X-ray and NQR studies of bromoindate(III) complexes: [C₂H₅NH₃]₄InBr₇, [C(NH₂)₃]₃InBr₆, and [H₃NCH₂C(CH₃)₂CH₂NH₃]InBr₅

2.1 Crystal structure determination of tetrakis(ethylammonium) heptabromoindate(III) [C₂H₅NH₃]₄InBr₇ (1)

A freshly prepared needle crystal (1N) was picked for the crystal structure determination, because the crystals turned gradually into massive ones on keeping them in the mother liquor. The crystal structure data obtained at 100 K are listed in Table 1. As described in the next section, it was found that the crystal of 1N underwent a phase transition. Thus, this structure corresponds to that of the low-temperature phase (LTP), 1N_LTP. The projection of a unit cell down the a axis is shown in Fig. 1. The asymmetric unit is composed of one octahedral [InBr₆]³⁻ ion, one Br⁻ ion, and four C₂H₅NH₃⁺ ions. This structure is similar to those of [CH₃NH₃]₄InBr₇ (P2₁/c, a=1683.8(6), b=772.2(2), c=2086.4(6) pm, β=128.67(2)°, Z=4) [3] and [CH₃NH₃]₄InCl₇ (P2/c, a=1611.3(3), b=746.6(5), c=2030.0(8) pm, β=127.9(4)°, Z=4) [4], though their asymmetric unit contains two octahedral [InBr₆]³⁻ or [InCl₆]³⁻ ions, but one in 1N_LTP. The In−Br bond lengths vary from 262.27(4) to 273.98(4) pm in 1N_LTP, as listed in Table 2, whereas variations from 266 to 268 pm are found in [CH₃NH₃]₄InBr₇ [3]. The bond angles ∠Br−In−Br (cis) range from 86.60(1) to 93.95(2)° in 1N_LTP, as shown in Table 3. The corresponding angles in [CH₃NH₃]₄InBr₇ [3] change from 89.6 to 90.2°. The slightly larger deviations from the regular octahedron for the [InBr₆]³⁻ ion in 1N_LTP than in [CH₃NH₃]₄InBr₇ arise from the differences in both shapes of cations and hydrogen bond schema around anions. The hydrogen bonds in 1N_LTP are depicted in Fig. 2 and their parameters are listed in Table 4.

Fig. 1:

Projection of the crystal structure of 1N_LTP at 100 K down the crystallographic a axis.

Fig. 2:

Hydrogen bond scheme around the bromoindate and bromide anion in 1N_LTP.

2.2 NQR and DTA measurements of [C₂H₅NH₃]₄InBr₇ (1); its self-decomposition to [C₂H₅NH₃]₃InBr₆+[C₂H₅NH₃]Br

The temperature-dependence curves of ⁸¹Br NQR frequencies of 1N measured in the range from 77 to around 300 K are shown in Fig. 3. In accordance with the crystal structure at 100 K, six ⁸¹Br NQR lines showing the same intensities were observed. These lines are assigned to the nonequivalent Br atoms of [InBr₆]³⁻, because the Br⁻ line should have, if observed, a far lower frequency, less than a few MHz. As seen in Fig. 3, the disappearance of the resonance lines was found between around 110 and 200 K for the highest- and lowest-frequency lines and between around 125 and 160 K for the remaining lines. The disappearance phenomenon may be due to a phase transition as detected by DTA measurements described below. The respective curves of the outer two lines connect smoothly between the low- and high-temperature parts on extrapolation. However, the close existence of the inner resonance lines makes not only the continuity of the curves but also the number of resonance lines themselves ambiguous in the higher temperature region. On the other hand, the NQR spectrum is outstanding for its widespread frequencies spanning to almost 32 MHz at 77 K despite the discrete anion. Furthermore, the ν versus T curve of the lowest-frequency line has an unusually large positive coefficient, though the line has normally a negative value due to the thermal lattice vibrations. The observation may be caused by a dynamical event of the intermolecular interaction related to the Br atoms of [InBr₆]³⁻. In many cases the intermolecular bonds lead to lower frequencies of the relevant lines compared with those of atoms free from such bonds. Then, when the intermolecular bonds are subjected to weakening at increasing temperatures, the positive temperature dependence is seen as a result. It seems that all Br atoms of the [InBr₆]³⁻ ion are more or less involved in the hydrogen bonds with short contacts of N…Br, as listed in Table 4. Among them, the Br3 atoms have the shortest contact with 337.6(3) pm and the longest In−Br bond length with 273.98(4) pm, which indicates the formation of the strongest hydrogen bond. By considering this situation the Br3 atom seems to be the most probable candidate assignable to the lowest NQR frequency line. If this is the case, the trans-effect may allow us the further assignment of the highest line to the Br6 atom situated in a trans-position of the Br3 atom. This effect predicts that the average NQR frequency of the atoms situated in a trans-position of the octahedral complex is kept nearly constant and the corresponding resonance lines, thus, changing their frequencies symmetrically for the counterpart with temperature. This is the consequence of the fact that the bonding in the octahedral complex is well explained using the 3c-4e bond model.

