Structure and luminescence colour change of an iridium(III) bis-terpyridine complex crystal triggered by water sorption/desorption

Abstract

A cationic complex crystal of [Ir(tpy)₂]Br₃·7H₂O (tpy = 2,2′:6′,2″-terpyridine) shows luminescence vapochromism as a reversible conversion between yellow and orange emissions in response to heating/cooling. X-ray crystallography reveals that [Ir(tpy)₂]Br₃·7H₂O exists as a triclinic crystal system, comprising Ir(III) complex layers and 2D hydrophilic sheets of H₂O and Br⁻. The orange emissive crystal is determined to be the orthorhombic [Ir(tpy)₂]Br₃·2H₂O, comprising 1D hydrophilic channels of H₂O and Br⁻, with Br⁻ in close proximity to the Ir(III) complex. The yellow emission of [Ir(tpy)₂]Br₃·7H₂O is due to the dual emission from the ligand-centered π–π* transition and the Br⁻-to-ligand charge transfer (XLCT). The π–π* emission is quenched by the switching to the XLCT emission by the close approach of Br⁻ to Ir(III) complex resulting from the H₂O desorption. The reversible H₂O sorption/desorption results in not only the luminescence colour change but also the drastic structural change between triclinic and orthorhombic crystal systems. The weak stimulus of H₂O sorption/desorption triggers the phase transition of the Ir(III) complex crystal; 5 H₂O molecules are incorporated into the 1D channel of the [Ir(tpy)₂]Br₃·2H₂O crystal to change the Ir–Br⁻ framework from orthorhombic to triclinic, and inversely, the elimination of 5 H₂O molecules from the 2D hydrophilic sheet in [Ir(tpy)₂]Br₃·7H₂O contributes to the reconstruction of the hydrophilic channel along with the reversion of the framework to orthorhombic.

---

Introduction

Most crystals consisting of transition metal complexes do not lose their crystallinity and intrinsic colour even when cracked and ground into fine powders because their periodic alignments of components remain the same. However, some Pt(II) complexes frequently show the loss of crystallinity resulting from amorphization upon mechanical grinding, thereby changing their colour and luminescence. For example, Pt(5mdpb)Cl (5mdpbH = 1,3-di(5-methyl-2-pyridyl)benzene)¹ and Pt(5fdpb)Cl (5fdpbH = 1,3-di(5-trifluoromethyl-2-pyridyl)benzene)² crystals show the colour change from yellow to orange with the switching of the excited state from the π–π* state to the MMLCT (metal metal-to-ligand charge transfer) state of the Pt–Pt dimeric form. Thereafter, reports on the MMLCT emission of Pt(II) complex crystals based on the mechanochromism increased.^3–11 Occasionally, such a weak mechanical stimulus causes not only the dimerization of the ground Pt(II) complex but also the excimer formation of the Pt(II) complex;^12–17 a Pt(5mdpb)Cl (5mdpbH = 1,3-di(5-methyl-2-pyridyl)benzene) crystal¹ shows no change in the crystal structure and colour even upon mechanical grinding, but it emits an orange luminescence different from the original yellow emission before grinding. The orange luminescence is attributed to the (Pt–Pt)* excimer formation on the surface of the powdered crystals.

Exposure to solvent vapor, which is another weak external stimulus, also induces a colour change in the Pt(II) complex crystal.^18–20 For instance, a purple crystal of Pt(Fmdpb)CN (FmdpbH = 4,6-difuluoro-1,3-di(5-methyl-2-pyridyl)benzene), showing no visible luminescence, suddenly emits a red luminescence along with a red colouration upon exposure to chloroform vapor.²¹ The characteristic colour change is attributed to the emergence of visible absorption due to the MMLCT transition in the dimeric [Pt(Fmdpb)CN]₂ moiety upon sorption of CHCl₃ vapor into the crystal. These neutral and planar Pt(N^C^N) complex crystals, exhibiting a colour change in response to mechanical grinding or exposure to CHCl₃ vapor, are classified as “soft crystals”,^22,23 one of the categories of smart materials, which can easily change their structure by weak stimulus while retaining their highly ordered structures. So far, many materials consisting of neutral compounds, including organic molecules, organometallic complexes, and transition metal complexes, have been found to show the characteristics of “soft crystals”, which is the change in morphology, colour, or luminescence as a response to weak external stimulus such as mechanical stress, exposure to vapor, or heating.²⁴ Recently, we presented vapochromic characteristics of a cationic complex salt, [Ir(tpy)₂]I₃·2H₂O (tpy = 2,2′:6′,2″-terpyridine), for which colour switches between orange and red during the sorption/desorption of H₂O from/to the atmosphere.²⁴ Reports on luminescence vapochromism of Ir(III) complexes seem to be limited compared to Pt(II) complexes.^25–31

