Hydrogen-Storing Salt NaCl(H$_2$) Synthesized at High Pressure and High Temperature

TL;DR

This study reports the synthesis of a novel hydrogen-storing salt NaCl(H₂) at high pressures and temperatures, revealing its structure, stability, and potential for hydrogen storage applications.

Contribution

It provides experimental evidence and theoretical predictions for new NaCl-hydrogen compounds formed under extreme conditions, expanding understanding of salt-hydrogen interactions.

Findings

NaClHₓ formed at >1500 K and >40 GPa with a specific crystal structure.

NaClHₓ remains stable down to 17 GPa upon pressure release.

Prediction of a hydrogen-rich phase NaCl(H₂)₄ stable above 40 GPa.

Abstract

X-ray diffraction and Raman scattering measurements, and first-principles calculations are performed to search for the formation of NaCl-hydrogen compound. When NaCl and H$_{2}$ mixture is laser-heated to above 1500 K at pressures exceeding 40 GPa, we observed the formation of NaClH$_{\textit{x}}$ with $\textit{P}$6$_{3}$/$\textit{mmc}$ structure which accommodates H$_{2}$ molecules in the interstitial sites of NaCl lattice forming ABAC stacking. Upon the decrease of pressure at 300 K, NaClH$_\textit{x}$ remains stable down to 17 GPa. Our calculations suggest the observed NaClH$_{\textit{x}}$ is NaCl(H$_{2}$). Besides, a hydrogen-richer phase NaCl(H$_{2}$)$_{4}$ is predicted to become stable at pressures above 40 GPa.

Tables

Table 1. Table S1: NaCl(H 2 ) with ABAC stacking at 30 GPa

Space group P63/mmc,		a = 3.515 Å, c = 6.465 Å
		x	y	z
Na	2c	1/3	2/3	1/4
Cl	2a	0	0	0
H	4f	1/3	2/3	0.808

Table 2. Table S2: NaCl(H 2 ) with AB stacking at 30 GPa

Space group P6m2,		a = 3.507 Å, c = 3.250 Å
		x	y	z
Na	1f	2/3	1/3	1/2
Cl	1a	0	0	0
H	2h	1/3	2/3	0.615

Table 3. Table S3: NaCl(H 2 ) 4 at 50 GPa

Space group Pm,		a = 4.155 Å, b = 3.146 Å
		c = 4.197 Å, $\beta$ = 118.91∘
		x	y	z
Na	1b	0.26767	1/2	0.66311
Cl	1b	-0.06614	1/2	0.00651
H	1a	-0.08452	0	0.40644
H	1a	-0.07692	0	0.58805
H	1b	0.67073	1/2	0.45016
H	1b	0.53023	1/2	0.25898
H	1a	0.53226	0	0.02106
H	1a	0.52772	0	0.58598
H	1a	0.35007	0	0.01730
H	1a	0.35405	0	0.40858

Table 4. Table S4: Raman active modes

Mode	$\omega$calc. (cm-1)
E2g	319.7
E1g	1030.0
E2g	1053.7
A1g	4278.7

Table 5. Table S5: Infrared active modes

Mode	$\omega$calc. (cm-1)
A2u	74.8
E1u	97.9
A2u	361.1
E1u	369.5
A2u	751.2
E1u	1047.9

Table 6. Table S6: Silent modes

Mode	$\omega$calc. (cm-1)
E2u	251.0
B2g	326.6
B1u	354.9
B2g	760.2
E2u	1021.0
B1u	4293.1

Full text

Gifu U] Department of Electrical, Electronic and Computer Engineering, Gifu University, Gifu 501-1193, Japan \alsoaffiliation[UTK] Joint Institute for Advanced Materials, The University of Tennessee, TN 37996, USA

\alsoaffiliation[KYOKUGEN] Center for Science and Technology under Extreme Conditions, Graduate School of Engineering Science, Osaka University, Osaka 560-8531, Japan

Nagoya U]Department of Materials Physics, Nagoya University, Nagoya 464-8603, Japan

