High-Pressure Spectroscopy Study of Zn(IO3)2 Using Far-Infrared Synchrotron Radiation

We report the first high-pressure spectroscopy study on Zn(IO3)2 using synchrotron far-infrared radiation. Spectroscopy was conducted up to pressures of 17 GPa at room temperature. Twenty-five phonons were identified below 600 cm−1 for the initial monoclinic low-pressure polymorph of Zn(IO3)2. The pressure response of the modes with wavenumbers above 150 cm−1 has been characterized, with modes exhibiting non-linear responses and frequency discontinuities that have been proposed to be related to the existence of phase transitions. Analysis of the high-pressure spectra acquired on compression indicates that Zn(IO3)2 undergoes subtle phase transitions around 3 and 8 GPa, followed by a more drastic transition around 13 GPa.


Introduction
Metal iodates [M(IO 3 ) x ] receive great attention because of their non-linear optical properties. In particular, they are promising materials for second-harmonic generation (SHG) [1][2][3][4]. They have been also proposed for applications as dielectric materials, in non-linear optics, and for desalination and water treatment [1][2][3][4]. Amongst the iodates, the quasi-two-dimensional zinc iodate, Zn(IO 3 ) 2 , has recently been studied by several research groups [5][6][7][8]. The crystal structure, thermal stability, and other physical properties of this compound have been characterized at ambient pressure [9][10][11]. However, nothing is known of its structural and vibrational behavior under high-pressure (HP) conditions. At ambient conditions, Zn(IO 3 ) 2 crystallizes into a monoclinic structure, which is represented in Figure 1. The monoclinic structure consists of ZnO 6 octahedral units connected by IO 3 triangular pyramidal units. In these units, the pentavalent iodine atom forms three covalent bonds with oxygen atoms, leaving two 5s electrons free to act as lone electron pairs (LEP). The presence of the LEP has been shown to induce an interesting HP behavior in iodates related to Zn(IO 3 ) 2 like Fe(IO 3 ) 3 [12,13], with two isostructural phase transitions taking place at pressures below 10 GPa in this compound. These transitions have been detected from changes in the pressure dependences of phonon frequencies, which show a nonlinear behavior, as well as kinks and slope changes in the frequency-versus-pressure plot at the identified transition pressures [12,13]. Such unusual behavior of phonons has been proposed to be connected with gradual modifications of the iodine coordination, which is favored by the LEP of iodine and has been connected with the isostructural phase transitions discovered in Fe(IO 3 ) 3 [12,13]. Since the LEP of iodine is a typical feature of many M(IO 3 ) x iodates, it would be interesting to explore if low-pressure phase transitions are a typical feature of this family of compounds. Given the structural similarities between Fe(IO 3 ) 3 and Zn(IO 3 ) 2 (both structures consist of MO 6 octahedral units connected by iodine atoms coordinated by three oxygen atoms with non-bonding LEP orbitals), the possibility of isostructural phase transitions at low pressures in the unexplored Zn(IO 3 ) 2 motivated the present work, in which we studied the pressure dependence of phonons by means of HP infrared (IR) spectroscopy. This is a technique useful for understanding the changes induced by compression in the physical-chemical properties of iodates. Additionally, the discovery of the existence of phase transition in Zn(IO 3 ) 2 could be very useful for the development of solid-state cooling technologies, which take advantage of the pressureinduced barocaloric effect [14]. Infrared spectroscopy has been shown to be an effective diagnostic for the detection of pressure-induced phase transitions in solids [15]. In the present work, IR spectroscopy provides information on the vibrational properties of Zn(IO 3 ) 2 , which, until the present work, have remained relatively underexplored. In fact, less than one-third of the 51 phonons predicted by group theory analysis have been reported experimentally for Zn(IO 3 ) 2 [5][6][7][8][9]16]. Therefore, the present far-infrared spectroscopy study of Zn(IO 3 ) 2 under high-pressure conditions makes a timely and valuable contribution. Zn(IO 3 ) 2 has not previously been studied under high-pressure conditions. The information obtained from this study is relevant not only to improve the knowledge on the aforementioned issues but also to provide an accurate determination of the phonon frequencies, which are needed to properly model other physical properties such as heat capacity and thermal expansion. The use of high-brilliance infrared synchrotron radiation facilitated the measurement of tiny samples loaded in diamond-anvil cells (DACs) with a very good signal-to-noise ratio. Using this methodology, we have been able to identify 25 phonons between 98 and 600 cm −1 . We then follow the pressure-induced evolution of these phonons up to 17 GPa. The analysis of the results leads us to propose the existence of three phase transitions, as will be discussed in the manuscript.