The strong visible colouration is due to the I⁻-to-tpy ligand charge transfer (XLCT) transition of [Ir(tpy)₂]I₃·2H₂O,³² where the HOMO, a 5p orbital of I⁻, is strongly influenced by the number of hydrated H₂O molecules, resulting in the shift of the XLCT band. The X-ray crystallography of [Ir(tpy)₂]I₃·2H₂O revealed that H₂O molecules are stored in hydrophilic channels and connected with I⁻ by hydrogen bonds. A Rietveld analysis of the PXRD patterns for [Ir(tpy)₂]I₃·nH₂O (n = 0–2) clarified that the [Ir(tpy)₂]I₃ moieties construct a rigid framework and retain the original packing structure during the H₂O sorption/desorption cycles.²⁴ Unfortunately, [Ir(tpy)₂]I₃·nH₂O does not emit luminescence at room temperature, and the vapochromic response is observed through a broad UV-vis absorption spectrum, and therefore the spectral changes were too small to analyze the interaction of the Ir(III) complex with the counter anion involved in vapochromism. In the present work, we prepared an emissive [Ir(tpy)₂]Br₃·7H₂O to observe the luminescence vapochromic response at room temperature. The complex crystal shows a reversible luminescence colour change during the sorption/desorption of H₂O molecules, along with a change in crystal structure, which is characteristic of “soft crystals”.

Experimental

Preparation

[Ir(tpy)₂]Br₃·7H₂O was prepared from [Ir(tpy)₂]I₃·2H₂O²⁴ by an ion-exchange method. [Ir(tpy)₂]I₃·2H₂O (74 mg, 0.071 mmol) and AgBr (53 mg, 0.28 mmol) were dissolved in water and subsequently stirred for 1 day under dark conditions. Removing the white precipitate by filtration afforded a clear yellow solution. After solvent evaporation, the resultant yellow solid material was washed with acetone and acetonitrile; yield 41 mg (67%). A single crystal [Ir(tpy)₂]Br₃·7H₂O suitable for X-ray crystal structure analysis was prepared by the slow evaporation of the solvent from an acetonitrile solution containing [Ir(tpy)₂]Br₃. Another single crystal [Ir(tpy)₂]Br₃·2H₂O suitable for X-ray crystal structure analysis was prepared by the vapor-diffusion method. To a large vial containing an acetone/dimethyl sulfoxide (2 : 1) mixture, as a poor solvent, a small vail containing an acetonitrile solution of [Ir(tpy)₂]Br₃ was placed. The large vial was capped and left to stand for 2 weeks, which yielded orange-emissive single crystals.

Measurements

UV-vis absorption spectroscopy was conducted using a JASCO spectrometer V-530ST. A 1 cm quartz cuvette attached with a grease-less valve was used for UV-vis measurements of a solid sample deposited as a thin-film on the inner wall of the cuvette. The emission spectra were measured by a JASCO FP-6500 spectrofluorometer. The samples for emission measurements were prepared in a quartz cuvette attached with a greaseless valve. The emission lifetimes were evaluated by the least squares method on a PC for time periods of luminescence recorded by a Tektronix TDS1102C digital oscilloscope after pulse excitation using a Japan Laser GL-3300 N₂ laser (337 nm). Thermogravimetric (TG) measurements were performed using a Hitachi Hightech Science STA7300. The solid sample stored in an aluminum pan was heated from 293 to 500 K at a heating rate of 5 K min⁻¹ under a nitrogen gas atmosphere. Differential scanning calorimetry (DSC) measurements were carried out in the range from room temperature to 400 K using a Hitachi High-Tech Science DSC 7020. Powder X-ray diffraction (PXRD) measurements were performed using a Bruker D8 ADVANCE equipped with a TTK450 for variable temperature measurement. The PXRD patterns for [Ir(tpy)₂]Br₃·7H₂O, [Ir(tpy)₂]Br₃·2H₂O, [Ir(tpy)₂]Br₃·H₂O, and [Ir(tpy)₂]Br₃ were measured at 25 °C, 80 °C, 110 °C, and 190 °C, respectively. A PXRD measurement to track the change from [Ir(tpy)₂]Br₃·2H₂O to [Ir(tpy)₂]Br₃·7H₂O was conducted at room temperature, where [Ir(tpy)₂]Br₃·2H₂O was prepared from [Ir(tpy)₂]Br₃·7H₂O by the desorption of 5H₂O molecules at 80 °C.

