of Cd increased. Page 3 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 4 1. Introduction Perovskites have many fascinating properties, including superconductivity, ferroelectricity, high thermoelectric power or optical properties. 1 Thanks to these properties, perovskites have been used as sensors, catalysts and optoelectronic material. 2-4 Nowadays, organic-inorganic perovskites based on halides are studied because they perform well in photovoltaic applications. 5-7 Perovskites have a specific crystal structure with the ABX 3 formula. The crystal structure of the ABX 3 compounds consist of a three-dimensional lattice of corner sharing BX 6 octahedra, where B is a bivalent metal cation, A is a monovalent organic cation that is small enough to fit into the structure, and X is a halogen. 8 A methyl ammonium cation ([CH 3 NH 3 ] + , MA) is the most usual monovalent organic cation in several applications. Moreover, interesting examples of this kind of perovskites are CH 3 NH 3 BX 3 , where B = Sn 2+ , Pb 2+ ; X = Cl - , Br - , I - . In turn, methyl ammonium lead iodide (MAPbI 3 ) has brought about a revolution in photovoltaic applications thanks to its high efficiency, which reaches 18%. 7 Thus, the crystalline structure and good photonic absorption are two vital factors in the high performance of this kind of compound in solar cells. The substitution of some elements in MAPbI 3 can make it possible to control the optical and electronic properties of perovskites, which can affect the performance of the devices. For example, the substitution of I - with Cl - or Br - has been widely studied; and doping with inorganic cations such as Sn 2+ , Sr 2+ , or Ca 2+ in Pb 2+ sites has been analysed to understand its effect on optical and electronic properties. 5, 9-12 Thus, this article studies the effect on the structural stability and electronic properties of the introduction of Pb 2+ or Cd 2+ ions in position B of an ABX 3 -type perovskite, where A = MA and X = I - . To do this, synthesized samples were characterised using elemental analysis techniques such as the CHNS technique or X-ray fluorescence (XRF) to corroborate the amount of Pb 2+ and Cd 2+ introduced; X-ray diffraction (XRD) Page 4 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 5 determined the crystalline phases present in the samples; X-ray photon spectroscopy (XPS) determined the chemical bonding states; and UV-Vis spectroscopy in the diffuse reflectance mode (DR-UV-Vis) was used to establish the optical band gap of the samples. In addition, from a theoretical perspective, periodic density functional theory (DFT-periodic) calculations were performed to rationalize the experimental information on this topic. Therefore, to explore the electronic and structural effects of increasing the concentration of Cd, the structures (MA) 4 Pb 4 I 12 , (MA) 4 Pb 3 CdI 12 , (MA) 4 Pb 2 Cd 2 I 12 (MA) 4 Pb 1 Cd 3 I 12 and (MA) 4 Cd 4 I 12 were optimised. The analysis of both the electron localization function (ELF) 13-17 and the non-covalent interaction (NCI) 18 shed light on the bonding interactions established within these hybrid systems, highlighting the important role of the Pb 2+ ions in the stabilization of this kind of hybrid structures. The density of State (DOS) and band structure results taking into account the spin-orbit coupling (SOC) effects were in agreement with the tendency forecast experimentally, showing a decrease in the optical band gap energy as the concentration of Cd 2+ increased. 2. Experimental and Computational section 2.1. Experimental 2.1.1. Reagents All the reagents were from commercial sources and used without further purification. The hydroidic acid (HI, 57 wt% in water) was from Aldrich; the methylamine (CH 3 NH 2 , 33 wt% in ethanol), diethyl ether (Et 2 O, purity ≥ 99.8%), led iodide (PbI 2 , purity 99%), cadmium iodide (CdI 2 , purity 99.999%), and γ-butyrolactone (purity ≥ 99%) were from Sigma-Aldrich. Page 5 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 6 2.1.2. Synthesis The synthesis of the perovskite phases was performed by means of the reaction of CH 3 NH 3 I and the nominal quantities of PbI 2 and CdI 2 . First, the CH 3 NH 3 I was synthesized following the procedure: 12, 19, 20 HI (10 mL, 0.076 mol) and CH 3 NH 2 (11.33 mL, 0.091 mol) were stirred in an ice bath for 2 h. Next, the mixture was evaporated at 50 ºC for 1 h, and CH 3 NH 3 I was obtained. Then, the solid was cleaned three times using Et 2 O. For the synthesis of the perovskite phase, a mixture was prepared of CH 3 NH 3 I (0.395 g) and the corresponding stoichiometric amount to obtain nominal concentrations of Cd (Cd/(Cd+Pb)) of 0, 25, 50, 75 y 100 at.