Materials and Methods
Experiments were performed on Zn(IO 3 ) 2 powders synthesized from aqueous solution according to the synthesis method and sample characterization found in Ref. [6]. The crystal structure was confirmed by powder XRD measurements (X'Pert Pro diffractometer, Panalytical, Almelo, The Netherlands) using Cu K α1 radiation, which corroborated the crystal structure reported by Liang et al. [10] (space group P2 1 ) with unit-cell parameters: a = 5.465(4), b = 10.952 (8), c = 5.129(4) Å, and γ = 120.37(8) • . HP Fourier transform infrared (FTIR) measurements (Vertex 70 spectrometer, Bruker Optik GmbH, Ettlingen, Germany) were conducted at MIRAS beamline of the ALBA synchrotron. Zn(IO 3 ) 2 samples were loaded in a DACs designed for IR spectroscopy, using IIAC-diamonds with culets of 300 µm. Stainless-steel gaskets were pre-indented to a thickness of 40 µm and drilled with a hole in the center of 150 µm in diameter. Cesium iodide (CsI) was used as the pressure-transmitting medium (PTM) [17]. The CsI PTM is not quasi-hydrostatic beyond 3 GPa; however, radial pressure gradients were smaller than 1 GPa in the pressure range of this study [18]. CsI was chosen because it has the widest IR transmission window amongst the possible PTMs [18]. Pressure was determined using the ruby scale [19]. Synchrotronbased FTIR-micro-spectroscopy experiments were performed in the transmission mode of operation. We used a masking aperture size of 50 × 50 µm 2 and a beam current inside the synchrotron ring of 250 mA. The measurements were performed by employing a 3000 Hyperion microscope coupled to a Vertex 70 spectrometer (Bruker Optik GmbH, Ettlingen, Germany). The microscope was equipped with a helium-cooled bolometer detector optimized for operation in the range covering the far-infrared spectral region. A Mylar beam splitter was used in the spectrometer. Spectra were measured using a 15× Schwarzschild magnification objective (NA = 0.52) coupled to a 15× Schwarzschild magnification condenser. Measurements at selected pressures were collected using the OPUS 8.2 software (Bruker Optik GmbH, Ettlingen, Germany) in the 90-600 cm −1 range with a spectral resolution of 4 cm -1 and 256 co-added scans per spectrum. The analysis of FTIR results was carried using the Multiple Peak Fit Tool of the OriginPro software (OriginLab Corporation, Northampton, MA, United States) and employing Gaussian functions to model the peaks.

Results and Discussion
The results at the lowest measured pressure (0.9 GPa) are shown in Figure 2 together with the multiple-peak fit used to identify phonons. We have identified 25 modes, as can be seen in the figure, thereby greatly extending the number of modes previously observed in zinc iodate. The frequencies of all 25 of these modes are summarized in Table 1 and compared with those reported in the literature. Zn(IO 3 ) 2 exhibits 51 modes (26A + 25B) according to group theory. All of these 51 modes are both Raman-active and IR-active. They can be described as internal vibrations of the of the IO 3 units and external vibrations (commonly known as lattice modes) involving the relative movements of IO 3 (behaving as rigid units) and Zn, which are observed in the low-frequency region [7,20]. Symmetric (ν 1 ) and asymmetric (ν 3 ) stretching vibrations of the pyramidal IO 3 units are in the 780−630 cm −1 and 820−730 cm −1 wavenumber regions, respectively [7,21,22]. The symmetric (ν 2 ) and asymmetric bending vibrations (ν 4 ) are in the 400−320 cm −1 wavenumber region and around 450−400 cm −1 , respectively [7,21,22]. In our experiments, we detected four ν 4 modes. The frequencies of the modes we found at 425, 440, and 452 cm −1 are less than 9 cm −1 larger than those measured in previous ambient-pressure Raman and IR experiments [6,7,16]. This is consistent with the fact that our experiment was performed at a pressure of 0.9 GPa. The expected phonon hardening [23], or increase in the phonon frequency, is due to the shortening bond