Advances and Physicochemical Property Investigations by Methylene Chain Length of Perovskite-type Layer Crystals [NH3(CH2)nNH3]CdCl4 (n=2, 3, and 4)

Hybrid perovskites have potential applications in several electrochemical devices such as supercapacitors, batteries, and fuel cells. However, despite various potential applications, there have been limited studies on compounds containing Cd. Therefore, in this study, the structures and phase transition temperatures T C of organic – inorganic perovskite [NH 3 (CH 2 ) n NH 3 ]CdCl 4 ( n =2, 3, and 4) were confirmed by X-ray diffraction and DSC experiments. The thermal stabilities obtained by TGA and DTA results were considered as a function of the length n of the CH 2 groups in the cation. In addition, structural characteristics and molecular dynamics of the cation and anion near T C were studied by 1 H MAS NMR, 13 C MAS NMR, 14 N static NMR, and 113 Cd MAS NMR experiments. From these results, regardless of whether n is even or odd, the differences in thermal and physical properties were minimal. Rather, the difference in molecular motion relative to the length of the cation was seen only at higher temperatures. These results provide insights into the thermal stability and molecular dynamics of [NH 3 (CH 2 ) n NH 3 ]CdCl 4 crystals and are expected to facilitate potential applications. MX 3, the the cations and the hybrid anion Cd, dynamics. These hybrid perovskites have potential applications in several electrochemical devices such as supercapacitors, batteries, and fuel cells.


Experimental
An aqueous solution containing NH2(CH2)nNH2·2HCl and CdCl2 was slowly evaporated at 300 K to produce single crystals of [NH3(CH2)nNH3]CdCl4 (n=2, 3, and 4). The structures of the [NH3(CH2)nNH3]CdCl4 crystals at 298 K were analyzed using an X-ray diffraction system equipped with a Cu-Kα radiation source. 30 The lattice parameters were determined by single-crystal X-ray diffraction at the Seoul Western Center of the Korea Basic Science Institute (KBSI). The crystals were mounted on a Bruker D8 Venture equipped with an 1 μS micro-focus sealed tube Mo-Kα and a PHOTON III M14 detector. 30 DSC (DSC 25, TA Instruments) measurements for the three crystals were carried out at a scanning speed of 10 °C/min between 190 and 600 K under nitrogen gas. TGA and DTA experiments were performed on a thermogravimetric analyzer (TA Instrument) at the same heating rate between 300 and 873 K under N2 gas. 31 In addition, optical observations were made using an optical polarizing microscope in the temperature range of 300-680 K, where the as-grown single crystals were placed on the heating stage of a Linkam THM-600.
NMR spectra of [NH3(CH2)nNH3]CdCl4 crystals were obtained using a Bruker 400 MHz Avance II+ solidstate NMR spectrometer equipped with 4 mm MAS probes at the Seoul Western Center, KBSI. The Larmor frequencies for 1 H MAS NMR and 13 C MAS NMR experiments were 400.13 and 100.61 MHz, respectively. The MAS rate to minimize the spinning sideband was 10 kHz, and the NMR chemical shifts were recorded using tetramethylsilane (TMS) as the standard. 32 The T1ρ values were obtained using a π/2−τ pulse followed by a spin-lock pulse of duration τ, and the width of the π/2 pulse for 1 H and 13 C was in the range Please do not adjust margins Please do not adjust margins of 3.4-3.62 μs. In addition, static 14 N NMR and 113 Cd MAS NMR spectra were measured with Larmor frequencies of 28.90 and 88.75 MHz, respectively. The 14 N NMR experiments were performed using a solid-state echo sequence: 4 μs-τ-4 μs-τ; τ=5 μs for n=2, τ=8 μs for n=3 and 4. The 113 Cd MAS NMR experiments were performed using a π/2−τ pulse followed by a spin-lock pulse of duration τ, and the width of the π/2 pulse for 113 Cd was 3.2 μs. The chemical shift measurements referenced NH4NO3 and CdCl2O8·6H2O as standard samples. The temperature was changed by adjusting the nitrogen gas flow and heater current, and it was maintained within ±0.5 K.