% in γ-butyrolactone (2 mL) at 60 ºC for 18 h. 2.1.2. Characterization The elemental analysis was performed using two instrumental techniques. The CHNS technique was used to obtain the amount of C, H and N, using an elemental analyser supplied by Leco®, model CHNS932. Also, X-ray fluorescence (XRF) was performed to determine the amount of Pb, Cd and I using a sequential XRF spectrometer supplied by Bruker®, model Pioneer S4. X-ray Diffraction (XRD) was used to determine the crystalline phases in the samples. The patterns were recorded using a diffractometer supplied by Bruker®, model D8 Discover, with Cu-Kα radiation. The scan conditions were from 10 to 70º in 2θ with a resolution of 0.02º, 40 kV and 40 mA. The optical band gap (E g ) values were determined from diffuse reflectance UV-Vis (DR-UV-Vis) measurements. The spectra were collected by using a system assembled in our laboratory composed of an integrating sphere supplied by Spectra Tech®, a USB2000+ spectrometer supplied by Ocean Optics®, and a Xe lamp, model ASB-XE-175, supplied by Spectral Products®, as the illumination source. Moreover, X-ray photoelectron spectroscopy (XPS) was used to study the chemical bonding states and the chemical Page 6 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 7 compositions of the samples. The spectra were obtained using a Kratos® Axis UltraDLD spectrometer, with monochromated Al-Kα radiation (1486.6 eV) and 20 eV pass energy. The binding energy was given with an accuracy of 0.01 eV. Electrostatic charging effects were stabilised with the help of a specific device developed by Kratos®. 2.2. Computational framework DFT-periodic calculations were performed using the Vienna Ab Initio Simulation Package (VASP) 21-24 with the projector-augmented wave (PAW) method. 25, 26 The number of plane waves in VASP was controlled by a cut-off energy, chosen according to the pseudopotential and set in our calculations to E cut =500 eV to satisfactorily describe the system. 12 In this way, the value of the cut-off was chosen by increasing around 30% the highest value defined for all pseudopotentials used, i.e. carbon and nitrogen. The electron exchange and correlation were treated within the generalised gradient approximation (GGA). 27 In the case of GGA, Perdew-Burke-Ernzerhof (PBE) 27 functionals were used that provide geometrical structures and relative stabilities for hybrid perovskites in good agreement with experimental data, as reported elsewhere. 11, 28-30 Both the cell shape and atomic positions were optimised using a conjugate-gradient algorithm, where the iterative relaxation of atomic positions was stopped when the forces on the atoms were less than 0.01 eV/Å. Also, a Gaussian smearing with k B T=0.1365 eV was applied. Perovskite tetragonal phase was characterised by the I4/mcm space group. 30 The tag KSPACING determined the number of k-points in the mesh. A value of 0.4 for sampling the Brillouin zone for perovskite bulk was enough to obtain negligible changes in the optimised cell parameters and energy. The resulting cell parameters were Page 7 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 8 a = 8.912 Å, c = 12.707 Å for the tetragonal phase, in agreement with the experimental values (a = 8.874 Å, c = 12.671 Å). 