Fig. 3:

Temperature dependence of ⁸¹Br NQR frequencies of the newly prepared sample 1N.

In the course of the NQR measurements, we found that the aged samples 1O_(10Y) and 1O_(2Y), kept at ambient temperatures for about 10 and 2 years, respectively, produced other extra NQR lines in addition to those of 1N, as shown in Fig. 4. The DTA measurements for the fresh sample of 1N showed an exothermic peak at 182 K on the cooling run, whereas the endothermic peaks appeared at 208 and 300 K on the heating run, as depicted in Fig. 5a. The lower-temperature phase transition at 182 K in the cooling run and 208 K in the heating run is a first-order type accompanied with a large hysteresis. The higher-temperature phase transition at 300 K is a second-order type by judging from the peak shape. This phase transition should be accompanied by only a slight structural change, as the corresponding anomaly was not found in the cooling run and also no anomalous changes occurred on the ν versus T curves of ⁸¹Br NQR lines in the vicinity of this temperature. The DTA curve of the aged sample of 1O_(6M) left for 6 months after preparation is shown in Fig. 5b. On cooling, the exothermic peak split into two peaks without any notable changes in the endothermic peak. Moreover, in the second run after 1 week, no heat anomaly appeared on cooling, but an exothermic peak strangely appeared at 163 K and an endothermic peak shifted to 213 K on heating. On the other hand, the sample 1O_(10Y) also exhibited heat anomalies but at different temperatures, at 301 K on the cooling run and at 310 K during heating, as shown in Fig. 6a. Furthermore, the sample 1O_(2Y) showed a different feature of the DTA curve, as shown in Fig. 6b, in which no heat anomaly appeared during cooling, but the endothermic peak which was observed at around 320 K in a first heating cycle did appear. Interestingly, the endothermic peak diminished in size on the successive second heating run, though no difference in the peak size could be observed in the case of 1O_(10Y). The NQR and DTA results indicate the occurrence of a structural change from 1N to 1O_(10Y). To confirm this, the C, H, and N elemental analyses were applied to the specimens of 1O_(10Y) and 1O_(2Y): found C 9.98, H 3.41, N 5.80 for 1O_(10Y); found C 10.41, H 3.52, N 6.12 for 1O_(2Y). When these values are compared with the calculated ones: C 9.83, H 3.30, N 5.73 for [C₂H₅NH₃]₃InBr₆ and C 11.19, H 3.75, N 6.52 for [C₂H₅NH₃]₄InBr₇ (1N), the values for 1O_(10Y) coincide closely with the former values and those for 1O_(2Y) are situated at an intermediate between the former and the latter. Therefore, it is deduced that a decomposition reaction, [C₂H₅NH₃]₄InBr₇ (1N)→[C₂H₅NH₃]₃InBr₆+[C₂H₅NH₃]Br, proceeds gradually at ambient temperature. The DTA curves in Fig. 6a therefore correspond to those of [C₂H₅NH₃]₃InBr₆. Furthermore, the DTA curves in Figs. 5b and 6b may be attributable to those of the mixtures in the transformation process from [C₂H₅NH₃]₄InBr₇ to [C₂H₅NH₃]₃InBr₆.