X-ray crystallography

The X-ray diffraction data were obtained at 123 K using a Rigaku XtaLAB Synergy-DW system with MoKα radiation. Indexing, integration, scaling, and absorption correction of the diffraction data were carried out using CrysAlis^Pro.³³ The structure was solved by the SHELXT 2014 program³⁴ and refined by full-matrix least-squares techniques against F² (SHELXL-2019/3)³⁵ with the Shelxle software package.³⁶ The non-hydrogen atoms were refined anisotropically. The crystal and molecular structures were drawn using the Mercury software.³⁷

Results and discussion

H₂O desorption/sorption

A thermogravimetry (TG) measurement was conducted for a [Ir(tpy)₂]Br₃·7H₂O (M7, tpy = 2,2′:6′,2″-terpyridine) salt prepared by the addition of acetone into an acetonitrile solution containing the Ir(III) complex. As shown in Fig. 1, the TG (black) and DTG (red) results clearly show a three-step weight loss; the peaks of the DTG curve are observed at 60 °C, 90 °C, and 140 °C. For the 1st stage, the weight change reaches −8.80% up to 65 °C, which matches the loss of 5 H₂O. A gradual decrease in weight at the initial stage below 50 °C suggests that one H₂O molecule among the water of hydration of M7 can be easily removed from the crystal framework, which likely occurs even at room temperature. The weight changes in the 2nd and 3rd stages correspond to −1.75% and −1.76%, respectively, each of which is due to the loss of a single H₂O molecule. Since there is no remarkable weight-loss in the region of 65–80 °C, it is reasonably assigned that the crystal heated at 80 °C is [Ir(tpy)₂]Br₃·2H₂O (M2). [Ir(tpy)₂]Br₃·H₂O (M1) and [Ir(tpy)₂]Br₃ (M0) are suggested to be obtained by heating at 90–100 °C and 140 °C, respectively. The intermediate phases of M5–M3 seem not to have been detected through the TG measurement. When M0 was cooled at room temperature under a humid atmosphere, the weight loss was recovered by the sorption of H₂O. Fig. S1 shows the TG results for two samples subjected to the heating–cooling treatment, where the number of water molecules of hydration is determined to be 6.0 and 6.4, respectively. This result suggests that the H₂O sorption from M0 to M7 is incomplete at room temperature.

Fig. 1 TG (black) and DTG (red) plots of [Ir(tpy)₂]Br₃·7H₂O (M7). Three distinct peaks are observed at 60 °C, 90 °C, and 140 °C in the DTG curve.

Fig. S2 shows the DSC result of M7 during heating under an N₂ atmosphere. Two distinct endothermic peaks are observed in the range of 30–70 °C and 105–125 °C, which correspond to endothermic processes for the desorption of 5H₂O and 2H₂O, respectively. Since the endothermic amount at 105–125 °C is +13 kJ mol⁻¹, we can estimate the endothermic amount of single H₂O desorption to be +6.5 kJ mol⁻¹. It is less than the distillation enthalpy of H₂O (ΔH = +44.016 kJ mol⁻¹ at 298 K),³⁸ suggesting that the elimination of H₂O from M2 and M1 easily occurs. In contrast, the endothermic amount of +107 kJ mol⁻¹ at 30–70 °C is much larger than the +32.6 kJ mol⁻¹ for 5H₂O desorption. It is suggested that the transition from M7 to M2 needs a large thermal energy for the change in the Ir–Br⁻ framework in the crystal in addition to the 5H₂O desorption.

Crystal structures of M7 and M2

We determined the crystal structure of M7 as a triclinic crystal system with the P1̄ space-group (Table S1). The X-ray diffraction study on this sample was performed at −150 °C and the number of H₂O molecules was determined to be 7 in the unit cell. When viewed from the a-axis of the unit cell, each Ir(III) complex aligns its principal axis of the D₂ structure parallel to the a-axis, as shown in Fig. 2. Molecules of the water of hydration are accommodated in voids between Ir(III) complexes on the bc-plane and form a hydrogen-bond network together with Br⁻. These hydrogen bonds are expanded along the a-axis to construct a 2D sheet structure. Two Br⁻ ions, coloured with green and blue, are buried in the 2D sheets, while the other ion (yellow) is close to Ir (l_Ir1–Br2 = 4.729 Å) and is located between two tpy ligands with a bite angle of 78.88° as to be pinched with a scissors of ligands when viewed from the (b–c)/2 axis of the unit cell (Fig. S3). The Ir–Br⁻ pair (A) and its turnover (B) are alternately ordered in a straight line, parallel to the hydrogen bonding networks. The distances between Br⁻ and O are in the range of 3.285–3.366 Å and those of O–O are at 2.751–2.814 Å (Table S2). An eight-membered ring consisting of 2Br⁻ (Br1) and 6O (O4, O5, O6) is observed in the unit cell. O5 and O6 connect to O5 and O6 in adjacent unit cells, respectively, thereby extending the hydrogen bonding network along the a-axis (Fig. S4). O1, O2, O3, and Br3 contribute to the extension of the network along the direction of the (b–c)/2 axis. Br2 connects to O1 and O7 through hydrogen bonding. It is observed that O7 forms a hydrogen bond with Br2 alone, suggesting the hydrogen bond between Br2 and O7 is weak.

Fig. 2 Crystal structure of M7. A and B denote the [Ir(tpy)₂]³⁺–Br⁻ pairs projected on the bc-plane when viewed along the a-axis. Red circles are O atoms from the water of hydration. The small green, red, and blue circles each represent Br⁻. The coloured lines between O and Br show the hydrogen bonds. Hydrogen atoms are omitted for clarity.

