Chinese Physics Letters, 2020, Vol. 37, No. 6, Article code 066201 Velocity and Stability of Condensed Polymorphic SiH$_{4}$: A High-Temperature High-Pressure Brillouin Investigation * Jiayu Wang (王佳钰), Qiang Zhou (周强), Siyang Guo (郭思洋), Yanping Huang (黄艳萍), Xiaoli Huang (黄晓丽), Lu Wang (王璐)**, Fangfei Li (李芳菲)**, Tian Cui (崔田) Affiliations State Key Laboratory of Superhard Materials, College of Physics, Jilin University, Changchun 130012, China Received 28 March 2020, online 26 May 2020 *Supported by the National Key Research and Development Program (Grant No. 2017YFA0403704), the National Natural Science Foundation of China (Grant Nos. 11474127, 11574112, 11274137, and 11504127), the Program for Changjiang Scholars and Innovative Research Team in University (Grant No. IRT1132), and China Postdoctoral Science Foundation (Grant No. 2015M570265).
**Corresponding author. Email: wanglu@jlu.edu.cn; lifangfei@jlu.edu.cn
Citation Text: Wang J Y, Zhou Q, Guo S Y, Huang Y P and Huang X L et al 2020 Chin. Phys. Lett. 37 066201    Abstract Silane (SiH$_{4}$) is a promising hydrogen-rich compound for pursing high temperature superconducting. Previous high pressure measurements of Raman, x-ray diffraction and theoretical studies on SiH$_{4}$ mainly focused on its polymorphic structures above 50 GPa, while the structure and the stability under lower pressure range are still unclear. Here we report an investigation of condensed SiH$_{4}$ by Brillouin scattering measurements at high temperature up to 407 K and high pressure up to 18 GPa. Brillouin scattering frequencies of fluid SiH$_{4}$ under compression are obtained under isothermal conditions of 300 K, 359 K and 407 K. The SiH$_{4}$ becomes unstable with increasing temperature and subsequently decomposes into silicon and H$_{2}$. We find that compression at room temperature induces two velocity anomalies corresponding to a fluid-solid state transition and a phase IV to phase V transition, respectively. Brillouin scattering spectra has been a powerful tool to investigate the fruitful phases and structure transitions in the hydrogen-rich compound under extreme conditions. DOI:10.1088/0256-307X/37/6/066201 PACS:62.50.-p, 62.60.+v, 64.60.-i, 78.35.+c © 2020 Chinese Physics Society Article Text Silane (SiH$_{4}$), a hydrogen-rich compound, has been extensively used for industrial productions of solar grade silicon, polysilicon in thin film photovoltaics, semiconductors and liquid-crystal displays.[1] Recently the condensed SiH$_{4}$ draws plenty of attention for possible high temperature superconducting.[1–8] Following the pioneering idea of metallic hydrogen and chemical precompression by Ashcroft, the metallization pressure of hydrogen-rich compounds is estimated to be much lower than that of pure hydrogen.[2,3] Several high-pressure structures of SiH$_{4}$ have been predicted,[2–5] in which metallization pressure are estimated to vary from 20 GPa to 380 GPa. Yao et al. predicted that SiH$_{4}$ transforms to a metallic phase with monoclinic $C2/c$ structure at about 60 GPa.[4] Feng et al. suggested the SiH$_{4}$ with an $O3$ structure should be metallic close to 91 GPa, and the superconducting transition temperature $T_{\rm c}$ would be 166 K at 202 GPa.[5] Pickard and Needs pointed out that the $C2/c$ structure of SiH$_{4}$ is probably a good metal with a high $T_{\rm c}$.[6] Zhang et al. predicted that two metallic structures may be stable above 383 GPa, and their $T_{\rm c}$ may reach up to 29.65 K at 300 GPa and 31.57 K at 400 GPa.[7] The stability of SiH$_{4}$ is a key issue for the experiments. Under ambient condition, spontaneous combustion occurs as soon as SiH$_{4}$ exposes to air.[8] Chen et al.