Fig. 4:

Temperature dependence of ⁸¹Br NQR frequencies of extra NQR lines of 1O_(10Y) and 1O_(2Y). The lowest-frequency line showed frequency changes on every measurment.

Fig. 5:

(a) DTA curve of 1N. The upper curve indicates the cooling run and the lower curve the heating run. (b) DTA curve of 1O_(6M). The upper curve indicates the cooling run and the lower curve the heating run.

Fig. 6:

(a) DTA curve of old sample 1O_(10Y). The upper curve is the cooling run and the lower curve the heating run. (b) DTA curve of the old sample 1O_(2Y). The upper curve is the cooling run and the lower curve the heating run.

2.3 Crystal structure determination of tris(guanidinium) hexabromoindate(III) [C(NH₂)₃]₃InBr₆ (2)

The crystal structure of 2 has been determined from data collected at 100 K. The projection of the unit cell of 2 down the a axis is shown in Fig. 7. The asymmetric unit consists of three octahedral [InBr₆]³⁻ ions and six C(NH₂)₃⁺ ions. The respective ions have the local symmetry shown in the brackets as follows: [(In1)Br₆]³⁻ [E], [(In2)Br₆]³⁻ [C_2v], [(In3)Br₆]³⁻ [D_2h], and all C(NH₂)₃⁺ [E]. Wyckoff positions are 8f, 4e, 4a, and 8f for In1, In2, In3, and all Br atoms, respectively. Therefore, 12 nonequivalent Br atoms exist in 2. The In−Br bond lengths are in the range of 261.86(2) to 272.39(5) pm (Table 2), and the bond angles ∠Br−In−Br (cis) range from 85.38(1) to 93.24(1)° (Table 3). These values are comparable with the corresponding values of 262.27(4) to 273.98(4) pm and 86.60(1) to 93.95(2)° in 1. The hydrogen bond network built up between anions and cations can be seen in Fig. 8.

Fig. 7:

Projection of the crystal structure of 2 at 100 K down the crystallographic a axis.

Fig. 8:

Hydrogen bond scheme around the bromoindate anion in 2.

2.4 NQR measurements of [C(NH₂)₃]₃InBr₆ (2)

The temperature dependence of ⁸¹Br NQR frequencies of 2 observed between 77 K and around 300 K is shown in Fig. 9. According to the crystal structure, 12 ⁸¹Br NQR lines with the same intensity are expected, but only 10 ⁸¹Br NQR lines could be observed, with the remaining 2 NQR lines missing. The overlapping between a number of ⁸¹Br and ⁷⁹Br NQR lines makes it difficult to completely resolve the respective lines. However, the monotonic temperature dependence suggests no occurrence of a phase transition in the measured temperature range. Interestingly, some NQR lines show positive temperature coefficients, whereas the others show negative but quite small temperature coefficients. This indicates that all the Br atoms are involved more or less in hydrogen bonds, as found in the crystal structure, which may be weakened according to the excitation of cation motions with increasing temperatures.

Fig. 9:

Temperature dependence of ⁸¹Br NQR frequencies of 2.

2.5 Crystal structure determination of 2,2-dimethyl-1,3-propanediammonium pentabromoindate (III) of [H₃NCH₂C(CH₃)₂CH₂NH₃]InBr₅ (3)