TITEC]Department of Earth and Planetary Sciences, Tokyo Institute of Technology, Tokyo 152-8550, Japan

JASRI] Japan Synchrotron Radiation Research Institute, Hyogo 679-5198, Japan

JASRI] Japan Synchrotron Radiation Research Institute, Hyogo 679-5198, Japan

JASRI] Japan Synchrotron Radiation Research Institute, Hyogo 679-5198, Japan

KYOKUGEN] Center for Science and Technology under Extreme Conditions, Graduate School of Engineering Science, Osaka University, Osaka 560-8531, Japan

Gifu U] Department of Electrical, Electronic and Computer Engineering, Gifu University, Gifu 501-1193, Japan

1 INTRODUCTION

Compressed hydrides have been predicted to be potential candidates for high-temperature (Tc) superconductivity.1, 2, 3 Recent discoveries of metallic SH3 and LaH10 exhibiting Tc above 200 K4, 5, 6 are attracting significant interest in novel hydrides. The synthesis of new hydrogen compounds, including unconventional stoichiometries, is of great interest.

At ambient pressure, sodium chloride (NaCl) is the only known binary compound formed by Na and Cl. The large electronegativity difference between Na and Cl stabilizes the highly ionic compound (Na+ and Cl-) which crystallizes in a rock-salt (B1) structure. Although it is not widely known, NaCl is one of the few materials7, 8, 9 that have been tested and revealed to be inert to hydrogen (H2). Even in a high-pressure H2 atmosphere reaching 100 GPa, there is no chemistry.7, 8 NaCl does not react with H2 either, when heated to 500 K at 20 GPa.8 This is why NaCl has been used to create an H2-sealing capsule in high-pressure experiments.7, 8, 10 However, we found the interstitial site in a B1 structure using the Pauling ionic radius11 is large enough to accommodate an H2 molecule whose molecular size is approximately 1.5 Å. Further, the recent experiments have revealed that alkali-metal polyhydrides (AH${}_{\textit{x}}$, A = Li and Na, x > 1), containing H2 molecules in their unit cells, can be synthesized by heating ionic compounds such as LiH or NaH in high-pressure H2.12, 13, 2 Research findings from the studies of AH${}_{\textit{x}}$ suggest that the formation of NaCl(H2)${}_{\textit{x}}$ may be anticipated by heating NaCl and H2 together at high pressures.

In this study, we performed the laser heating on the mixture of NaCl and H2 (NaCl/H2 mixture) to above 1500 K at pressures. By using Raman scattering and X-ray diffraction (XRD) measurements, we observed the formation of NaClH${}_{\textit{x}}$ with P63/mmc structure having H2 molecules in the interstitial sites of NaCl that formed ABAC stacking, when the NaCl/H2 mixture was heated at pressures exceeding 40 GPa. The synthesized NaClH${}_{\textit{x}}$ is suggested to be NaCl(H2) by our structure search based on a generic algorithm technique and first-principles calculations. It is implicated that NaCl(H2) is formed by the H2 molecules insertion into the crystal lattice of NaCl which has CsCl-type structure. Upon the decrease of pressure at 300 K, NaCl(H2) remains stable down to 17 GPa. Besides, our calculations predicted a hydrogen richer NaCl(H2)4 with a monoclinic (Pm) structure was stable at pressures above 40 GPa.