[9] compressed the SiH$_{4}$ in a diamond anvil cell (DAC) and speculated an insulator-to-metal transition by the Raman and IR reflectivity measurements. Meanwhile, Eremets et al.[10] reported the observation of SiH$_{4}$ metallization based on electrical resistance measurement. However, it was found later that their measured resistance drop is caused by the decomposition of SiH$_{4}$ into Si and H$_{2}$ at pressures above 50 GPa. The conduction mostly arises from a chemical reaction between H$_{2}$ and environment materials like Pt or Re rather than from the SiH$_{4}$.[11,12] Hanfland et al.[13] showed that SiH$_{4}$ becomes amorphous above 60 GPa and recrystallizes above 90 GPa into a polymeric structures. Nevertheless, the structures and the stability of SiH$_{4}$ are still under debate, and in addition no superconducting phase has been confirmed yet. Current investigations on SiH$_{4}$ are mainly focused on their polymorphic structures above 50 GPa by x-ray diffraction (XRD), Raman scattering, infrared reflectivity (IR) and theoretical calculations.[12–16] However, the structural stability under low pressure range before its decomposition is still unclear. Brillouin scattering spectra can provide velocity and elastic properties of materials, through which we can extract the structural stability under extreme conditions. They offer a good means to disclose the structure properties of compressed SiH$_{4}$.[17,18] In this Letter, we first report a high-temperature high-pressure (HTHP) Brillouin scattering measurements on condensed SiH$_{4}$ in a DAC up to the temperature of 407 K and the pressure of 18 GPa, the refractive indexes of fluid SiH$_{4}$ and the velocities of solid SiH$_{4}$ are presented and the elastic properties are unveiled. The SiH$_{4}$ gas from the Alfa Aesar (99.999% purity) was stored in a compressed gas cylinder. We loaded the gaseous SiH$_{4}$ into the BX90 DAC (with wide opening) chamber in a glove box purged with argon. The DAC was cooled down to a temperature below saline's boiling point when SiH$_{4}$ gas was puffed on, followed by an immediate sealing of the DAC as soon as the chamber was fully filled with liquid SiH$_{4}$. The diameter of the diamond anvil was 400 µm with a pre-drilled rhenium gasket as the sample chamber. The DAC could be heated by a resistance heater around the diamonds and the sample chamber, whose temperature was calibrated with a K-type thermocouple attached on the flank of the diamond. During measurements the temperature was controlled by a feedback power and the temperature error was normally controlled within 1 K. The pressure was also monitored by ruby fluorescence.[19] Taking into account of the temperature gradient around the sample chamber, the temperature dispersion between the anvil center and the side face of the diamond could not be ignored. In order to obtain a precise temperature on the sample, a ruby ball was placed in the sample chamber to measure the temperature dependence of the ruby fluorescence. The shifts of the luminescence at ambient pressure were recorded and compared with early publications.[20] The actual temperature in the center of sample chamber was thus calibrated for the following experiments. In this study, the Brillouin scattering spectrum could be collected in both back (180$^{\circ}\!$) and symmetric platelet (60$^{\circ}\!$) scattering geometries.[21] A single-frequency green 532 nm laser was used as the incident excitation source. The laser was a solid-state diode-pumped frequency-doubled Nd:Vanadate laser from Coherent Company. The scattered signals were collected by 3$+$3 pass tandem Fabry–Pérot interferometer designed by Sandercock.