Novel Organic-Inorganic Perovskite Compounds Having Phosphonium Groups.

Organic-inorganic perovskites are composed of organic cations and [PbX6]4- octahedra, and the properties change depending on the type of organic cations. To identify the effect of organic cations and control the properties of the perovskite, thin films were prepared using quaternary alkylammonium and quaternary alkylphosphonium cations, which have big steric effects. A big steric effect can generate the distortion of [PbX6]4- octahedra leading to changes in properties. A thin film of a Pb-based organic-inorganic perovskite having quaternary alkylphosphonium cations was prepared for the first time. An exciton absorption was observed at a lower wavelength than other perovskites prepared from primary and quaternary ammonium salts. The perovskite with phosphonium groups was thermally stable compared with ammonium groups.

Introduction

Organic–inorganic metal halide-based perovskite-type compounds are artificially synthesizable hybrid materials. By dissolving stoichiometric amounts of an organic ammonium halide and lead halide in suitable solvents, these microcrystalline and crystalline perovskite compounds can be easily obtained because of their high self-organizational properties. Compounds composed of methylamine and lead iodide are rapidly gaining attention worldwide as next-generation solar cell materials because their initial application in photovoltaic devices by Miyasaka et al. in 2009.[1] As the study of perovskite solar cells (PSCs) has progressed dramatically over the last decade, the power conversion efficiency of PSCs has increased from 3.8% to more than 22%.[2−11] Although the perovskite-type materials used for photovoltaic solar cells have been limited thus far to three-dimensional (3D) perovskites with the general formula AMX3 (A = Cs+, CH3NH3+, and CH(NH2)2+; M = Pb2+ and Sn2+; X = Cl–, Br–, and I–), organic–inorganic perovskites are a relatively large family, thanks to various combinations of organic and inorganic components. For example, two-dimensional (2D) perovskites with the general formula A2MX4 (A = CH3(CH2)NH3+ and C6H5(CH2)NH3+; M = Pb2+ and Sn2+; X = Cl–, Br–, and I–) can be prepared from reacting alkylammonium salts with longer chain lengths than ethyl. 2D perovskites form quantum-well structures, in which inorganic layers and organic ammonium layers are alternately stacked. The inorganic layers comprise 2D sheets of [MX6]4– octahedra connected at four corners with halide ions in the plane. Accordingly, excitons are confined strongly in the inorganic layers. These compounds exhibit outstanding optical properties such as strong photoluminescence (PL) and high optical nonlinearity.[12−18] Furthermore, intermediate dimensional perovskites (2D–3D) and lower dimensional perovskites (1D, 0D) can be obtained by using various kinds of organic amines.[19−21] Nevertheless, most perovskites are fabricated from primary ammonium cations. Research of perovskite compounds with quaternary cations especially for phosphonium groups has not been progressed much.[22,23] Most of the articles on quaternary ammonium or phosphonium hybrids preferentially evaluated electrical properties of these hybrids having different metals from Pb, and thus the effects of bulky ammonium or phosphonium cations on other properties are still unclear. The three methyl groups around an ammonium ion generate a steric effect and affect the organic–inorganic interface, such that the changes in optical properties can be expected. Additionally, the ionic radius of phosphorus is larger than that of nitrogen, thus creating a larger steric effect. In this study, we introduced primary ammonium, quaternary ammonium, and phosphonium groups into the organic layer of perovskites to examine the steric effects of these organic cations on the physical properties of perovskites (Figure ).

Figure 1

Schematic view of the perovskite containing the organic compounds used in this study.

Results and Discussion

Quaternary ammonium bromide, C6H13N(CH3)3Br (CTNBr), or quaternary phosphonium bromide, C6H13P(CH3)3Br (CTPBr), were mixed with a stoichiometric amount of PbBr2 in N,N-dimethylformamide to make perovskite precursor solutions. Films of CTN–PbBr and CTP–PbBr were fabricated by spin-coating. As a comparison, a (C6H13NH3)2PbBr4 (CN–PbBr) spin-coated film was also fabricated by a similar procedure using primary ammonium hydrobromide, C6H13NH3Br (CNBr). Figure a,b show the out-of-plane X-ray diffraction (XRD) profiles of the CN–PbBr, CTN–PbBr, and CTP–PbBr films. CTN–PbBr and CTP–PbBr films showed different peaks from their precursors CTNBr, CTPBr, and PbBr2, suggesting that new structures were formed by complexation. A series of diffraction peaks corresponding to the interlayer d-spacing between the inorganic layers was observed for the CTN–PbBr and CN–PbBr films. From the diffraction peaks observed at 4.8° and 4.0° for CN–PbBr and CTN–PbBr, the d-spacing values were calculated to be 18.3 and 22.2 Å, respectively. Thus, by introducing the quaternary ammonium cation, the d-spacing increased by about 4 Å. The additional three methyl groups at the ammonium center produced a greater steric effect than a primary ammonium group. On the other hand, many peaks were observed for the CTP–PbBr film. CTP–PbBr generated a peak at 5.5° and the d-spacing value was 16.1 Å. Because the ionic radius of phosphorus is larger than that of nitrogen, the d-spacing value was expected to increase, but it was actually the reverse. This result suggested that the alkyl moieties of CTP–PbBr are greatly tilted or strongly packed between the inorganic layers.