The crystal structure of 3 has been determined from diffraction data collected at 100 K. The projection of the unit cell of 3 down the b axis is shown in Fig. 10. The asymmetric unit consists of one tetrahedral [InBr₄]⁻, one double-octahedral [In₂Br₁₁]⁵⁻, and three H₃NCH₂C(CH₃)₂CH₂NH₃²⁺ ions. The [InBr₄] tetrahedron of the central In3 atom is severely distorted with In3−Br bond lengths ranging from 250.2(10) to 256.34(10) pm and ∠Br−In3−Br bond angles from 100.20(3) to 132.00(4)°, as listed in Tables 2 and 3, respectively. The [In₂Br₁₁]⁵⁻ ion is a corner-sharing dimer of two [InBr₆] octahedra with the central In1 and In2 atoms connected through the Br6 atom. The bridging bonds In1−Br6 and In2−Br6, being 297.82(9) and 301.66(9) pm, respectively, are considerably longer compared to the terminal bonds ranging from 262.01(9) to 270.36(9) pm in the In1 octahedron and from 260.05(10) to 267.26(10) pm in the In2 octahedron. There are thus 15 nonequivalent Br atoms in the crystal structure. One of three 2,2-dimethyl-1,3-propanediammonium cations, including C11 as a central atom of the propane skeleton, is disordered, with a CH₃ and a CH₂NH₃ groups placed between two positions leaving the remaining CCH₂NH₃ groups ordered. The hydrogen bond network of the ions is shown in Fig. 11.

Fig. 10:

Projection of the crystal structure of 3 at 100 K down the crystallographic b axis. Cations with the center cabon atom C11 are disordered.

Fig. 11:

Hydrogen bond scheme around [In₂Br₁₁]⁵⁻ and [InBr₄]⁻ ions.

2.6 NQR and DTA measurements of [H₃NCH₂C(CH₃)₂CH₂NH₃]InBr₅(3)

The temperature dependence of the ⁸¹Br NQR frequencies in 3 observed up to around 310 K is shown in Fig. 12. All NQR lines except for the line situated at 70 MHz were observed as doublets with the intensity ratio of about 1:2 at 77 K. The spectra of representative doublet lines at 59, 62, and 80 MHz are shown and compared with those of two lines well separated in frequency in Fig. 13. The frequency difference between two lines in each doublet is less than 200 kHz, which is nearly equal to the quench frequency of 180 kHz in the super-regenerative oscillator-detector used for measurements. However, the spectra of the doublets are apparently different in their appearance from the side bands, as shown in Fig. 13. The stronger components of these doublets take the higher- or lower-frequency positions independently of each other. With increasing temperature the weaker components of respective doublets are weakened rapidly in their intensities to become unobservable under the noises, whereas the stronger ones can be observed. As a result, five ⁸¹Br NQR lines are observed in the searched range from 40 to 110 MHz, though the crystal structure requires 15 nonequivalent Br atoms. The resonance frequencies of the observed lines are almost comparable with those of the ⁸¹Br NQR lines in 1 and 2, indicating that these lines are attributable to the terminal Br atoms of the double-octahedral [In₂Br₁₁]⁵⁻ ion, even though the number of lines is decreased to a half of those expected from the crystal structure at 100 K. On the other hand, the DTA measurements of 3 revealed the occurrence of phase transitions shown by an exothermic heat anomaly at 184 K accompanied by a small peak at 187 K on the cooling run and the endothermic one at 189 K while heating, as shown in Fig. 14. The hysteresis for the anomaly and the peak shape show the phase transition to be a first-order one. On the other hand, the ν versus T curves of ⁸¹Br NQR lines show smooth changes in the vicinity of the phase transition points without showing any recognizable anomalies. This fact may be understandable if the phase transition is induced by the change in cation dynamics and is followed by almost no substantial changes in the anion structure. The crystal structure at 100 K in the LTP has revealed that one of the three cations is partially disordered. Furthermore, it is noticed that the disordered NH₃ groups of N6 and N6′ are hydrogen-bonded to 5 atoms of the 10 terminal Br atoms (Br2, Br3, Br7, Br8, and Br11), as plotted in Fig. 15. When the disorder fluctuates with the electric field around these Br atoms, it is plausible that the intensities of ⁸¹Br NQR lines are weakened to become unobservable and only half of the lines of the 10 terminal Br atoms are thus observed. Then, the weaker components of the doublets observed around 77 K should be attributable to the Br atoms being hydrogen bonded to the disordered cation, which become quickly unobservable with increasing fluctuation. On the other hand, the absence of an indication on the temperature dependence curves of ⁸¹Br NQR lines at the phase transition can be realized, if the phase transition is characterized by the conversion between a static and a dynamic disorder structure of the disordered cation. The disordered cations fixed at one of two sites in the LTP can thus exchange their positions between two possible sites dynamically in the room-temperature phase (RTP). Furthermore, the dynamic disordering seems to be frozen via two steps to the static disordered state, as two exothermic peaks follow the phase transition from RTP to LTP. As a result, the ⁸¹Br NQR lines of five terminal Br atoms in the [In₂Br₁₁]⁵⁻ ions which take part in hydrogen bonding with the ordered cation have been observed throughout the measured temperature range, whereas those of the remaining five terminal Br atoms which take part in the hydrogen bonding with the disordered cation have been observed partially only in the vicinity of 77 K. The ⁸¹Br NQR line of the bridging Br6 atom with the angle of 164.23(4)° in the [In₂Br₁₁]⁵⁻ ion and the ⁸¹Br NQR lines of the [InBr₄]⁻ ion have not been observed in this experiment, as they appear beyond the frequency range we searched up to 110 MHz. Higher frequencies are expected according to the MO calculations as described in the next section. The NQR frequencies of the bridging halogen atoms having a bridging angle of nearly 180° are known to be higher than those of terminal halogen atoms [5].