2 METHODS

2.1 Experimental Methods

To detect the formation of NaCl-hydrogen compounds and reveal their crystal structure, we performed XRD and Raman scattering measurements on the NaCl/H2 mixture heated to above 1500 K at pressures between 4 and 46 GPa. The high-pressure experiments were carried out using a diamond anvil cell (DAC) equipped with type Ia diamond anvils with culets between 0.2 and 0.3 mm. Figure 1 shows a schematic drawing of the sample chamber in a DAC. The 0.25 mm-thick gaskets made of Re (Aldrich, 99.98$\%$) were pre-indented to 50-60 $\mu$m-thickness, and a 150 $\mu$m-diameter hole (sample chamber) was drilled through the center of the gasket. We loaded two NaCl (Wako, 99.95$\%$+) thin plates, each of which has a thickness of about 10 $\mu$m, in the sample chamber and filled it with fluid H2. We used IR-lasers (SPI laser, $\lambda$ = 1070 nm and $\lambda$ = 1090 nm) to heat the NaCl/H2 mixture compressed in the DAC. Because NaCl does not absorb IR radiation, tiny chips of platinum (Pt, Nilaco, 99.95$\%$) in an experiment (Exp. 1) and gold (Au, Nilaco, 99.95$\%$) in other two experiments (Exp. 2 and 3) were supplied as laser absorbers, and the IR laser was focused on Pt or Au. Loading the H2 fluid was performed by using a cryogenic gas-loading system.15 We first compressed the NaCl/H2 mixture at room temperature to the desired pressure and then heated the mixture to above 1500 K using an IR laser. We note that we obtained the same results using different laser absorbers.

During compression at room temperature, we confirmed that NaCl and H2 did not react by observing the structural transformation via Raman scattering measurements. The heating temperature was estimated by collecting thermal radiation from the sample and analyzing it within a wavelength range of 600-800 nm to convert the radiation spectrum to a temperature in accordance with the Planck’s blackbody radiation law.16 Laser heating was performed at BL10XU/SPring-8 (SPI laser, $\lambda$ = 1070 nm) and Nagoya University (SPI laser, $\lambda$ = 1090 nm). During the laser heating at BL10XU/SPring-8, the reaction between NaCl and H2 was detected by monitoring the crystal structure change of NaCl by in situ synchrotron XRD measurements using an X-ray flat panel detector (Perkin Elmer XRD0822 CP23, 1024 $\times$ 1024 pixels, pixel size: 0.20 mm). After heating for several minutes or after we confirmed the change of crystal structure of NaCl by XRD, the sample was quenched to room temperature. After quenching, we performed Raman scattering and XRD measurements at room temperature. The XRD patterns at 300 K were collected using an imaging plate (RIGAKU R-AXIS IV++, 300 $\times$ 300 mm2, pixel size: 0.10 mm) in a forward scattering geometry. The Raman scattering measurements were performed in a backscattering geometry using a triple polychromator (JASCO NR1800) equipped with a liquid-nitrogen-cooled charge-coupled device. Radiation of 532 nm from a solid-state laser was used for excitation. The focused spot size of the radiation on the samples was approximately 10 $\mu$m in diameter. To estimate pressure, we used the shift of the first-order Raman band spectra of the diamond anvil facing the sample with a proposed calibration.17

2.2 Computational Methods

To investigate the possible formation of NaCl-hydrogen compounds, we performed a crystal structure search based on a genetic algorithm technique and first-principles calculations. We developed a crystal structure prediction code based on a genetic algorithm and combined it with the Quantum ESPRESSO.18, 19 We previously used this structure prediction code in the search for thermodynamically stable phases in a sulfur-hydrogen system20 and argon-hydrogen system.21 In our structure search, eight structures were created by ‘mating’, six by ‘distortion’, and six by ‘permutation’ in each generation. We searched for stable structures at pressures of 30, 50, and 100 GPa using supercells including two formula units for NaClH, NaClH2, NaClH3, NaClH4 and NaClH8. We used the Perdew-Burke-Ernzerhof exchange-correlation functional22 and the Vanderbilt ultrasoft pseudopotential.23 The k-space integration over the Brillouin zone was performed on an 8 $\times$ 8 $\times$ 8 grid, and the energy cut-off of the plane-wave basis was set to 80 Ry. For the phonon calculation of P63/mmc NaCl(H2), we used an 8 $\times$ 8 $\times$ 4 k-point grid and a 4 $\times$ 4 $\times$ 4 q-point grid.