X-ray crystallography analysis determined that the crystal system of M2 is orthorhombic with the Pna2₁ space-group (Table S3). As well as M7, two Ir–Br⁻ pairs (C and D) are observed in the M2 crystal, as shown in Fig. 3. Each Ir(III) complex is in an almost perfect D_2d symmetry and aligns its principal C₂ axis parallel to the a-axis of the unit cell. Br⁻ (yellow) is located on the C₂ axis and seems not to form any hydrogen bond with H₂O. The closest distance between Br⁻ and Ir is determined as l_Ir1–Br38 = 4.640 Å for the Ir–Br⁻ pair. The pairs C and D in Fig. 3 are equivalent to each other, yet D is the turn-over of C with respect to the bc-plane and located with an offset of a/2 with respect to C. The other Br⁻ ions (green and blue) contribute to hydrogen-bonding with the water of hydration, respectively, to construct 1D channels along the a-axis. The distances between O and Br⁻ are l_Br40–O42 = 3.358 Å and l_Br39–O41 = 3.355 Å, respectively (see Table S4 and Fig. S5). The hydrogen bonds are extended along the a-axis of the unit cell. In addition, we found a very interesting and rare phenomenon, which is the electron deficiency around Br⁻ (yellow) with no hydrogen bonding (Fig. S6). Although this might be due to a partial electron transfer from Br⁻ to the Ir(III) complex, we could not observe any increase in electron density around [Ir(tpy)₂]³⁺. This could be because Ir, as a heavy atom with many electrons, causes any such increase to fall within the experimental error range and thus remain undetected.

Fig. 3 Crystal structure of M2. C and D denote the [Ir(tpy)₂]³⁺–Br⁻ pairs projected on the bc-plane when viewed along the a-axis. Red circles are the O atoms of the water of hydration. The small green, red, and blue circles each represent Br⁻. The coloured lines between O and Br show the hydrogen bonds. Hydrogen atoms are omitted for clarity.

The X-ray crystallography for M7 and M2 clarified that the difference in the number of water of hydration in the crystal causes the difference in the hydrogen-bond networks. H₂O can be eliminated through the 2D hydrophilic sheet in the desorption process from M7 to M2. The 5H₂O loss results in the rearrangement of the Ir–Br⁻ framework, along with the change in the hydrophilic channel from 2D sheets to 1D columns and with the release of [Ir(tpy)₂]³⁺ from the stress due to the distorted structure. The distortion energy for the [Ir(tpy)₂]³⁺ moiety was predicted through a DFT calculation³⁹ performed by using CAM-B3LYP as a functional and the basis set of 6-31G* for H, C, N, O, Br, and LANL2DZ for Ir. In this calculation, the geometries for the distorted and undistorted structures were fixed to those of [Ir(tpy)₂]³⁺ from the X-ray crystal structures for M2 and M7, respectively, and Br⁻ and H₂O were omitted. The relative energy of the distorted structure in M7 with respect to the D_2d structure in M2 was predicted to be +7.32 kJ mol⁻¹. Even though this exothermic energy is consumed for the conversion from M7 to M2, it is too small to compensate for the endothermic energy of +107 kJ mol⁻¹ for the phase transition from M7 to M2 determined by the DSC measurement.

Rearrangement of the Ir–Br⁻ framework

To detect the Ir–Br⁻ framework change experimentally, the H₂O sorption process for M2 was tracked through the PXRD measurement, where the diffraction pattern at 2θ = 8–12° for M2 prepared by heating at 80 °C was recorded every 8 min to observe the change in peaks during the sorption of H₂O molecules by cooling. As shown in Fig. 4, three distinct peaks at 2θ = 9.4°, 10.7°, 11.5° and a small peak at 2θ = 10.4° are observed initially (a, black), consistent with the powder pattern simulated from the result of the single crystal X-ray diffraction. The peaks at 9.4°, 10.4°, and 11.5° gradually become small and disappear in 64 min, as displayed in (b)–(j). Instead, new peaks emerge at 8.6° and 9.9° and grow larger over time. Eventually, three peaks at 2θ = 8.6°, 9.9°, and 10.7° remain in the powder pattern (j). The peak shape at 10.7° appears asymmetric due to the shoulder at the higher angle. These peaks are consistent with the PXRD pattern for M7 (k). Clearly, the change in diffraction pattern corresponds to the change in the Ir–Br⁻ framework in the crystal during the H₂O sorption process from M2 to M7.

Fig. 4 Variation in the PXRD pattern of M2 prepared by heating at 80 °C upon cooling to room temperature. The diffraction patterns are observed every 8 min. For M2 (a) and M7 (k), each experimental PXRD pattern (black) is reproduced with four Gauss functions, and the resultant simulations are displayed as green and purple, respectively. The red curves (b–i) superimposed by the simulations for M2 (green) and M7 (purple) are the best fit to the PXRD patterns (c–k, black), respectively.