[22] For forward platelet scattering geometry measurements, the sound velocity of sample is calculated by[23] $$ V=\frac{\lambda _{0}\Delta V}{2n \sin{\left({\theta }/{2}\right)}},~~ \tag {1} $$ where $V$ is the acoustic velocity, $\Delta V$ is the corresponding Brillouin frequency shift in gigahertz (GHz), $\lambda_{0}$ is the wavelength of the incident beam, $n$ is the refractive index of sample, and $\theta$ is the angle between the incident and scattered beams. When a plate sample in DAC was studied, its sound velocity can be obtained to be independent of the sample's refractive index, which has been discussed elsewhere in detail.[24] To put it simply, in 60$^{\circ}\!$ platelet geometry, Eq. (1) can be expressed as $$ V _{60} = \Delta V _{60}\cdot\lambda_{0} ,~~ \tag {2} $$ It should be noticed that a complete dynamical theory of fluid sample is still absent, and has been approached by different levels of theory, including thermodynamic, hydrodynamic and more microscopic, detailed mechanism can be found elsewhere if interested.[25,26] The elementary excitation induced Brillouin shift in fluid is complicated and may not follow the linear dispersion relation, while we just calculate and use the velocity concept in later comparison and discussion for convenient. Then due to the isotropy in fluid, by collecting both 60$^{\circ}\!$ platelet and 180$^{\circ}\!$ back scattering signals, the refractive index of the sample can be estimated based on certain assumptions.[18] In the backscattering geometry Eq. (1) is rewritten as $$ V _{180}=\frac{\lambda _{0}\Delta V_{180}}{2n}.~~ \tag {3} $$ In this study, the liquid SiH$_{4}$ was heated up to a constant temperature, and then the pressure was increased manually until the sample finally solidifies. The temperature effect on liquid SiH$_{4}$ was compared between the isothermals. After solidification, only forward 60$^{\circ}\!$ scattering geometry measurements are performed as the velocity anisotropism appears. The solidification process was monitored either by a microscope or by analyzing the Brillouin scattering spectra.
cpl-37-6-066201-fig1.png
Fig. 1. High-pressure Brillouin spectra of liquid SiH$_{4}$ at 300 K (a) in the 60$^{\circ}\!$ forward platelet scattering geometry and (b) in the back scattering geometry. R refers to the Rayleigh peak.
Brillouin spectra of liquid SiH$_{4}$ under high pressures at 300 K were measured in both the 60$^{\circ}\!$ forward platelet scattering geometry and the back scattering geometry as shown in Fig. 1. The spectra are of good quality with a high signal-to-noise ratio. Both the longitudinal frequency shift peaks ($V_{60}$) and weak back scattering signal ($V_{180}$) can be observed in the platelet scattering geometry, but their intensity decreases dramatically with increasing pressure. The velocities of liquid SiH$_{4}$ are calculated by Eq. (2) as the foregoing explanation, and their pressure dependence is represented in Fig. 2(a). The velocity increases smoothly and monotonously up to 4.5 GPa at 300 K. Since the velocity in an isotropic medium is independent of direction, its refractive index can be calculated from the ratio of the Brillouin scattering frequency shifts by Eqs. (2) and (3). As shown in Fig. 2(b), the refractive index for liquid SiH$_{4}$ also increases smoothly and monotonously with pressure.
cpl-37-6-066201-fig2.png
Fig. 2. Pressure dependence of the velocity (a) and the refractive index (b) for liquid SiH$_{4}$ at 300 K.
cpl-37-6-066201-fig3.png
Fig. 3. Pressure evolutions of velocity along different isothermals. The upper inset in (b) presents the image showing a laser damage to the liquid SiH$_{4}$ at about 3.8 GPa and 407 K, and the lower inset gives the Raman spectrum of vibron of hydrogen detected in the sample chamber.