30 The (MA) 4 Pb 4 I 12 cell was optimised as a model for the tetragonal phase of perovskite. The incorporation of each proportion of Cd into the structure was performed in substitutional positions. The Cd atom substitutes the Pb atom in the perovskite structure taking into account the equivalent sites. The structures (MA) 4 Pb 3 CdI 12 , (MA) 4 Pb 2 Cd 2 I 12 (MA) 4 Pb 1 Cd 3 I 12 and (MA) 4 Cd 4 I 12 were optimised. The selected structures were adequate to represent qualitatively the experimental concentrations within DFT with a periodic boundary condition framework. It is well known the structural disorder associated to the MA rotation in this kind of hybrid perovskites. 5, 30, 31 In a previous work we tested different orientations of MA within the tetragonal structure, reporting the most stable configuration for each doped structure. 12 This structure has been used as origin for optimizing the structures used in this work. The density of states (DOS) and projected density of states (PDOS) for the relaxed structures were obtained using the tetrahedron method with Blöchl corrections 32 and a KSPACING of 0.3. For all the systems, DFT+U 33 calculations were performed to take into account the on-site Coulomb interaction. A U value of 5 eV was used for d orbitals. 34 Recent works have confirmed a relevant SOC in MAPbI 3 due to the strong SOC on Pb atoms leading to a strong band-gap reduction by inducing a large splitting of the conduction levels while the valence band levels are nearly unaffected. 35, 36 Although more computationally demanding, the inclusion of many body perturbation theory, within the GW approach 37, 38 and SOC lead to a good correlation between the band gap values calculated and the experimental ones, 11, 35 because of GW corrections are large and in opposite direction to SOC effect. 35 Nonetheless, a computationally viable Page 8 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 15 Figure 2. UV-Vis spectra, in diffuse reflectance mode, of the samples synthesized. Figure 2B shows the spectra for the samples with a nominal composition of Cd of 75%, and 100%, and the spectrum of CdI 2 used as reagent in the synthesis. For these samples, the presence of the perovskite absorption band is not observed. Moreover, the presence of the typical absorption band for CdI 2 at about 390 nm, seen in the DR-UV-Vis spectrum for CdI 2 used as reagent, is observed in the samples with a nominal composition of Cd of 75 at.% and 100 at.%. Moreover, for the sample 75 at.% of Cd, it is possible to observe the typical absorption band for PbI 2 . So, these spectra are further evidence of the non-reaction between CH 3 NH 3 I, PbI 2 and CdI 2 for higher proportions of Cd, or when PbI 2 is not present, for obtaining perovskites structures. Thus, the results seem to indicate that a high proportion of Cd 2+ in the samples (75 at.% and 100 at.% for the Cd/(Cd+Pb ratio)) impedes the formation of the perovskite structure, the presence of the reactants used for its synthesis being observed. The presence of CH 3 NH 3 I is not detected with UV-Vis, as its absorption band overlaps the one assigned to the CdI 2 . The UV-Vis spectrum, in diffuse reflectance mode, for CH 3 NH 3 I used as a reagent is shown in Figure S1 in the ESI. Page 15 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 16 3.4. XPS XPS spectra were recorded to study the oxidation and bonding states of the elements in the samples. The samples with a nominal composition of 25 and 75 at.% of Cd were studied. Figure S2 in the ESI shows the general spectra for these samples. Moreover, Figure 3A shows the Pb 4f spectra for these samples. The Pb 4f 7/2 signals appear at the binding energy (BE) of about 139 eV, in good agreement with the values reported previously. 56, 57 Moreover, the spectra show well-separated spin-orbit components, with a separation of about 5 eV in both cases, which is typical for Pb 2+ . Figure 3. XPS spectra for: (A) Pb 4f; (B) Pb 4d and Cd 3d; and (C) I 3d for the samples with a nominal composition of 25 and 75 at.% Cd. Page 16 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 17 Figure 3B shows the spectrum of the Cd 3d. The signal of the Cd 3d 3/2 completely overlaps that of the Pb 4d 5/2 at approximately 413 eV. This signal is asymmetrical due to the presence of several contributions to it. Furthermore, the signal changes shape depending on the Cd content in the samples. The BE of Cd 3d 5/2 was found at about 406.2 eV, in agreement with the literature for Cd 2+ . 