3 RESULTS

3.1 Synthesized NaClH${}_{\textit{x}}$

When the NaCl/H2 mixture was heated below 30 GPa, we detected no structural transformation of NaCl. On the other hand, at 46 GPa (Exp. 1) and 42 GPa (Exp. 2 and 3) the sample exhibited visible change after laser heating (Figure 2a,b). The laser-focused area was distinguishable from other areas, suggesting that new material was synthesized (Figure 2b). The visible difference became apparent at 33 GPa (Figure 2c,d) upon a pressure decrease and revealed the presence of a compound that had a refractive index different from that of either NaCl or H2. Hereafter, we will refer the newly formed compound as NaClH${}_{\textit{x}}$.

Figure 3a and b present the synchrotron XRD profiles obtained in Exps. 1 and 2. We note that NaCl is in a cesium-chloride-type (B2, Pm$\bar{3}$m) structure at room temperature and pressures above 29 GPa.24, 25, 26 The XRD peaks of the unheated area were indexed with NaCl (B2) and PtH (P63/mmc). We did not observe the formation of gold hydride (AuH${}_{\textit{x}}$) in Exp. 2, which is in agreement with previous reports.9, 7, 27 In Exp. 1 at 46 GPa after the heating, six new XRD peaks appeared (Figure 3a) in the laser-focused area in addition to the unreacted NaCl and PtH peaks. At 42 GPa in Exp. 2, we confirmed three additional peaks at 2$\theta$ = 7.9*∘, 13.8∘, and 17.2∘* (inset to Figure 3b, upper panel). Using these nine XRD peaks, NaClH${}_{\textit{x}}$ was indexed with a hexagonal P63/mmc structure where the Cl and Na layers stack in the ABAC manner along the c axis. The lattice parameters were a = 3.410(7) Å and c = 6.277(8) Å at 46 GPa. At 20 GPa in Exp. 1 (Figure 3b, bottom panel), the presence of 010, 110 and 013 diffraction lines from the P63/mmc structure becomes evident due to the small overlap of the lines from NaClH${}_{\textit{x}}$, NaCl, and the laser absorber becomes minimal.

In Figure 3c, the volume per NaCl unit (Vf.u. (Å3/NaCl)), which is calculated by dividing the unit cell volume by the number of NaCl units in a unit cell, is plotted for NaCl+H2 and NaClH${}_{\textit{x}}$ as a function of pressure. At 42 GPa, NaClH${}_{\textit{x}}$ is 1.8% smaller than NaCl+H2 which is calculated using the XRD data of NaCl and H2 observed simultaneously with NaClH${}_{\textit{x}}$. These observations suggest that NaClH${}_{\textit{x}}$ is denser than NaCl+H2 and energetically favorable at elevated pressures. Before experiments, there was a concern that the laser heating of NaCl/H2 might result in the formation of Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7).1, 2 However, none of Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ have a P63/mmc lattice. Further, their Raman spectra do not match the data of NaClH${}_{\textit{x}}$ (See the Supporting Information for the detailed comparisons of the Raman scattering data.).

Figure 4a shows the representative Raman scattering spectra measured in the heated area where examined by XRD measurements together with the data of unheated areas after the laser heating at 46 GPa. As the heated area was the mixture of NaClH${}_{\textit{x}}$, NaCl that had no Raman-active modes, and excess solid H2, the Raman scattering peaks from NaClH${}_{\textit{x}}$ and H2 were observed. Across the heated area, in addition to the rotational mode (roton) and the intramolecular stretching mode (vibron) of H2 in excess solid H2, three new peaks appeared at the wavenumbers 272.8, 966.4, and 4323.5 cm*-1* (hereafter referred to as $\nu$1, $\nu$2, and $\nu$3, respectively). Notably, no other new peaks appeared in the wavenumbers between 100 and 4400 cm*-1*. The high wavenumber of $\nu$3 suggests that $\nu$3 is the vibron of H2. Furthermore, the shift of $\nu$3 from the vibron of solid H2 indicates that $\nu$3 originates from the H2 molecules placed in a potential field different from that of solid H2. Thus, NaClH${}_{\textit{x}}$ contains H2 molecules in its crystal lattice. Figure 4b presents the Raman scattering spectra in Exp. 1 upon pressure release at 300 K. The Raman shift of $\nu$1-3 obtained in Exp. 1-3 are plotted together as a function of pressure in Figure 4c. Peaks $\nu$1-3 have positive pressure dependence. The pressure dependences of $\nu$1, $\nu$2, and $\nu$3 obtained in the three experiments are in excellent agreement, indicating that the presence of Pt (PtH) or Au does not affect the results. When pressure is released to 17 GPa, $\nu$1-3 disappear, and only the roton and vibron of solid H2 persist (Figure 4c). Because NaCl has no Raman-active phonon mode in the B1 structure, we conclude that NaClH${}_{\textit{x}}$ decomposes into NaCl and H2 below 17 GPa.