When the pressure is released at room temperature, the velocity decreases back to the original state, indicating stable molecular fluids within the measured pressure range. By resistance heating, the fluid SiH$_{4}$ in DAC was heated to higher temperatures, and pressure dependence of the velocities along 359 K and 407 K isothermals were collected. In Fig. 3(a) we can see that the variation of the velocity with temperature is smaller compared with its pressure dependence and just decreases slightly with increasing temperature at a similar constant pressure. The solidification pressure extends to a higher pressure of nearly 7 GPa at 359 K. However, when temperature increases up to 407 K, the transparent SiH$_{4}$ becomes opaque gradually at about 3.2 GPa, as shown in the photo in Fig. 3(b). Though the measured velocities along 407 K isothermal present monotonous increasing as usual, they most probably correspond to the SiH$_{4}$ decomposition. As the opaque area appeared at the locations irradiated by the laser, it might be caused by heating effect of the focused laser. However, no changes were seen during 300 K and 359 K measurements, suggesting that the SiH$_{4}$ becomes unstable at higher temperatures. After the DAC was released to lower pressure, the opaque position did not disappear, implying a nonreversible process. If we heated the sample again up to 407 K and gradually increased pressure to measure the Brillouin frequencies on those transparent areas with ultralow laser power, a clear drop of the refractive index could be observed in this run, as presented in Fig. 4(b). Therefore, the SiH$_{4}$ is speculated to disintegrate in the first run that releases H$_{2}$ and silicon microcrystal. When a certain amount of silicon microcrystal appears, the chamber becomes opaque. The mixture of H$_{2}$ and undecomposed SiH$_{4}$ gives a decreased velocity and a drop of refractive index along the second 407 K isothermal run. To confirm this, the vibron of hydrogen was measured through Raman scattering measurements, as shown in the inset of Fig. 3(b).
cpl-37-6-066201-fig4.png
Fig. 4. Pressure dependence of refractive index $n$ in fluid SiH$_{4}$ along different isothermals.
At room temperature the liquid state of SiH$_{4}$ exists to about 4.5 GPa, then it turns into solid state with the transverse mode velocity appearing, as shown in Fig. 5. The longitudinal ($V_{\rm L}$) mode has a strong intensity than the transverse ($V_{\rm T}$) mode, and their intensities both decrease remarkably with increasing pressure. After crystallization, the SiH$_{4}$ has different sound velocities in planar 180$^{\circ}\!$ rotational measurements. The Brillouin scattering frequencies of solid SiH$_{4}$ increase monotonically in all directions, at one of which the frequencies is compared under compression and presented in Fig. 5.
cpl-37-6-066201-fig5.png
Fig. 5. Brillouin scattering spectra of SiH$_{4}$ at different pressures. R refers to the Rayleigh scattering, $V_{\rm L}$ is the longitudinal mode and $V_{\rm T}$ is the transverse mode.
Upon compression, a continuing increase of longitudinal frequency could be seen. However, above 13 GPa a sudden drop was observed followed by a monotonic increase of the Brillouin frequency. To better understand these behaviors, the direction dependence of sound velocities at different pressures and the pressure dependence of the velocities within the whole pressure range are plotted in Fig. 6. After crystallization, the rollercoaster of the velocity anisotropy is clearly seen along each direction, while the curvatures of velocity variations keep constant under pressure up to 11.9 GPa. Then the anisotropy becomes dramatically larger than before. The maximal difference of the velocities between different direction is about 1.1 km/s, while it was only 0.5 km/s below 13 GPa. Thus there may be a phase transition above 13 GPa.
cpl-37-6-066201-fig6.png
Fig. 6. Direction dependence of velocities for SiH$_{4}$ at different pressures and room temperature. Longitudinal and transverse velocities are denoted by $V_{\rm L}$ (a) and $V_{\rm T}$ (b), respectively. (c) Pressure dependence of velocities of the SiH$_{4}$. The maximum (red circles), minimum (green circles) and averaged (black square) velocities are shown by different symbols for clearness. The upper red dotted lines show a series of phase separations from Chen's document in Ref. [9].