56 The intensity ratio between the signals shown in Figure 3B varies due to the different amounts of Cd in the samples. The signal observed at a BE of about 402.5 eV in Figure 3B is assigned to N 1s. Figure 3C shows the I 3d 3/2 and I 3d 5/2 spectra for the samples. The spectra show wellseparated spin-orbit components, with a separation of about 11.5 eV in both cases, which is in agreement with values reported for I - . 56 Moreover, a shift towards a lower BE is observed in the samples with a higher nominal composition of Cd, which is coherent for compounds of a more ionic nature. So, this typical shift suggests small modifications in the oxidation state of I - occurred in accordance with the amount of Cd 2+ in the samples. In fact, the decrease in the optical band gap for the perovskite samples suggests that the presence of Cd slightly modifies the interactions of the ions in the lattice, which leads to changes in the band gap and geometrical distortions in the lattice, as theoretical calculations show. 3.5. Structure and local geometry analysis From a theoretical perspective, an analysis was performed into the effect of increasing the concentration of Cd on the structural and electronic properties of MAPbI 3 perovskites. Likewise, a study was performed of the tetragonal structures of MAPb 1- x Cd x I 3 with x=0.00, 0.25, 0.50, 0.75 and 1.00. The local geometry for the most stable configuration of the optimised structures is shown in Figure 4. Our theoretical results show that the incorporation of Cd 2+ ions caused distortions in the structure. Generally, the inclusion of Cd 2+ ions produced a Page 17 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 18 Figure 4. The local geometry in 2D and 3D and ELF contour plots for the tetragonal MAPb 1-x Cd x I 3 perovskite with x=0.00, 0.25, 0.50, 0.75 and 1.00. shortening of the Cd-I distances compared with the Pb-I ones. This change is more noticeable for the MAPb 0.75 Cd 0.25 I 3 and MAPb 0.50 Cd 0.50 I 3 structures ( c.a. 3-5% in both) than for the MAPb 0.25 Cd 0.75 I 3 structures ( c.a. 1-2%) and MACdI 3 ( c.a. 1%). These modifications to the Cd-I and I-Pb distances are reflected in the changes in the values of the Pb-I-Cd and Cd-I-Cd angles (Figure 4, Table 4). As an example and for clarity, these angles are represented as α, β, γ, and δ in Figure 4. The ESI includes the cut corresponding with the front face of the tetragonal structure (Figure S3) and a table with its angles (Table S2). As Table 4 shows, in general the α, β, γ and δ angles increase in accordance with the concentration of Cd, reaching their highest values in the MACdI 3 structure. In turn, the γ and δ angles in the MACdI 3 structure are c.a. 10 and 5º bigger than the same angles for the MAPbI 3 structure. These geometry results lead us to believe that the structural tensions in the theoretical structures MAPb 0.25 Cd 0.75 I 3 and MACdI 3 are greater than in the structures with low concentrations of Cd 2+ ions. Page 18 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 19 Therefore, it would seem that a certain concentration of Pb 2+ ions must be present to stabilise the structure, as shown experimentally and as the results from ELF and NCI indicate, and which will be discussed below. Table 4. Geometrical parameters, identified in Figure 4, of the structures studied. Structure α β γ δ MAPbI 3 157.11 155.03 156.52 158.35 MAPb 0.75 Cd 0.25 I 3 157.98 155.22 158.38 160.33 MAPb 0.50 Cd 0.50 I 3 158.71 155.33 160.02 161.95 MAPb 0.25 Cd 0.75 I 3 159.03 155.30 161.19 163.19 MACdI 3 159.85 157.02 166.40 163.48 The lower part of Figures 4 and S3 show the ELF plots corresponding to the geometries represented in 2D. Figure 4 shows how the contour plot for Pb is different to that of Cd. It shows that for the experimentally-formed MAPbI 3 , structure, the outlines of the ELF for I and Pb almost overlap by means of slight electron localization (sky blue colour) that provides more stability to the Pb-I interaction. The introduction of Cd produces directionality