3.2 Predicted stable phases of NaCl-hydrogen compounds: NaCl(H2) and NaCl(H2)4 above 20 GPa

Figure 5a illustrates the formation enthalpy of NaCl-hydrogen compound ((NaCl)${}_{1-\textit{x}}$H${}_{\textit{x}}$) from NaCl and H2. The convex hull diagram indicates that NaCl(H2) emerges as the thermodynamically stable phase above 15 GPa. Our calculations predict that NaCl(H2) crystallizes in a hexagonal P63/mmc structure at 30 GPa with lattice parameters a = 3.515 Å and c = 6.465 Å. The atomic positions are Na:2c(1/3, 2/3, 1/4), Cl:2a(0, 0, 0), and H:4f(1/3, 2/3, 0.808). The Cl and Na layers stack in the ABAC manner along the c axis, and H2 molecules occupy the center of mass in the Na triangles in the B (C) layer (Figure 5b). The predicted crystal symmetry agrees with that of the experimentally observed NaClH${}_{\textit{x}}$. Besides, the Vf.u. of NaCl(H2) calculated from the predicted lattice parameters shows good agreement with the experimentally observed NaClH${}_{\textit{x}}$ (Figure 3c). The volume reduction caused by the formation of NaCl(H2) is 6.1% at 20 GPa, 1.8% at 40 GPa, which is in agreement with the experiment, and 1.8% at 50 GPa.

In the crystal structure search, we also identified another hexagonal structure of NaCl(H2) whose space group is P6m2 with lattice parameters a = 3.507 Å and c = 3.250 Å. In the P6m2 structure, the atoms are at Na:1f(2/3, 1/3, 1/2), Cl:1a(0, 0, 0), and H:2h(1/3, 2/3, 0.615) positions, and the layers of Na and Cl stack in AB manner (Figure 5c). The NaCl(H2) with AB stacking possesses a higher enthalpy by 0.15 mRy/atom than that of the ABAC stacking. Besides, we predicted that a hydrogen-richer phase, NaCl(H2)4, is stabilized at pressures above 40 GPa (Figure 5a and 5d). This NaCl(H2)4 is obtained by deforming NaCl(H2) with AB stacking. An NaCl(H2) layer is formed by overlapping the B layer with A, and the hexagonal lattice is slightly distorted owing to the H2 molecular axis orientation into the layer. Then, H2 molecules are intercalated between the layers, and an (H2)3 layer is created (Figure 5d). The space group of this structure is a monoclinic Pm with lattice parameters a = 4.155 Å, b = 3.146 Å, c = 4.197 Å, $\beta$ = 118.91*∘. The detailed structural parameters for NaCl(H2)4 is shown in the Supporting Information. Although our calculations predicted the formation of these two phases (P6m*2 and Pm), we did not observe them in our experiments. We will later discuss the difference between experimental and theoretical results in relation to the criteria for the NaCl-hydrogen compound formation.

Figure 6 shows the calculated phonon dispersion of NaCl(H2) with ABAC stacking at different pressures. Imaginary phonon frequency appears in $\Gamma$-M, M-K, K-$\Gamma$, $\Gamma$-A, and A-M lines at 10 GPa (Figure 6a), which indicates that this structure is unstable at this pressure. The imaginary phonon frequency disappears above 15 GPa (Figure 6b-d). This observation is in agreement with our experiments that the synthesized NaClH${}_{\textit{x}}$ decomposes into NaCl and H2 below 17 GPa. The phonon at around 4250 cm*-1* is the H2 vibron that occupies the 4f site in NaCl(H2). This H2 vibron is slightly dispersive at 50 GPa owing to the increases of the interaction between H2 molecules and the other atoms with compression (Figure 6d). We note that our first-principles calculations predict that the H2 molecules in NaCl(H2) with ABAC stacking rotationally oscillate with limited angles (librons), and their molecular axes are roughly aligned along the c axis.