Fig. 12:

Temperature dependence of ⁸¹Br NQR frequencies of 3.

Fig. 13:

NQR spetra collected at around 77 K. These spectra are obtained after numerical integration of raw recorded spectra.

Fig. 14:

DTA curve of 3. The upper curve is the cooling run and the lower curve is the heating run.

Fig. 15:

Hydrogen bond scheme around disordered N6 and N6′ atoms at 100 K.

2.7 MO calculations

MO calculations were applied to the respective anions of 1 and 3 to estimate the ⁸¹Br NQR data (frequency and asymmetry parameter) and the bond angles defined as the angles between the z principal axes of EFG tensors. In the calculations, the crystal field effect was not considered. Details of calculation are described elsewhere [5]. The ⁸¹Br NQR frequencies (MHz) and the asymmetry parameters calculated are as follows: For the [InBr₆]³⁻ ion in 1, (Br1: 92.02 MHz, 0.00), (Br2: 91.37 MHz, 0.00), (Br3: 87.23 MHz, 0.00), (Br4: 93.31 MHz, 0.00), (Br5: 92.72 MHz, 0.00), and (Br6: 92.40 MHz, 0.00) with the averaged frequency of 91.51 MHz; for the [InBr₄]⁻ ion in 3, (Br12: 145.83 MHz, 0.01), (Br13: 165.84 MHz, 0.00), (Br14: 164.05 MHz, 0.00), and (Br15: 145.52 MHz, 0.01) with the averaged frequency of 155.24 MHz ; for the [In₂Br₁₁]⁵⁻ ion in 3, (Br1: 76.59 MHz, 0.00), (Br2: 98.09 MHz, 0.01), (Br3: 94.58 MHz, 0.01), (Br4: 93.56 MHz, 0.01), (Br5: 91.06 MHz, 0.01), (Br6: 163.73 MHz, 0.12), (Br7: 112.38 MHz, 0.02), (Br8: 106.54 MHz, 0.02), (Br9: 97.62 MHz, 0.02), (Br10: 99.16 MHz, 0.03), and (Br11: 80.73 MHz, 0.00) with the averaged frequency of 95.03 MHz except for bridging Br6. The average values of the calculated frequencies are approximately 1.32 and 1.37 times higher than the observed ones for the [InBr₆]³⁻ and [In₂Br₁₁]⁵⁻ ions, respectively. The contribution of the crystal field effects to the EFGs around the relevant Br atoms is thus considered to be roughly 25%, which may reduce the ⁸¹Br NQR frequencies. The angles between the z principal axes of the EFG tensors are listed in Table 3 and compared with the bond angles from the X-ray structures. For the [InBr₆]³⁻ ion of 1, both angles are in good accordance with deviations within 1°. For the [InBr₄]⁻ and [In₂Br₁₁]⁵⁻ ions of 3, the angles are also in good accordance, but with several exceptions where the maximum difference reaches up to around 10°. Especially large discrepancies are seen between the angles related to the terminal Br14 atom of [InBr₄]⁻ and the bridging Br6 atom of [In₂Br₁₁]⁵⁻. These discrepancies for the [InBr₄]⁻ and [In₂Br₁₁]⁵⁻ ions of 3 can be attributed to the packing of the molecules which incorporates two kinds of complex anions with different coordination structures. The packing with maximum crystallization energy may cause deviations of the bond directions from the principal axes.