The pressure dependence of the velocity variations in Fig. 6(c) can be divided it into three areas: one fluid phase and two crystallized phases. The liquid-to-solid phase transition is confirmed by the appearance of transversal velocity, and the enhancement of the velocity anisotropy of solid-to-solid phase transition is at about 13 GPa. In the figure, the maximum and minimum velocities propagating along different directions are shown with red and green symbols separately, and the averaged velocity is given by solid black circles. All velocities increase smoothly with pressure without any changes of the anisotropy. However, above 13 GPa, the minimum $V_{\rm L}$ decreases about 6.4% and the minimum $V_{\rm T}$ decreases about 15.8%. The velocity anisotropic increases more than double across the boundary. The structural transition of SiH$_{4}$ within the low pressure range is complicated and unclear till now. Chen et al. reported the Raman and IR spectra of SiH$_{4}$ and showed that there are three phase transitions below 20 GPa,[9] including the fluid state to phase III, phase III to phase IV and finally to phase V. Their phase separations are plotted in Fig. 6(c) for a direct comparison. The fluid-to-solid phase transition around 4.5 GPa observed in this work is consistent with Chen's report. Based on Raman measurements, they observed a discontinuity in the $V_{1}$, $V_{2}$, and $V_{3}$ Raman-active modes at 6.5 GPa, which indicates the transition from phase III to IV. The transition from phase III to phase IV however cannot be identified from this velocity measurement. This suggests that the transition between phase III and phase IV may be an isostructural transition where no enormous change occurs. The transition at about 13 GPa we found in this study may correspond to the phase IV to V transition presented by Chen et al., though there appeared a little difference between these two transition pressures.
cpl-37-6-066201-fig7.png
Fig. 7. Pressure–temperature phase diagram of SiH$_{4}$. The measured pressure points of three isotherms are labeled by circles are given in the figure.
In addition, a previous structural investigation on SiH$_{4}$ suggested the $P2_{1}/c$ symmetry between 10 GPa and 25 GPa,[27] corresponding to the phase V existing between 10 GPa and 27 GPa.[9] Then, there is no other structural information available for the complex phase transitions of SiH$_{4}$. The lack of the space group information becomes a main constraint on elastic constants analysis. However, we can still provide probable stability estimations based on the velocity variations. The Brillouin scattering spectra can be used as a novel method to investigate the phase transitions and stability under extreme conditions. The high pressure Brillouin scattering spectra of SiH$_{4}$ are collected along three isotherm experiments at 300, 359, and 407 K. The measured data points from this work are shown in Fig. 7 together with previous literature data.[9] The phase diagram of SiH$_{4}$ is extended up to 407 K. In summary, by the high-pressure high-temperature Brillouin scattering measurements, the Brillouin frequency shifts, velocities, refractive index $n$ and their respective pressure dependencies are all calculated and presented here. The velocities and refractive indexes of fluid SiH$_{4}$ along 300 K, 359 K and 407 K isothermal experiments are obtained and compared under pressure, both of which increase monotonously and smoothly with increasing pressure but decrease slightly with the increasing temperature at a similar pressure. High temperature weakens the stability of SiH$_{4}$ molecules, leading to partial decomposition, into silicon and H$_{2}$. The pressure partly enhances the interactions in solid SiH$_{4}$. The velocity variations prove the transitions from fluid to solid and finally to phase V, consistent with early Raman reports. The transition from phase III to phase IV may be an isostructural process so that no obvious velocity jumps or curvature change occurs.