distance caused by increasing pressure in our experiment. The mode reported at 405 cm −1 in Ref. [5] was not detected in previous works from the literature, but in our case could be probably assigned to a shoulder of the peak with wavenumber 388 cm −1 , which can be identified at 402 cm −1 . Additionally, the mode previously reported at 524 cm −1 in Ref. [8] is not consistent either with the present or previous experiments [5][6][7]16]. Indeed, there are no phonons in Zn(IO 3 ) 2 , or in any divalent metal iodate [7], in the 600−500 cm −1 range. The origin of such a phonon could be related with twinning of the crystal structure when large single crystals are grown [9], being a possible overtone, since 524 cm −1 is approximately double of the frequency of the lattice modes.  Table 1. Table 1. A comparison of the phonon frequencies determined by IR-and Raman-spectroscopy from this work (at 0.9 GPa) and from previous studies (ambient pressure). For the present study, frequencies are given with errors. The horizontal dashed line indicates the cut-off of the IR set-up, below which it was not possible to detect signal due to experimental constraints. Regarding the other bending vibrations, five ν 2 modes were detected in our experiments. The frequencies of three of them agree well with previous studies [6,7,16], with ours observed at 3-5 cm −1 higher frequencies due to the higher pressure (0.9 GPa) in the sample. The brilliance and resolution of the experimental set-up enabled the identification of two modes that were previously undetected. Additionally, the mode reported at 366 cm −1 in Ref. [5] was not observed in our measurements, and it was not reported in the rest of the works in the literature [6][7][8]16]. We note that the frequency of this mode is approximately double the frequency of the 183 cm −1 mode detected in the present work and elsewhere in the literature [6][7][8]16] and that it is therefore likely to be an overtone.

Assignment
Regarding the lattice modes, sixteen in total were observed. Ten of these modes have been detected in previous studies [7,16], thereby providing good agreement with the results of the present work (especially considering the frequency increase due to the 0.9 GPa experimental sample pressure). In the literature, four modes have been detected by Raman and IR spectroscopy [7,8,16], which could be detected in our experiment because it is below the cut-off frequency of our setup. Lattice modes are mainly related to translation and libration movements of IO 3 , and the coupling of these movements with movements of Zn atoms. In particular, the 145 cm −1 phonon has been assigned to a O-Zn-O deformation of the ZnO 6 octahedron [8]. This is consistent with the fact that a mode involving the same deformation of ZnO 6 has the same frequency in ZnWO 4 [24] and ZnMoO 4 [25]. Additionally, the modes below 120 cm −1 are likely to be due to pure translational or librational movements of IO 3 , because of the large mass of the iodine atom. This hypothesis is consistent with the observation that these modes have nearly the same frequency in Mn(IO 3 ) 2 , Ni(IO 3 ) 2 , Co(IO 3 ) 2 , and Zn(IO 3 ) 2 [8] as well as in Fe(IO 3 ) 3 [10].
We will now comment on the data acquired at higher pressures. For the sake of accuracy, we will concentrate on wavenumbers higher than 150 cm −1 because the signals at lower energy became noisy with increasing pressure. Figure 3 displays IR spectra measured at different pressures (indicated in the figure). It is clear in Figure 3 that there are qualitative changes in the spectra at 3.6 GPa, in particular for wavelengths smaller than 300 cm −1 . In particular, three modes (marked by asterisks) increase in intensity. The transition pressure is close to the loss hydrostaticity of the pressure medium (3 GPa) [18]. The influence of non-hydrostaticity in phase transitions induced by pressure in Zn(IO 3 ) 2 is beyond the scope of this study. Additional changes occur in the IR spectra at 8.8 GPa.