Phase transition temperatures and thermal properties
The DSC curves of [NH3(CH2)nNH3]CdCl4 crystals at a heating rate of 10 K/min under a nitrogen atmosphere are shown in Fig. 3. No peak was observed for the case of n=2, whereas only one endothermic peak for Please do not adjust margins Please do not adjust margins n=3 was observed at 374 K (=TC). Finally, for the case of n=4, two endothermic peaks were observed at 341 K (=TC2) and 366 K (=TC1). These phase transition temperatures are consistent with those reported previously. 6,16,27 To verify whether the endothermic peaks correspond to phase transition or decomposition, TGA and DTA experiments were performed at the same heating rate. The TGA and DTA curves displayed in Fig. 4 show that the crystals with n = 2, 3, and 4 are almost stable up to approximately 493, 539, and 536 K, respectively; according to the number n of CH2 groups in the carbon chain, the molecular weight loss near 493, 539, and 536 K marks the onset of partial thermal decomposition (at temperature Td).
[NH3(CH2)nNH3]CdCl4 undergoes loss in the molecular weight with increasing temperature. The amount remaining as solid residues can be calculated from the molecular weights. When n=2, the loss of 12% and 23% of its weight at temperatures of about 622 K and 804 K was due to the decomposition of HCl and 2HCl, respectively (see Fig. 4(a)). The small endothermic peak at 374 K on the DTA curve for [NH3(CH2)3NH3]CdCl4 is assigned to the phase transition detected in the DSC experiment. Additionally,   weight losses of 11% and 22% occurred at temperatures of 613 K and 623 K, respectively (see Fig. 4(b)). Finally, in the case of [NH3(CH2)4NH3]CdCl4 crystals with n=4, the two small endothermic peaks at 341 K and 366 K on the DTA curve are attributed to the phase transition seen in the DSC result. At temperatures of 612 K and 623 K, 11% and 21% of its weight was respectively lost (see Fig. 4(c)). The molecular weight of the three crystals decreased sharply between 550 and 650 K. In the case of n=3 and 4 near 800 K, weight losses of 45% occurred, whereas when n=2, the weight loss was the smallest at 23%.
To support the TGA results, the appearance of single crystals with changing temperature was observed with an optical polarizing microscope (Fig. 5). For the case where n=2, the crystal formed at 300 K was colorless and transparent, while it appeared slightly opaque above 547 K. Upon increasing the temperature to 622 K, HCl was eliminated, and the crystal turned orange. Finally, the surface near 633 K appeared to melt slightly. For the case where n=3, the crystal had an opaque white color at room temperature. Upon increasing the temperature to 673 K, it remained opaque even though the 2HCl was blown away. For the case where n=4, the crystal was transparent at 300 K, and it turned opaque white with increasing temperature, likely indicating the elimination of HCl. Finally, it turned bright brown as 2HCl was lost near 670 K.  Fig. 6. At 300 K, the 1 H chemical shift for NH3 and CH2 were obtained at 7.96 and 4.78 ppm, respectively, in the case of n=2, and at 7.57 and 3.23 ppm, respectively, in the case of n=3. Finally, for n=4, they were obtained at 6.90 and 4.69 ppm, respectively. It can be seen that the 1 H chemical shifts for NH3 when n=2, 3, and 4 are similar, while the 1 H chemical shifts for CH2 are very different. In the case where n=3, no change was observed near TC. In particular, in the case where n=4, the 1 H chemical shift for CH2 shows discontinuity near TC2 and continuity near TC1. The chemical shifts for NH3 of all three crystals are nearly independent of temperature, which means that the surrounding environment of the 1 H in NH3 does not significantly change with temperature. However, in the case where n=4, the surrounding environments of 1 H in CH3 do significantly change with temperature. The 1 H MAS NMR spectra were measured with several delay times at each given temperature, and the plot of spectral intensities vs. delay times were found to follow a single exponential function. The decay rate of the spin-locked proton magnetization is characterized by the spin-lattice relaxation time T1ρ as [33][34][35] where P(τ) and P(0) are the signal intensities at time τ and τ=0, respectively. From the slope of the logarithm of intensities vs. delay times plot, the 1 H T1ρ values were determined for NH3 and CH2 at several temperatures. The 1 H T1ρ results are shown in Fig. 7 for the three compounds as a function of inverse temperature. In the cases where n=2 and 3, as the temperature increases, the T1ρ values increase rapidly from 1 to 700 ms, and then rapidly reduces at temperatures above 350 K. In the case where n=3 and 4, two 13 C resonance lines were obtained for CH2-1 and CH2-2, respectively. At 300 K, the 13 C chemical shift for n=2 was recorded at 37.36 ppm, and those for n=3 were observed at 25.00 and 39.09 ppm for CH2-1 and CH2-2, respectively. The 13 C chemical shifts for n=4 were observed at 25.30 and 42.85 ppm, respectively. Here, the chemical shifts for CH2-1 are similar for these crystals, but those for CH2-2 are different between n=2, 3, and 4. The full width at half maximum (FWHM) for 13 C NMR at 300 K are relatively narrow, ranging from 1.2 to 1.6 ppm. Meanwhile, the chemical shifts for the in-situ 13 C MAS NMR spectra for the three compounds are shown in Fig. 8 with increasing temperature. In the case where n=2 (Fig. 8(a)), the 13 C chemical shifts for CH2-2 increases slightly as the temperature increases. Some chemical shift changes around 290 K can also be seen. In case where n=3 ( Fig. 8(b)), the slopes of the chemical shifts marked by red dotted lines are temperature dependent, with more variation for CH2-2 than for CH2-1. In addition, there was no   change in chemical shifts near TC. However, in the case where n=4 (Fig. 8(c)), the chemical shifts show discontinuity near TC2, whereas they show continuity near TC1. It can be seen that the chemical shifts of CH2-2 near TC2 changes more than those of CH2-1. In the three compounds, the chemical shifts of CH2-2 (more so than that of CH2-1) are thought to be affected by the N sites bonded to both ends of CH2-2.