In Figure 4c, we also display the predicted Raman-active zone-center phonon modes of NaCl(H2) with ABAC stacking at pressures. See the Supporting Information for the other phonon modes that are infrared active and silent. Mode $\sharp$1 (E2g) is the translational oscillation of the B (C) layers in the a-b plane, and mode $\sharp$2 (E1g) is the H2 libron. Mode $\sharp$3 originates from the translational oscillation of H2 in the a-b plane, and mode $\sharp$4 is the H2 vibron. Here, we compare the prediction of phonon modes with experimental results. The predicted frequencies in modes $\sharp$1, $\sharp$2 and $\sharp$3, and $\sharp$4 emerge at frequencies similar to $\nu$1, $\nu$2, and $\nu$3, respectively. In the Raman spectra, $\nu$2 has a wider peak width than $\nu$1, suggesting the presence of multiple peaks in a broad single peak. Assuming that modes $\sharp$2 and $\sharp$3 are broadened and overlap, $\nu$2 can be assigned to the overlapped $\sharp$2 and $\sharp$3 modes. In addition, the predicted positive pressure dependence of modes $\sharp$1-3 agrees with that of $\nu$1-3 (Figure 4c). We also note that the pressure dependence of the observed H2 vibron shows an excellent agreement with the reported data of pure H230, which supports our conclusion that the observed Raman peaks except for $\nu$1-3 are from solid H2.

4 DISCUSSION

Considering all experimental and theoretical data together, we conclude that hexagonal NaCl(H2) with ABAC stacking is synthesized by heating the NaCl/H2 mixture to approximately 1500 K above 40 GPa. While a pressure near 40 GPa is necessary for synthesis, NaCl(H2) is stable down to 17 GPa.

The questions remained are; (i) What kinds of the chemical bonding are formed between H2 and NaCl host lattice?, (ii) How the compound is formed at high pressure and high temperatures?, (iii) What is the criteria for NaCl-hydrogen compound formation?, and (iv) Why the salt is incompatible with hydrogen at room temperature?

On the question (i), the Raman scattering measurements show that the H2-vibron frequency of NaCl(H2) continues to increase beyond that of pure solid hydrogen with the increase of pressure, which implies that there is little interaction between NaCl host lattice and H2 molecules. The results of the phonon calculations also support this hypothesis, in which the H2-vibron frequency is non-dispersive up to at least 40 GPa. That is to say, NaCl host lattice acts as a compressor for the H2 molecule up to at least 46 GPa, similarly to the case of argon-hydrogen compounds reported earlier.21

On the question (ii), we speculate that the NaCl(H2) is formed by the insertion of H2 molecules into the B2-type NaCl host lattice. The temperature used in the present study was about 1500 K which is below the melting point of NaCl at 42-46 GPa31 and well above the melting point of H2.32 Therefore, NaCl(H2) is a product of the reaction between solid B2-type NaCl and fluid H2. The P63/mmc structure for NaCl(H2) can be obtained by the small distortion of the B2-type structure and the displacement of the body-centered Na atom.

To obtain additional insight relating to the question (iii), we laser-heated (2000 K) cesium chloride (CsCl) at 7 GPa and potassium chloride (KCl) at 17 GPa. Both of these compounds have a B2 structure. KCl at 17 GPa has a larger interstitial site than NaCl at 46 GPa. In contrast, CsCl has inadequate interstitial site volume to accommodate an H2 molecule. KCl appears to have the potential to form a KCl(H2)${}_{\textit{x}}$ while CsCl does not. However, no structural transformations were observed for both KCl and CsCl. Although further studies are necessary, we think that many factors such as crystal structure, the ionic radii of elements that form salts, and pressure and temperatures dictate the formation criteria of salt-hydrogen compounds.