Figure 2

Out-of-plane XRD profiles of films: (a) CN–PbBr and CTN–PbBr, (b) CTP–PbBr, and in-plane XRD profiles of films: (c) CN–PbBr and CTN–PbBr, and (d) CTP–PbBr.

Out-of-plane XRD profiles of films: (a) CN–n class="Chemical">PbBr and CTN–PbBr, (b) CTP–PbBr, and in-plane XRD profiles of films: (c) CN–PbBr and CTN–PbBr, and (d) CTP–PbBr.

Figure c,d show the in-plane XRD profiles of the CN–PbBr, CTN–PbBr, and CTP–PbBr films. CN–PbBr showed a diffraction peak at 15.3° (5.8 Å), which corresponds to the size of a [PbBr6]4– octahedra. Thus, the octahedra in this compound shared corners with one another to form an inorganic sheet parallel to the substrate. For CTN–PbBr with quaternary cations, the 15.3° diffraction peak disappeared, and a new peak was observed at 11.6° (7.6 Å). For CTP–PbBr, a peak at 9.6° (9.2 Å) was observed at a smaller angle. The d-spacing values are larger than that of ordinary corner-sharing perovskites, suggesting that the introduction of quaternary cations causes a change in the connectivity of [PbBr6]4– octahedra. It was predicted that the cation portion, with its large steric effect tilted, the octahedra such that the cations could fit between them when forming the perovskite structure. In addition, the perovskites with quaternary cations produced diffraction peaks at 3.9° and 5.9° for CTN–PbBr and at a smaller angle (3.8°) for CTP–PbBr, indicating that crystalline growth was partially perpendicular to the substrates. CTN–PbBr was expected to form a layered structure because a series of diffraction peaks was observed. However, a pure layered structure did not appear to be formed, which would have a molar ratio of CTNBr and PbBr2 of 2:1 with a theoretical elemental analysis of C, 26.51; H, 5.44; N, 3.44. Instead, the actual value was significantly different at C, 21.30; H, 4.28; N, 2.75. This result suggests that the molar ratio of CTNBr and PbBr2 is 1.33:1 (calcd C, 21.64; H, 4.44; N, 2.80). An ordinary layered perovskite has a structure in which organic cations are contained between [PbBr6]4– octahedra constituting the inorganic layer. However, because quaternary ammonium has a large steric effect, not all quaternary ammonium cations can fit between the [PbBr6]4– octahedra. Thus, it is considered that the steric effects induce the formation of a structure in which the organic cations are partially extruded or other connection of [PbBr6]4– octahedra. When ammonium cations are close to each other in the perovskite structure, van der Waals forces act between adjacent cations to maintain the layered structure. On the other hand, CTP–PbBr has a structure different from CTN–PbBr. There are many cations that cannot fit between the [PbBr6]4– octahedra due to the large steric effect.

Figure a shows the UV–vis absorption spectra of CN–PbBr, CTN–PbBr, and CTP–PbBr films. Absorption peaks attributed to quantum confinement structures were observed for each film, suggesting the formation of excitons in the perovskites. There are only a few reports of a phosphonium-based perovskite compound. It is notable that exciton absorption was also observed for CTP–PbBr to which alkylphosphonium groups were introduced.[24] As the steric effect of the cation portion increased, the exciton absorption peak shifted to shorter wavelengths, indicating that the quantum confinement structure was changed by introducing various cations with different steric effects. These peak shifts are attributed to the distortion of the [PbBr6]4– octahedra due to the increase in the steric effect of the cation portion. As the distortion of the [PbBr6]4– octahedra becomes energetically unstable, an exciton absorption peak shift to the higher energy side, namely the lower wavelength side, is observed. Figure b shows the PL spectra of CN–PbBr, CTN–PbBr, and CTP–PbBr films. Emission peaks corresponding to exciton absorption were observed for CN–PbBr and CTN–PbBr.