The 13 C MAS NMR spectrum showed a change in intensity with increasing delay time at each temperature. All these decay curves could be described by a single exponential function, and from the slope of their recovery traces, the 13 C T1ρ values for CH2-1 and CH2-2 were obtained for the three compounds and plotted as a function of 1000/T, as shown in Fig. 9. In the case where n=2, the 13 C T1ρ value first decreased slightly with increasing temperature and then decreased rapidly at higher temperatures. In the case where n=3, it decreased slightly with increase in temperature, then increased again, and finally decreased at temperatures above TC. Meanwhile, the minimum T1ρ values (34.74 and 28.40 ms for CH2-1 and CH2-2, respectively) occur at 280 K. In the case where n=4, T1ρ decreases slightly as the temperature increases, and then increases rapidly near TC2. Similar to that in the case of n=3, this tendency results from molecular motion below the phase transition temperature. There are distinct molecular motions, and the minimum T1ρ is due to the molecular motion of CH2-1 and CH2-2 in the [NH3(CH2)nNH3] cations in the case where n=3 and 4, respectively. These T1ρ values could be described by the correlation time τC for the molecular motion, and the T1ρ value for the molecular motion is given by 36,37 T1ρ -1 = C (γH 2 γC 2 ħ 2 / r 6 )[4fa + fb + 3fc + 6fd + 6fe], where fa = τC /[1 + ω1 2 τC 2 ], fb = τC /[1 + (ωH -ωC) 2 τC 2 ], fc = τC /[1 + ωC 2 τC 2 ], fd = τC /[1 + (ωH + ωC) 2 τC 2 ], and fe = τC /[1 + ωH 2 τC 2 ]. Here, C is a coefficient, γH and γC are the gyromagnetic ratios for 1 H and 13 C, respectively, ħ is the reduced Planck constant, r is the internuclear distance, ωH and ωC are the Larmor frequencies of 1 H and 13 C, respectively, and ω1 is the frequency of the spin-lock field. Here, the 13 C T1ρ values were measured using the spin-locking pulse sequence with a locking pulse of ω1=75.76 kHz for n=3 and ω1=70.42 kHz for n=4. When ω1τC=1, T1ρ has the minimum value. Therefore, a relationship between T1ρ and ω1 was applied to obtain the coefficient C in Eq. (2). Using this coefficient, τC was calculated as a function of temperature. According to the Bloembergen-Purcell-Pound (BPP) theory, the local field fluctuation is governed by the thermal motion of CH2-1 and CH2-2. The correlation time τC for molecular motion at several temperatures follows the Arrhenius equation 33 where Ea and kB are the activation energy of the motions and Boltzmann constant, respectively. The magnitude of Ea depends on the molecular dynamics. The plot of log τC vs. 1000/T provided the Ea values for CH2-1 and CH2-2, as shown in Fig. 10 38 The 14 N NMR spectrum for n=2, 3, and 4 differed with increasing temperature, as shown in Fig. 11. Here, the measurements were performed by keeping the c-axis of the single crystals parallel to the direction of the magnetic field. In the case of n=2, the two resonance lines of one pair decreased with increasing temperature, then decreased to a minimum near 400 K, and then increased again. In the case of n=3, the chemical shifts for six resonance lines due to the three pairs caused a large change near TC. The symbols with the same color below TC indicate the same pairs for 14 N. Near 374 K (=TC), the number of resonance lines and chemical shifts of the NMR spectrum showed abrupt changes; three pairs turned into just two pairs. The changes in the 14 N chemical shift as a function of temperature were attributed to the variations in the structural geometry. In addition, the chemical shifts of the 14 N signals below TC changed almost continuously, and the chemical shifts for 14 N above TC remained constant with temperature. The 14 N NMR spectrum exhibits a reduction in the NMR lines from three to two pairs of lines at the phase transition TC. Finally, when n=4, the four resonance lines due to two inequivalent N sites showed no change in temperature below TC2; however, it was difficult to detect because the line width suddenly increased above TC2. In the cases of n=3 and 4, the different N spectra were explained as follows. None of the previously reported X-ray results 6,16,24,27 include different N sites; therefore, two or three different N sites are thought to have a twin domain due to the ferroelastic property in materials with the organic-inorganic perovskite structure reported recently.  