in the outlines of the electron localization of the I towards the Cd, which explains the decrease produced in the Cd-I distances. As an example, this directionality is shown as an arrow in the enlarged image of the ELF for the MAPb 0.75 Cd 0.25 I 3 structure. However, for the MAPb 0.75 Cd 0.25 I 3, MAPb 0.50 Cd 0.50 I 3 and MAPb 0.25 Cd 0.75 I 3 structures, there is a sky blue area between the Pb and I atoms due to the electron localization between them. For the MACdI 3 structure, a double directionality can be observed of the I towards each Cd, which explains the variation in the angles stated above (Table 4). This effect is also observed in the ELF image of the MAPb 0.25 Cd 0.75 I 3 structure in Figure S3 in the ESI. The results of the analysis of the ELF show that a certain proportion of Pb must be present for the perovskite structure to be formed and that the Pb-I interaction provides the structure with a stabilizing effect. At this point, it is interesting to use the NCI index, 18 an analysis tools that complements ELF and is Page 19 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 20 useful for studying weak interactions. This index enables the qualitative identification and characterisation of weak interactions of various strengths as chemically intuitive isosurfaces that reveal both stabilizing (strong and attractive interactions in blue, weak interactions in green) and destabilizing interactions (strong and repulsive interactions in red). Figure 5 shows a comparison of the NCI plots for the I-Pb-I, I-Cd-I-Pb-I and I-Cd- I interactions for the MAPbI 3 , MAPb 0.50 Cd 0.50 I 3 and MACdI 3 structures, respectively. The ESI includes the NCI plots for the remaining structures (Figure S4) as well as multimedia files (video) with the NCI analysis for all the structures. Comparing Figures 5A and 5B, it is interesting that the experimentally formed structures (MAPbI 3 and MAPb 0.50 Cd 0.50 I 3 ) show isosurfaces in blue for the Pb-I and Cd-I-Pb interactions. However, for the MACdI 3 structure, which was not formed experimentally, the Cd-I isosurface lobes are a green-blue colour, which shows that this interaction is weaker compared with the previous ones. Bearing in mind the tones of the isosurfaces, it is clear that the presence of Pb, particularly above a certain concentration, provides stabilizing characteristics to the Pb-I interaction, which promotes the formation of these perovskite-type structures. In this sense, the MAPb 0.25 Cd 0.75 I 3 structure that was not formed experimentally provides crucial evidence of this (see Figure S4 and multimedia files in the ESI). Figure S4 shows a majority of Cd-I interactions of a green-blue colour, compared with the uniform blue colour of the Pb-I interaction. This greater strength of the Pb-I interactions means that the structures with a certain amount of Pb are more stable, so they can be formed experimentally. In the case of the MAPb 0.25 Cd 0.75 I 3 structure, the smaller presence of strong interactions leads us to believe that the structure is less stable, which is coherent with the fact that it is not formed experimentally. Of course, all of the above does not mean that the interactions of other perovskite-type experimental structures will present green-blue interactions, as the Page 20 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 21 formation of these structures is known to present a wide range of exceptions. 49, 50 These results are coherent with the experimental results shown previously. Figure 5. NCI plots for tetragonal structures: (A) MAPbI 3 ; (B) MAPb 0.5 Cd 0.5 I 3 ; (C) MACdI 3 . 