In particular, several modes broaden, and three low-frequency modes (also marked with asterisks) become enhanced. Finally, there is a considerable broadening of phonon bands at 13 GPa and higher pressures. Similar changes have been assigned to phase transitions in the case of Fe(IO 3 ) 3 [12,13]. Thus, phase transitions could be the cause of the changes observed in the IR spectra. Further evidence supporting the existence of phase transitions in Zn(IO 3 ) 2 is obtained via the analysis of the pressure dependence of the mode frequencies described immediately below.   We will comment now on the pressure dependence of the ten lattice modes that were observed under sample compression. For these modes, the changes in the frequency pressure dependence, at phase transitions, are more evident than in the bending modes, as can be seen in Figure 5. In particular, slope changes can be seen at the transition pressures (3.6, 8.8, and 13 GPa), which are represented by vertical lines in the figure. Such changes are very noticeable in the lowest-frequency mode shown in Figure 5. They are also quite evident for the mode at 208 cm −1 at 0.9 GPa. As happened in the bending modes, the changes at 13 GPa are more noticeable than changes at the other transitions. In particular, there are discontinuities in the frequencies of at least two modes. The changes in the frequency pressure dependences at the phase transitions detected at 3.6 and 8.8 GPa, and the strongly non-linear pressure dependences, are qualitatively similar to those observed for phonons in Fe(IO 3 ) 3 at 2 and 6 GPa [12,13]. In Fe(IO 3 ) 3 , these changes have been related to isostructural phase transitions occurring in Ref. [12,13], which are themselves related to gradual pressure-induced changes in the coordination sphere of iodine, which are affected by the presence of lone electron pairs. The similarities between Zn(IO 3 ) 2 and Fe(IO 3 ) 3 therefore provide further support to the observation of phase transitions in the Zn(IO 3 ) 2 investigated here. Future studies using other characterization techniques, including X-ray diffraction and Raman spectroscopy should be performed to confirm the present interpretation of results and to determine if the transitions at 3.6 and 8.8 GPa are isostructural or not. We hope our work will trigger such studies as well as computer simulation studies.
Before concluding, we would like to add a comment on the transition at 13 GPa. The changes observed around 13 GPa are more apparent than those observed around 3.6 and 8.8 GPa. This suggests that the changes around 13 GPa can be linked to the occurrence of a first-order structural transition. The fact that the changes are accompanied by a marked broadening of phonon bands suggests a disorder of the crystal structure [26], which could be related to the presence of non-hydrostatic stresses at 13 GPa and higher pressures [27].
To conclude, we note that all observed modes harden under compression. In the pressure range up to 3.6 GPa (before the first phase transition), the pressure coefficients are all within the range of 1-9 cm −1 /GPa. This can be seen in Table 2, where we represent phonons frequencies (ω) and pressure coefficients (dω/dP) for different modes in the different phases. We have named the phases as follows: phase I (low-pressure phase), II, III, and IV (successive HP phases). Table 2 shows clearly the change in the pressure dependences at 3.6 GPa and 8.8 GPa as well as the changes in frequencies and pressure dependencies at 13 GPa, supporting the existence of the proposed transitions. The pressure coefficients in the low-pressure phase are comparable to pressure coefficients in Fe(IO 3 ) 3 [13], LiIO 3 [28], and KIO 3 [29] in the same frequency region, indicating that Zn(IO 3 ) 2 is extremely compressible, similar to these other iodates. Additionally, the pressure coefficients for the different HP phases are of the same order of magnitude as in the low-pressure phase, suggesting that the changes in the coordination polyhedra are not drastic and that the transition is probably related to gradual changes in cation coordination number [30]. Table 2. Phonon frequencies (ω) of the modes observed under compression for phases I (at 0.9 GPa), II (at 3.6 GPa), III (at 8.8 GPa, and IV (at 13 GPa). The pressure coefficients (dω/dP) obtained from linear fits are also given.

Conclusions
Synchrotron far-infrared spectroscopy measurements have allowed us to determine that Zn(IO 3 ) 2 undergoes three phase transitions at 3.6, 8.8, and 13 GPa. The phase transitions are identified from changes in the infrared spectra. The first two transitions resemble those previously observed in Fe(IO 3 ) 3 at similar pressures and are probably isostructural transitions favored by the presence of lone electron pairs in Zn(IO 3 ) 2 . The third phase transition appears to be a first-order transition that is connected to the occurrence of more important structural changes. Assignment of phonon modes has been discussed and their pressure dependence reported. We found that the lattice modes are more sensitive than the bending modes of IO 3 to pressure-induced structural changes.