113 Cd has an isotopic abundance of 12.3% and a spin of I=1/2. 113 Cd NMR spectroscopy has been used to examine the structure and dynamics of various inorganic and organic materials. The 113 Cd MAS NMR spectra for the three crystals were obtained as a function of temperature, as shown in Fig. 12. The chemical shifts are related to the local field at the position of the resonating nucleus in the crystals. 39 The 113 Cd chemical shifts for the three crystals shift slightly in the negative direction as the temperature increases. However, in the case where n=4, it was discontinuous near TC2. This result suggests that in the cases when n=2 and 3, there is no significant change in the environment near Cd depending on the temperature, whereas in the case of n=4, the environment around Cd changes near the phase transition temperature. In addition, the difference in Cd chemical shifts for the three materials seems to be due to the difference in the Cd-Cl distance as shown in Table  1. In other words, it is consistent that the Cd-Cl distance increases as the n value increases. Similarly, the change in the chemical shift near TC2 is attributed to the shortening of the Cd-Cl distance from the average 2.646 Å to 2.6198 Å. 16 The 113 Cd MAS NMR spectrum for three crystals measured the change in intensity with various delay times at 300 K. The decay curves were described by a single exponential function, and the 113 Cd T1ρ values were obtained from the slope of their recovery traces. In the cases where n=2, 3, and 4, the T1ρ values were 2058, 1512, and 1101 ms, respectively. All the T1ρ values for 1 H, 13 C, and 113 Cd at 300 K are listed in Table 2. 113 Cd T1ρ was very long compared to 1 H T1ρ and 13 C T1ρ. The long 113 Cd T1ρ values with n=2 at 300 K were considered to be more rigid than the others (n=3 and 4) because the Cd-Cl length is short as shown in Table 1; a longer T1ρ indicates that the transfer of energy from the nuclear spin system to the surrounding environment is not very easy.  13 C, and 113 Cd chemical shifts are continuous near TC1, but discontinuous around TC2. This is attributed to a large change in the lattice parameters c and β near TC2. The molecular dynamics near the phase transition temperatures were analyzed in terms of the T1ρ values for 1 H and 13 C. 1 H and 13 C T1ρ are almost continuous when n=2 and 3, but when n=4, they show a discontinuity near TC2, as with the chemical shifts. Overall, n=2 and 3 showed a rapid decrease in T1ρ of 1 H and 13 C at high temperatures, whereas n=4 showed an increase at high temperatures. The short 13 C T1ρ values at low temperatures in the case of n=4 were considered to be more flexible than the others because the N-C-C-C-C-N lengths are long.

Conclusion
The crystal structures were found to be monoclinic when n was even and orthorhombic when n was odd, but there were little differences in thermal and physical properties. The difference in molecular motion according to the length of the cation was observed only at high temperatures. In the [NH3(CH2)nNH3] cations, 13 C T1ρ with n=4 at low temperature has shorter values than those with n=2 and 3. It was found that as the carbon length increased, it became more flexible. Overall, 113 Cd T1ρ at 300 K is very long compared to the 1 H and 13 C T1ρ values. This is thought to be related to the Cd-Cl bond length. These results provide insight into the thermal stability and molecular dynamics of [NH3(CH2)nNH3]CdCl4 (n=2, 3, and 4) crystals as a function of the length of CH2 groups in the carbon chain and are expected to facilitate future research as well as potential applications in supercapacitors, batteries, and fuel cells.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.