Figure 3

(a) UV–vis absorption spectra and (b) PL spectra of CN–PbBr, CTN–PbBr, and CTP–PbBr films.

(a) UV–vis absorption spectra and (b) PL spectra of CN–n class="Chemical">PbBr, CTN–PbBr, and CTP–PbBr films.

A Stokes shift was about 16 nm for CN–PbBr and 21 nm for CTN–PbBr. The peak intensity of CTN–PbBr was lower than that of CN–PbBr. On the other hand, CTP–PbBr generated a broad emission peak unlike the other two hybrids. This dramatic change is due to structural distortions of the inorganic lattice.[25]

Figure shows the thermogravimetric (TG) curves of CTN–PbBr and CTP–PbBr microcrystals. The perovskite with alkylphosphonium groups possessed higher thermal stability. The thermal decomposition point of CTP–PbBr significantly increased compared to CTN–PbBr, 399 °C for CTP–PbBr, and 279 °C for CTN–PbBr. In general, hydrogen bonds act between ammonium cations (N+) and halogen anions (Cl–, Br–, and I–) in the perovskite structure. In the case of the perovskite with phosphonium cations, hydrogen bonds also act between phosphonium cations and bromine anions. Thus, the higher thermal stability of CTP–PbBr is attributed to the stronger hydrogen bonds between phosphonium cations and bromine anions compared to those between ammonium cations and bromine anions.

Figure 4

TG curves of CTN–PbBr and CTP–PbBr.

TG curves of CTN–n class="Chemical">PbBr and CTP–PbBr.

When CTPBr was introduced to the perovskite, multiple structures were formed instead of a single-layered structure. In order to examine the additive effect of CTPBr to conventional 2D perovskites, we prepared perovskite films, (CN)(CTP)–PbBr (n = 0, 0.25, 0.50, 0.75, and 1.0), by mixing CNBr with CTPBr. Figure a,b show out-of-plane XRD profiles and UV–vis absorption spectra of the (CN)(CTP)–PbBr (n = 0, 0.25, 0.50, 0.75, and 1.0) films. For n = 1.0, a series of diffraction peaks was observed, indicating that the compound formed a layered structure. For n = 0.50 and 0.75, in addition to an original diffraction of 2D at 4.8°, diffraction peaks at 3.9° were observed. The peak intensity at 3.9° increased as n increased. For n = 0.25, the peak at 4.8° was completely disappeared and only a simple diffraction pattern at 3.9° was observed. Additionally, this film generated higher order diffraction peaks and had high crystallinity. This peak is consistent with the peak observed for in-plane diffraction patterns of pure CTP–PbBr which was shown in Figure d. This indicates that by mixing CNBr with CTPBr, the orientation of the layered structure was dramatically changed. Optical absorption properties also support this results. As shown in Figure b, only a single absorption peak was observed for each sample. The wavelength of the exciton absorption peak was dramatically changed between n = 1.0–0. The exciton absorption peak was observed around 400 nm for n = 0.50, 0.75, and 1.0, suggesting that these electronic states were affected by CN–PbBr. The small shift about 10 nm observed for n = 0.50 and 0.75 compared with n = 1.0. This implies that the energy loss was a structural perturbation. On the other hand, the exciton absorption peak was observed around 320 nm for n = 0.25 and 0, suggesting that these electronic states were dominated by CTP–PbBr. This is well consistent with the results of XRD patterns shown in Figure a. From the out-of-plane XRD results shown above, the diffraction intensity at 4.8° is stronger than that of 3.9° for n = 0.50 and 0.75, therefore, the optical properties were mainly dominated by the conventional 2D compound, CN–PbBr. For n = 0.25, the XRD profile shows a strong diffraction at 3.9°, so the electric state of CTP–PbBr was dominated. These results showed that the structural orientation and electronic state of perovskite compounds can be controlled by mixing primary and quaternary cations.

Figure 5

(a) Out-of-plane XRD profiles and (b) UV–vis absorption spectra of (CN)(CTP)–PbBr (n = 0, 0.25, 0.50, 0.75, and 1.0) films.