References From monosilane to crystalline silicon. Part III. Characterization of amorphous, hydrogen-containing silicon productsHydrogen Dominant Metallic Alloys: High Temperature Superconductors?Metallic Hydrogen: A High-Temperature Superconductor?Superconductivity in high-pressure SiH 4Structures and Potential Superconductivity in SiH 4 at High Pressure: En Route to “Metallic Hydrogen”High-Pressure Phases of SilaneHigh-temperature Superconductivity in compressed Solid SilaneSome mechanistic problems in the kinetic modeling of monosilane pyrolysisPressure-induced metallization of silaneSuperconductivity in Hydrogen Dominant Materials: SilaneFormation of transition metal hydrides at high pressuresHigh-pressure study of silane to 150 GPaHigh-Pressure Synthesis, Amorphization, and Decomposition of SilaneEstimation of the superconducting parameters for silane at high pressureHydrogen segregation and its roles in structural stability and metallization: silane under pressureElectrical Resistivity of Silane Multiply Shock-Compressed to 106 GPaAcoustic velocities, refractive index, and elastic constants of liquid and solid CO 2 at high pressures up to 6 GPaBrillouin scattering study of liquid methane under high pressures and high temperaturesTemperature dependence of the ruby luminescence method for measuring high pressuresCalibration of the ruby R 1 and R 2 fluorescence shifts as a function of temperature from 0 to 600 KBrillouin scattering at high pressure: an overviewBrillouin scattering study of SbSI using a double-passed, stabilised scanning interferometerThe velocity, refractive index, and equation of state of liquid ammonia at high temperatures and high pressuresElastic moduli of NaCl by Brillouin scattering at high pressure in a diamond anvil cellA continuum mechanics theory of depolarized and polarized Rayleigh-Brillouin light scattering spectra of supercooled liquidsThermal relaxation and brillouin scattering in liquidsCrystal structure of Si H 4 at high pressure
[1] Odden J O, Egeberg P K and Kjekshus A 2005 J. Non-Cryst. Solids 351 1317
[2] Ashcroft N W 2004 Phys. Rev. Lett. 92 187002
[3] Ashcroft N W 1968 Phys. Rev. Lett. 21 1748
[4] Yao Y, Tse J S, Ma Y et al 2007 Europhys. Lett. 78 37003
[5] Feng J, Grochala W, Jaron T et al 2006 Phys. Rev. Lett. 96 017006
[6] Pickard C J and Needs R J 2006 Phys. Rev. Lett. 97 045504
[7] Zhang H, Jin X, Lv Y et al 2015 Sci. Rep. 5 8845
[8] Becerra R and Walsh R 1992 J. Phys. Chem. 96 10856
[9] Chen X J, Struzhkin V V, Song Y et al 2008 Proc. Natl. Acad. Sci. USA 105 20
[10] Eremets M I, Trojan I A, Medvedev S A et al 2008 Science 319 1506
[11] Degtyareva O, Proctor J E, Guillaume C L et al 2009 Solid State Commun. 149 1583
[12] Strobel T A, Goncharov A F, Seagle C T et al 2011 Phys. Rev. B 83 144102
[13] Hanfland M, Proctor J E, Guillaume C L et al 2011 Phys. Rev. Lett. 106 095503
[14] Durajski A P and Szczȩśniak R 2014 Mod. Phys. Lett. B 28 1450052
[15] Cui W W, Shi J M, Liu H Y et al 2015 Sci. Rep. 5 13039
[16] Zhong X F, Liu F S, Cai L C et al 2014 Chin. Phys. Lett. 31 126201
[17] Shimizu H, Kitagawa T and Sasaki S 1993 Phys. Rev. B 47 11567
[18] Li M, Li F F, Gao W et al 2010 J. Chem. Phys. 133 044503
[19] Yen J and Nicol M 1992 J. Appl. Phys. 72 5535
[20] Ragan D D, Gustavsen R and Schiferl D 1992 J. Appl. Phys. 72 5539
[21] Polian A 2003 J. Raman Spectrosc. 34 633
[22] Sandercock J 1970 Opt. Commun. 2 73
[23] Li F F, Li M, Cui Q L et al 2009 J. Chem. Phys. 131 134502
[24] Whitfield C H, Brody E M and Bassett W A 1976 Rev. Sci. Instrum. 47 942
[25] Wang C H 1986 Mol. Phys. 58 497
[26] Mountain R D 1966 J. Res. Natl. Bur. Stand. Sect. A 70A 207
[27] Degtyareva O, Martinez Canales M, Bergara A et al 2007 Phys. Rev. B 76 064123