Despite the fact that this H₂O-sorption process took around 60 min for completion, no apparent peaks for M3–M6 were detected. This suggests that (1) the H₂O sorption is too rapid to detect the peaks for M3–M6 by our apparatus, (2) the crystal systems of M3–M6 are the same as that of M7 or M2, or (3) M3–M6 are not detected by PXRD because of the lack of crystallinity. In the case of (1), considering that the M2 used is a powdery sample consisting of many fine crystals, the slow change in peak intensity might be due to the statistical change of the M2 content in the entire fine crystals, where the rapid H₂O sorption of each fine crystal occurs. In that case, any PXRD pattern should be the superposition of those of M2 and M7. In the case of (2), the crystal systems of M3–M6 are expected to be the same as that of M7. Since there is no void space in the M2 crystal to accommodate extra H₂O molecules, if H₂O molecules are forced to insert into the M2 crystal, the Ir–Br⁻ framework of M2 will change to that of M7 to accommodate extra H₂O molecules. Considering that the X-ray diffraction intensity is proportional to the number of electrons, the XRD peak intensities for M3–M7 may be varied by the number of H₂O molecules, even though the peaks are observed at the same 2θ angles (Fig. S7). The observed PXRD patterns during the H₂O sorption will not be reproduced from the simple superposition of those of M2 and M7. Regarding case (3), the lack of crystallinity might be provided by a variety of structural changes in the Ir–Br⁻ framework due to the squeezing of extra H₂O molecules into the M2 crystal. In this case, the transient domains of M3–M6 would contribute to the depletion of the PXRD intensity.

In Fig. 4, simulations of the PXRD patterns are exhibited as coloured curves. The patterns of M2 (a) and M7 (k) were reproduced using four Gauss functions and displayed as green and purple curves, respectively. For (b)–(j), the patterns simulated by the superposition of green and purple curves are displayed as red curves, respectively, which sufficiently match each experimental result. This concludes that the intermediate phases are not detected in the present work; that is, the H₂O sorption process along with the framework change is too rapid to detect though the PXRD measurement. Fig. S8 shows plots of the composition of M2 and M7 in the simulation, where the change in composition is shown to be neither linear nor exponential but stepwise. This result might suggest some difference in the rate of framework change depending on the inhomogeneity in the crystal size of M2.

Structure of M1 and M0

The PXRD measurements were carried out for [Ir(tpy)₂]Br₃·7H₂O at 190, 110, 75, and 25 °C, respectively. The results are exhibited as red lines in Fig. S9, where we assigned the samples at 190, 110, 75, and 25 °C to M0, M1, M2, and M7, respectively, according to the TG results. The blue line of M7 is derived from a simulation of the single crystal X-ray diffraction analysis, which is in good agreement with the experimental pattern at 25 °C. For M2, as well, the pattern observed at 75 °C is very similar to the simulation from the single crystal X-ray diffraction result, suggesting the crystal structure of the Ir(III) complex prepared by heating at 75 °C from M7 is the same as that of M2 directly prepared by crystallization from the acetonitrile solution. This indicates that the dehydration processes from M7 to M2 are a single-crystal to single-crystal transition. Considering V = 1750.47 ų for M7 (Z = 2) and V = 2963.01 ų for M2 (Z = 4) from the single crystal X-ray analyses, the net difference in volume is roughly estimated as 134 ų when converted into the case of Z = 1. It should be noted that this volume difference is comparable to 150 ų, the sum of volumes of 5H₂O. Both the patterns of M1 and M0 coincide with that of M2, which suggests that their packing structures consisting of [Ir(tpy)₂]³⁺ and 3Br⁻ are very similar to one another; that is, the crystal frameworks are robust and retain the same structure during H₂O desorption by heating. This result suggests that the dehydration processes from M2 to M1 to M0 are single-crystal to single-crystal transitions. The temperature effect on peak angles is recognized; for example, the peaks at the lowest angle are observed at 9.337° (M2), 9.408° (M1), and 9.428° (M0). The variation in peak angles would be due to the combination of the thermal expansion of the crystal lattice and the contraction resulting from the elimination of H₂O. Rietveld analyses⁴⁰ of the PXRD patterns were performed for M2, M1 and M0, where the geometry of the Ir(III) complex and the space group are the same as those in M2 determined by the single crystal X-ray structural analysis. Results are in good agreement with the experimental patterns (Fig. S9). As listed in Table S5, although the lattice parameters a, b, and c decrease slightly or remain the same with increasing temperature, the volumes clearly decrease. The decrease in volume from M2 to M0 is 14.8 ų, which is very small, considering the volume of a single H₂O molecule is 30 ų and Z = 4 in the Pna2₁ space group. The small change in volume might guarantee the robustness of the crystal framework during the sorption/desorption of H₂O.