(a) Out-of-plane XRD profiles and (b) UV–vis absorption spectra of (CN)(CTP)–n class="Chemical">PbBr (n = 0, 0.25, 0.50, 0.75, and 1.0) films.

Conclusions

Perovskite films were prepared from quaternary long alkylphosphonium cations for the first time. Perovskite compounds can be prepared not only by the ammonium type, but also by the phosphonium type cations. Introducing quaternary cations into the perovskite structure shifted the absorption peaks to shorter wavelengths upon increasing the steric effect of the cation portion. The phosphonium-based perovskites possessed higher thermal stability due to the strong hydrogen bonds between phosphonium cations and bromine anions. Furthermore, the structural orientation and electronic state of perovskite compounds can be controlled by mixing primary and quaternary cations.

Experimental Section

Materials

Quaternary ammonium bromide, C6H13N(CH3)3Br, was purchased from Tokyo Chemical Industry Co., Ltd. and used as received. Hexylamine, C6H13NH2, was also purchased from Tokyo Chemical Industry Co., Ltd. Acetone was purchased from Kanto Chemical Co., Inc. 1-Bromohexane, hydrobromic acid (HBr, 48 wt % aqueous solution), and DMF (super dehydrated) were FUJIFILM Wako Pure Chemical Industries, Ltd. Trimethylphosphine was purchased from Sigma-Aldrich. Quaternary phosphonium bromide, C6H13P(CH3)3Br, was synthesized by reacting with equivalent 1-bromohexane and trimethylphosphine.[26] Primary ammonium hydrobromide, C6H13NH3Br, was synthesized by neutralizing C6H13NH2 and with stoichiometric amounts of HBr.

Elemental analysis data for C6H13P(CH3)3Br, n class="CellLine">C9H22PBr: Calcd C, 44.83; H, 9.20. Exp. C, 44.63; H, 9.22.

Elemental analysis data for C6H13N(CH3)3Br, n class="CellLine">C9H22NBr: Calcd C, 48.22; H, 9.89; N, 6.25. Exp. C, 48.42; H, 9.38; N, 6.26.

Elemental analysis data for C6H13NH3Br, C6H16n class="Chemical">NBr: Calcd C, 39.57; H, 8.86; N, 7.69. Exp. C, 39.60; H, 8.70; N, 7.64.

Sample Preparation

C6H13P(CH3)3Br and PbBr2 were dissolved in DMF at 50 °C for 1 h to obtain the precursor solutions for the preparation of C6TP–PbBr. Similarly, precursor solutions of C6TN–PbBr and C6N–PbBr were prepared. Films were fabricated on preheated hydrophilic substrates by spin-coating at 2000 rpm using a MIKASA 1H-D7 SPINCOATER. The substrates were heated at ca. 100 °C during the spin-coating process to obtain high-quality films because of the high boiling point of DMF. As a comparison, the (C6H13NH3)2PbBr4 (CN–PbBr) spin-coated film was also fabricated by a similar procedure using primary ammonium hydrobromide, C6H13NH3Br (CNBr).

(CN)(CTP)–n class="Chemical">PbBr (n = 0, 0.25, 0.50, 0.75, and 1.0) films were also fabricated by a similar procedure. The precursor solution was prepared by mixing CNBr and CTPBr in any ratio. The following microcrystalline powder are for the thermal analysis. A microcrystalline powder of C6TN–PbBr was obtained by pouring the precursor solution into acetone. After vacuum drying, a white powder was obtained. The microcrystalline powder of C6TP–PbBr was obtained by drying the precursor solution, and a white powder was obtained.

Elemental analysis data for C6TN–n class="Chemical">PbBr: Calcd (assuming that molar ratio C6TNBr and PbBr2 is 2:1). C, 26.51; H, 5.44; N, 3.44%. Calcd (assuming that molar ratio C6TNBr and PbBr2 is 1.33:1). C, 21.64; H, 4.44; N, 2.80%. Exp. C, 21.35; H, 4.28; N, 2.75%.

Characterization

The optical absorption spectra of the spin-coated films were obtained with a Shimadzu UV-3100PC UV–vis–NIR spectrophotometer at room temperature. The XRD profiles were obtained over the 2θ range of 1.5–35° with a Rigaku SmartLab X-ray diffractometer operating at 45 kV and 200 mA using a Ni-filtered Cu Kα target. Thermal analysis was performed using a SII NanoTechnology TG-DTA7200 thermogravimeter-differential thermal analyzer under a nitrogen flow of 200 mL min–1.