Answering the question (iv) is still difficult at current. The laser heating at 25 GPa, where NaCl is in B2-type structure at temperatures above 1100 K,26 did not result in the formation of the NaClH${}_{\textit{x}}$, while our theoretical calculations predicted NaCl(H2) becomes stable above 15 GPa. Along with the fact that the formation of NaClH${}_{\textit{x}}$ is not induced by the compression of the NaCl/H2 mixture at room temperature, it is thought that there is a large energy barrier between NaCl+H2 and NaClH${}_{\textit{x}}$. Therefore, sufficiently high temperature is necessary to overcome the barrier depending on pressure. It would be worth trying the synthesis of NaCl(H2) with AB stacking and NaCl(H2)4, which are predicted by first-principles calculations, at much higher pressures and temperatures. The information of the present study along with the studies of other ionic compounds13, 2 will contribute to a deeper understanding of the inertness against H2.

5 CONCLUSION

The present study showed that NaCl can indeed form a compound with hydrogen, NaClH${}_{\textit{x}}$. The NaClH${}_{\textit{x}}$ is suggested to be NaCl(H2) which has a P63/mmc structure accommodating two H2 molecules in the interstitial sites of the unit cell. Upon the decrease of pressure at 300 K, NaCl(H2) remains stable at pressures down to 17 GPa. Such a large hysteresis in pressure at room temperature suggests the possibility of recovering NaCl(H2) to ambient pressure at low temperatures when the mobility of hydrogen is significantly suppressed. Our calculations predict that hydrogen-richer phase NaCl(H2)4 is stabilized at pressures above 40 GPa. A further question remains whether the synthesized NaClH${}_{\textit{x}}$ phase has the stoichiometric composition NaCl(H2). In-situ neutron diffraction measurements, which include the recovery of materials at low temperature, are expected to provide an answer to the question.

6 Supporting Information - Hydrogen-Storing Salt NaCl(H2) Synthesized at High Pressure and High Temperature

Contents

1. 1 INTRODUCTION

2. 2 METHODS

3. 2.1 Experimental Methods

4. 2.2 Computational Methods

5. 3 RESULTS

6. 3.1 Synthesized NaClH${}_{\textit{x}}$

7. 3.2 Predicted stable phases of NaCl-hydrogen compounds: NaCl(H2) and NaCl(H2)4 above 20 GPa

8. 4 DISCUSSION

9. 5 CONCLUSION

10. 6 Supporting Information - Hydrogen-Storing Salt NaCl(H2) Synthesized at High Pressure and High Temperature

11. S.1 Structural parameters for NaCl(H2) and NaCl(H2)4 -results of ab-initio calculations-

12. S.2 Comparison of Raman scattering data between NaCl(H${}_{\textit{x}}$), Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7)

13. S.3 Calculated zone-center phonon frequencies of NaCl(H2) with ABAC-stacking at 46 GPa

• •

Structural parameters for NaCl(H2) and NaCl(H2)4 -results of ab-initio calculations-

• •

Comparison of Raman scattering data between NaCl(H${}_{\textit{x}}$), Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7)

• •

Calculated zone-center phonon frequencies of NaCl(H2) with ABAC-stacking at 46 GPa

S.1 Structural parameters for NaCl(H2) and NaCl(H2)4 -results of ab-initio calculations-

S.2 Comparison of Raman scattering data between NaCl(H${}_{\textit{x}}$), Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7)

Here, we compare the Raman scattering spectra obtained for NaCl(H${}_{\textit{x}}$) with the data of Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7). The Raman scattering spectrum and the pressure dependence of the observed Raman scattering peaks of NaCl(H${}_{\textit{x}}$) do not match the data for any of Na3Cl, NaCl3, and NaH${}_{\textit{x}}$ (x = 3, 7).

S.3 Calculated zone-center phonon frequencies of NaCl(H2) with ABAC-stacking at 46 GPa