Vapochromic behaviour

As shown in Fig. 5, M7 emits yellow luminescence showing vibronic structure peaks at 507, 544, and 581 nm, which can be assigned to the π–π* transition of the ligand.³² Upon heating at 80 °C for 5 min, surprisingly, the yellow luminescence changes to orange resulting in a broad spectrum centered at 623 nm. This broad emission is likely due to the Br⁻-to-ligand charge transfer (XLCT) transition of M2, according to the emission study on [Ir(tpy)₂]₂X₃ (X⁻ = Cl⁻, Br⁻, and I⁻) at 77 K.³² The luminescence colour returns to original yellow in 40 min when the crystalline sample is cooled under ambient atmosphere. Since the luminescence change is more sensitive than the colour change, the [Ir(tpy)₂]Br₃ crystal showing the visible luminescence vapochromism is superior to the non-emissive [Ir(tpy)₂]I₃ system previously reported.³² We note that the emission profile of M7 appears to be a superposition of the structured π–π* emission and the broad XLCT emission. This spectrum is similar to that of the dual emission of [Ir(tpy)₂](PF₆)₃ in acetonitrile,³² where structured and broad emissions were assigned to the individual emissions of [Ir(tpy)₂]³⁺ and [Ir(tpy)₃]³⁺·PF₆⁻, respectively, in the ion-association equilibrium. As well, in the present case, the broad emission component of M7 is probably assigned to the XLCT emission of the [Ir(tpy)₃]³⁺·Br⁻ pair, in which the Br⁻ bearing H₂O is located on the pseudo C₂ axis of the Ir(III) complex with l_Ir1–Br2 = 4.729 Å. The broad emission of M7 is blue shifted, along with sharpening at 77 K, as shown in Fig. S10, which is a familiar phenomenon for the CT emissions of metal complexes. On the other hand, the structured emission at around 500 nm attributable to the π–π* emission remains even at 77 K. These results support that the emission of M7 is assigned to the dual emission.

Fig. 5 Photographs (insets) and emission spectra of the [Ir(tpy)₂]Br₃·7H₂O (M7) and [Ir(tpy)₂]Br₃·2H₂O (M2) crystals observed at room temperature.

An emission lifetime measurement was conducted for M7 and M2 at room temperature. The time course of the M7 emission monitored at 500 nm, a peak wavelength of the vibronic band of the π–π* emission, consists of fast and slow decay curves (Fig. S11a); the lifetimes of the fast and slow components are estimated to be 20 ns and 300 ns, respectively, by a least square fitting using a double exponential function. When monitored at 650 nm, the peak wavelength of the broad emission, the emission lifetime was determined as τ = 540 ns, which is longer than τ = 300 ns. The longer the monitoring wavelength, the longer the emission lifetime. On the other hand, the emission lifetime of M2 monitored at 650 nm is determined as τ = 860 ns (Fig. S11b). H₂O desorption likely elongates the XLCT emission lifetime, probably because of the prevention from nonradiative deactivation due to the O–H vibration of H₂O. Consequently, the luminescence vapochromism of M7 is established to originate from switching of the emitting state from the π–π* state to the XLCT state, probably due to the proximity of Br⁻ and [Ir(tpy)₂]³⁺.

This luminescence-colour conversion is repeatedly observed, as shown in Fig. S12, which are plotted as the emission intensities of the sample at 540 nm upon heating (red) for 10 min and cooling (black) for 20 min under a humid atmosphere alternately. Although the emission intensities of the heated sample are stable during the heating–cooling cycles, those after cooling seem to vary. Fig. S13 shows a change in the emission spectra of the heated sample (M2) observed at room temperature under a humid atmosphere. The broad band at 610 nm gradually decreases in intensity, along with the growth of vibronic bands in the region of 500–550 nm. The iso-emissive point observed at 570 nm at the initial stage eventually disappears. By analogy to the TG results, it is suggested that M2 is not converted to M7 completely upon cooling but to the M7/M6 mixture.

Although we cannot perceive the remarkable colour difference in the luminescence of M2–M0, the emission spectrum of M2 slightly shifted to the longer wavelength region upon removing H₂O from the crystal by heating. The resultant spectra of M1 and M0 are similar to each other, as shown in Fig. S14. The UV absorption spectrum in the 250–400 nm region of M2 is also shifted to the longer wavelength region by heating to produce M1 and M0. This phenomenon is very similar to that found for the [Ir(tpy)₂]I₃·2H₂O–[Ir(tpy)₂]I₃ system previously, where the reversible colour change between yellow and orange due to the sorption/desorption of 2H₂O is observed.