4.1 Sample preparation

A stock solution of InBr₃ was prepared by dissolving In metal (>99.9%, Sigma-Aldrich) in 48% aqueous HBr solution (special grade, Wako) and by adding water resulting in 20% HBr content for the syntheses of all compounds. The crystals of 1 were prepared by adding 0.2 mol of [C₂H₅NH₃]Br (special grade, Wako) into InBr₃ stock solution containing 0.05 mol of InBr₃. Colorless needle-like crystals (1N) appeared after the removal of water from the solution by P₂O₅ in a desiccator at room temperature. On keeping the crystals in the mother liquor, 1N turned gradually into massive crystals. A needle crystal of 1N for the X-ray measurement was picked from the freshly prepared solution. The crystals of 2 were obtained during an attempt of preparation of [C(NH₂)₃]InBr₄ by adding 0.1 mol of [C(NH₂)₃]₂CO₃ (first grade, Wako) and 0.3 mol of HBr as 48% aqueous solution into InBr₃ stock solution containing 0.2 mol of InBr₃. On removing the water from the solution by P₂O₅ in a desiccator, colorless square prismatic crystals appeared. A suitable crystal was selected from them for the X-ray measurement. The crystals of 3 were obtained by adding 0.1 mol of 2,2-dimethyl-1,3-proanediamine and 0.3 mol of HBr as 48% aqueous HBr solution into InBr₃ stock solution containing 0.1 mol of InBr₃. Concentrating the solution by heating, feather-like crystals appeared initially, which turned into colorless tabular crystals after dissolving by the addition of water and removing water by P₂O₅ in a desiccator. A tabular crystal was picked for the X-ray measurement. – Elemental analyses for 1N: calcd. C 11.19, H 3.75, N 6.52; found C 10.80, H 3.65, N 6.30; for 2: calcd. C 4.65, H 2.34, N 16.27; found C 4.59, H 2.30, N 15.84; for 3: calcd. C 9.70, H 2.60, N 4.52; found C 9.77, H 2.56, N 4.52.

4.2 NQR and DTA measurements

The ⁸¹Br and ⁷⁹Br NQR spectra were recorded by a home-made super-regenerative type spectrometer at temperatures above 77 K. The resonance frequencies were determined by a counting method. The accuracy of the frequency measurements is estimated to be within ±0.05 MHz. DTA was measured by a home-made apparatus at temperatures above 100 K.

4.3 Crystal structure determination

All measurements were made on a Bruker D8 Venture PHOTON area detector (MoK_α radiation, λ=71.073 pm). Data were collected at 100 K and processed using Olex2 1.2 [6], [7]. The structures were solved by Direct Methods [8], [9] and expanded using Fourier techniques. All calculations were performed using the Olex2 1.2 crystallographic software package except for refinements which were performed using Shelxl [10], [11], [12]. Positional and displacement parameters and other experimental details can be obtained from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_requetst/cif (CCDC 1431879 for 1N, 1431880 for 2, and 1431881 for 3).

4.4 MO calculation

The detailed description of the MO calculation on PM3 basis using Gamess (Windows version 2013), a quantum chemistry package [13], was reported elsewhere [5]. The population analysis was done for 4p orbitals of Br atoms in the [InBr₆]³⁻ ion in 1, and the [In₂Br₁₁]⁵⁻ and [InBr₄]⁻ ions in 3. The structural information of the ions was taken from the crystal structures determined. Facio 19.1.4 [14], [15] and Mercury 3.6 [16], [17] were used for the support of the MO calculations.