3.6. DOS and PDOS analysis To deepen our understanding of the electronic properties that underlie the geometrical effects mentioned above, we will discuss the electronic structure in terms of the density of states (DOS), the projected density of states (PDOS) and the band structure analysis. The band gaps calculated using the DOS analysis decreased with higher concentrations of Cd (see Figure 6). This smaller band gap is in agreement with the tendency observed experimentally for the perovskite structures formed. Figure 6 compares the DOS with and without SOC. Although, qualitatively, they follow the same trend, the band gap values calculated when the SOC is taken into account are slightly lower than those without SOC. Page 21 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 22 Figure 6. Total DOS for tetragonal perovskite structures analysed with and without SOC. Figure 7 shows the contribution of Pb, Cd, I and MA groups to the valence band (VB) and conduction band (CB) for the MAPbI 3 , MAPb 0.50 Cd 0.50 I 3 and MACdI 3 structures, which were chosen as being representative for the discussion below. Figure S5 in the ESI includes the same contributions for the MAPb 0.75 Cd 0.25 I 3 and MAPb 0.25 Cd 0.75 I 3 structures. Figure S6 shows the PDOS for all structures without SOC. In the figures, the contributions of I atoms and MA groups are represented as total I (s, p and d states of I atoms) and total CHN (s and p states of C and N, and s states of H), respectively. The analysis of the projected density of states (PDOS) of the structures simulated indicates that the VB is mostly composed of I p states, and the CB of the s, p and d states of the Pb 2+ and Cd 2+ ions, depending on the structure analysed. For the MAPbI 3 structure, the Page 22 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 23 Figure 7. Total CHN, total I, Pb and Cd projected states with SOC+U for tetragonal structures: (A) MAPbI 3 ; (B) MAPb 0.5 Cd 0.5 I 3 ; (C) MACdI 3 . VB and CB are mostly composed of I p and Pb p states (see Figure 7A) respectively, as reported elsewhere. 12, 30 Likewise, Figure 7A shows that the s and p states of the Pb make the biggest contribution to the edges of the VB and CB, respectively, while the MA groups contribute more internally. In Figure 7B, for the MAPb 0.50 Cd 0.50 I 3 structure, as well as the I and CHN states, it is possible to see that the s states of the Pb together with the d states of the Cd make the biggest contribution to the edge of the VB. In particular, the d states that are near the edge of the VB must play an important role in the charge transfer processes. The origin of the CB is dominated by the s states of the Cd followed by the p states of the Pb (Figure 7B). A different situation is observed for the MACdI 3 structure, where the I states contribute almost exclusively to the edge of the VB while the s states of the Cd make an extraordinary contribution to the origin of the conduction band (Figure 7C). The analysis of the PDOS shows that the contribution of the s states of the Cd to the origin of the conduction band (CB) plays a fundamental role in the decrease in the band gap (Figure 6). This decrease in the band gap values can be seen more clearly in Figure 8, where the Cd s states in the CB have been superimposed Page 23 of 30 Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript 24 for the MAPb 0.75 Cd 0.25 I 3 , MAPb 0.50 Cd 0.50 I 3 , MAPb 0.25 Cd 0.75 I 3 and MACdI 3 structures. Figure 8 shows the shift to lower gap values of the Cd s states as the concentration of Cd increases. Figure 8. Cd projected s-states for tetragonal perovskite structures analysed with SOC+U. In turn, Figure 9 shows the band structure for the MAPbI 3 , MAPb 0.50 Cd 0.50 I 3 and MACdI 3 structures. The Brillouin Zone for the tetragonal lattice was chosen as described in the literature. 28 As Figure 9A and 9C show, the transition from VB towards CB for the MAPbI 3 and MACdI 3 structures is direct and takes place at point Γ . However, the transition for MAPb 0.50 Cd 0.50 I 3 is indirect (Figure 9B). The valence-band- maximum (VBM) corresponds to point Γ and the conduction-band-minimum (CBM) to point R. For the MAPbI 3 structure, a splitting of the bands away from the critical point 36 associated with the loss of inversion symmetry has been reported. This splitting in the presence of SOC is also produced in the MAPb 0.50 Cd 0.50 I 3 and MACdI 3 structures. Nevertheless, strangely, the splitting for both the MAPb 0.50 Cd 0.50 I 3 and MACdI 3 around Page 24 of 30Physical Chemistry Chemical Physics Physical Chemistry Chemical Physics Accepted Manuscript