Conclusion

We found the vapochromism due to the reversible H₂O sorption/desorption for the cationic complex crystal [Ir(tpy)₂]Br₃·nH₂O (n = 7–0). The X-ray crystallography revealed the hydrophilic 2D sheets and channels in [Ir(tpy)₂]Br₃·7H₂O (triclinic) and [Ir(tpy)₂]Br₃·2H₂O (orthorhombic), respectively, both consisting of H₂O and Br⁻ interacting via hydrogen bonding. The heating of the yellow [Ir(tpy)₂]Br₃·7H₂O crystal results in the removal of H₂O molecules hydrated to Br⁻, followed by the orange colouration due to the emergence of the XLCT absorption band in the visible region. This vapochromic process is accompanied by the drastic change in the Ir–Br⁻ framework in the crystal, which is a characteristic behaviour for “soft crystals” and a novel phenomenon for cationic transition metal complexes. For [Ir(tpy)₂]Br₃·2H₂O, the rigidity of the Ir–Br⁻ framework makes it possible for the H₂O desorption process producing empty channels, whereas flexibility is necessary to insert H₂O molecules into the occupied channels. It is notable that a weak stimulus, such as exposure to moisture, causes a drastic change in the crystal from triclinic to orthorhombic. Unfortunately, analysis of the mechanism of the dynamic structural change from triclinic to orthorhombic was not achieved through the detection of intermediate phases for [Ir(tpy)₂]Br₃·nH₂O (n = 6–3) in the present work. If successful, it would greatly contribute to our scientific understanding of “Soft Crystals,” changing crystal structure upon weak external stimulus. [Ir(tpy)₂]Br₃·7H₂O was found to emit the dual emission from the π–π* and XLCT states depending on the strength of interaction between the Ir(III) complex and Br⁻. The heating of [Ir(tpy)₂]Br₃·7H₂O contributes to not only the framework change but also the detachment of H₂O from the hydrated Br⁻. Thereby, the interaction between the Ir(III) complex and Br⁻ is strong and stabilizes the XLCT state in energy, resulting in the quenching of the π–π* emission. The switching of the excited state between the π–π* and XLCT states is one of the valuable findings in the present work.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI).

Supplementary information: crystallographic data and additional figures and tables. See DOI: https://doi.org/10.1039/d6ce00194g.

CCDC 2527091 ([Ir(tpy)₂]Br₃·7H₂O) and 2527097 ([Ir(tpy)₂]Br₃·2H₂O) contain the supplementary crystallographic data for this paper.^41a,b

Acknowledgements

We thank the Institute of Science Tokyo Core Facility Center for the provision of the dual wavelength single-crystal X-ray diffractometer (Rigaku XtaLAB Synergy-DW system). This work was supported by JSPS KAKENHI (grant number JP21K14647).

References

1. T. Abe, T. Itakura, N. Ikeda and K. Shinozaki, Luminescence color change of a platinum(II) complex solid upon mechanical grinding, Dalton Trans., 2009, 711–715.
2. Y. Nishiuchi, A. Takayama, T. Suzuki and K. Shinozaki, A Polymorphic Platinum(II) Complex: Yellow, Red and Green Polymorphs and X-ray Crystallography of [Pt(fdpb)Cl] (Hfdpb = 1,3-Di(5-trifluoromethyl-2-pyridyl)benzene), Eur. J. Inorg. Chem., 2011, 2011, 1815–1823.
3. J. Ni, X. Zhang, N. Qiu, Y.-H. Wu, L.-Y. Zhang, J. Zhang and Z.-N. Chen, Inorg. Chem., 2011, 50, 9090–9096.
4. X.-P. Zhang, J.-F. Mei, J.-C. Lai, C.-H. Li and X.-Z. You, J. Mater. Chem. C, 2015, 3, 2350–2357.
5. Y. Shigeta, A. Kobayashi, M. Yoshida and M. Kato, Cryst. Growth Des., 2018, 18, 3419–3427.
6. M. Martinez-Junquera, R. Lara, E. Lalinde and M. T. Moreno, J. Mater. Chem. C, 2020, 8, 7221–7233.
7. C.-Y. Lien, Y.-F. Hsu, Y.-H. Liu, S.-M. Peng, T. Shinmyozu and J.-S. Yang, Inorg. Chem., 2020, 59, 11584–11594.
8. Y. Yin, Z. Chen, R.-H. Li, C. Yuan, T.-Y. Shao, K. Wang, H. Tan and Y. Sun, Inorg. Chem., 2021, 60, 9387–9393.
9. M. Martínez-Junquera, E. Lalinde and M. T. Moreno, Inorg. Chem., 2022, 61, 10898–10914.
10. Y. Shigeta, T. Nomoto, M. Kato and M. Mizuno, Inorg. Chem., 2023, 62, 66–74.
11. H.-H. Zhang, S.-X. Wu, Y.-Q. Wang, T.-G. Xie, S.-S. Sun, Y.-L. Liu, L.-Z. Han and X.-P. Z.-F. Shi, Dyes Pigm., 2022, 197, 109857.
12. M. Cocchi, J. Kalinowski, V. Fattori, J. A. G. Williams and L. Murphy, Appl. Phys. Lett., 2009, 94, 073309.
13. S. J. Farley, D. L. Rochester, A. L. Thompson, J. A. K. Howard and J. A. G. Williams, Inorg. Chem., 2005, 44, 9690–9703.
14. S. Develay, O. Blackburn, A. L. Thompson and J. A. G. Williams, Inorg. Chem., 2008, 47, 11129–11142.
15. M. Cocchi, D. Virgili, V. Fattori, D. L. Rochester and J. A. G. Williams, Adv. Funct. Mater., 2007, 17, 285–289.
16. W. Mróz, C. Botta, U. Giovanella, E. Rossi, A. Colombo, C. Dragonetti, D. Roberto, R. Ugo, A. Valore and J. A. G. Williams, J. Mater. Chem., 2011, 21, 8653–8661.
17. M. Cocchi, J. Kalinowski, V. Fattori, J. A. G. Williams and L. Murphy, Appl. Phys. Lett., 2009, 94, 073309.
18. N. M. Barker, S. D. Taylor, E. Ferguson, J. A. Krause, A. G. Oliver, W. B. Connick and P. Zhang, Inorg. Chem., 2021, 60, 14731–14743.
19. C. L. Stubbs, M. J. Bryant, L. E. Hatcher and P. R. Raithby, CrystEngComm, 2025, 27, 7498–7513.
20. Y. Li, L. Chen, Y. Ai, E. Y.-H. Hong, A. K.-W. Chan and V. W.-W. Yam, J. Am. Chem. Soc., 2017, 139, 13858–13866.
21. S. Hattori, T. Nakano, N. Kobayashi, Y. Konno, E. Nishibori, T. Galica and K. Shinozaki, Dalton Trans., 2022, 51, 15830–15841.
22. M. Kato and K. Ishii, Soft Crystals: Flexible Response Systems with High Structural Order (The Materials Research Society Series), Springer, 2023.
23. M. Kato, H. Ito, M. Hasegawa and K. Ishii, Chem. – Eur. J., 2019, 25, 5105–5112.
24. S. Hattori, M. Kondo, A. Sekine and K. Shinozaki, Vapochromism of an iridium(III) bis-terpyridine complex based on the modulation of halide-to-ligand charge transfer transition, Dalton Trans., 2022, 51, 7068–7075.
25. P. Alam, M. Karanam, D. Bandyopadhyay, A. R. Choudhury and I. R. Laskar, Eur. J. Inorg. Chem., 2014, 3710–3719.
26. P. Alam, M. Karanam, D. Bandyopadhyay, A. R. Choudhury and I. R. Laskar, Eur. J. Inorg. Chem., 2014, 3710–3719.
27. E. V. Nykhrikova, M. A. Kiseleva, P. Kalle, S. S. Mariasina, S. A. Kozyukhin, S. V. Tatarin and S. I. Bezzubov, Inorg. Chem., 2025, 64, 5210–5220.
28. X. Zhu, P. Cui, S. Kilina and W. Sun, Inorg. Chem., 2017, 56, 13715–13731.
29. P. Bolle, H. Serier-Brault, R. Génois, E. Faulques, A. Boulmier, O. Oms, M. Lepeltier, J. Marrot, A. Dolbecq, P. Mialane and R. Dessapt, J. Mater. Chem. C, 2016, 4, 11392–11395.
30. Y. Jiang, G. Li, D. Zhu, Z. Su and M. R. Bryce, J. Mater. Chem. C, 2017, 5, 12189–12193.
31. Y. Wang, T. Yang, X. Liu, G. Li, W. Che, D. Zhu and Z. Su, J. Mater. Chem. C, 2018, 6, 12217–12223.
32. W. Ito, S. Hattori, M. Kondo, H. Sakagami, O. Kobayashi, T. Ishimoto and K. Shinozaki, Dual Emission from an Iridium(III) Complex/Counter Anion Ion Pair, Dalton Trans., 2021, 50, 1887–1894.
33. CrysAlisPro, Agilent Technologies Ltd, Yarnton, Oxfordshire, England, 2014.
34. G. M. Sheldrick, SHELXT – Integrated space-group and crystal-structure determination, Acta Crystallogr., Sect. A: Found. Adv., 2015, 71, 3–8.
35. G. M. Sheldrick, Crystal structure refinement with SHELXL, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 3–8.
36. C. B. Hübschle, G. M. Sheldrick and B. Dittrich, ShelXle: a Qt graphical user interface for SHELXL, J. Appl. Crystallogr., 2011, 44, 1281–1284.
37. C. F. Macrae, P. R. Edgington, P. McCabe, E. Pidcock, G. P. Shields, R. Taylor, M. Towler and J. van de Streek, Mercury: visualization and analysis of crystal structures, J. Appl. Crystallogr., 2006, 39, 453–457.
38. P. Atkins and J. de Paula, Atkins' Physical Chemistry, Oxford New York, 9th edn, 2010.
39. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman and D. J. Fox, Gaussian 16, Revision C.01, Gaussian, Inc., Wallingford CT, 2016.
40. F. Izumi and K. Momma, Solid State Phenom., 2007, 130, 15–20.
41. (a) CCDC 2527091: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2qtn29; (b) CCDC 2527097: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2qtn8h.

---