Directed Dehydration as Synthetic Tool for Generation of a New Na₄SnS₄ Polymorph: Crystal Structure, Na⁺ Conductivity, and Influence of Sb‐Substitution

Abstract

We present the convenient synthesis and characterization of the new ternary thiostannate Na₄SnS₄ (space group $I{4}_1/acd$ ) by directed removal of crystal water molecules from Na₄SnS₄⋅14 H₂O. The compound represents a new kinetically stable polymorph of Na₄SnS₄, which is transformed into the known, thermodynamically stable form (space group $P\bar{4}{2}_{1}c$ ) at elevated temperatures. Thermal co‐decomposition of mixtures with Na₃SbS₄⋅9 H₂O generates solid solution products Na_4−x Sn_1−x Sb _x S₄ (x=0.01, 0.10) isostructural to the new polymorph (x=0). Incorporation of Sb⁵⁺ affects the bonding and local structural situation noticeably evidenced by X‐ray diffraction, ¹¹⁹Sn and ²³Na NMR, and ¹¹⁹Sn Mössbauer spectroscopy. Electrochemical impedance spectroscopy demonstrates an enormous improvement of the ionic conductivity with increasing Sb content for the solid solution (σ _25°C=2×10⁻³, 2×10⁻², and 0.1 mS cm⁻¹ for x=0, 0.01, and 0.10), being several orders of magnitude higher than for the known Na₄SnS₄ polymorph.

Dehydration of Na₄SnS₄⋅14 H₂O at moderate temperature generates the new, kinetically stabilized Na₄SnS₄ polymorph exhibiting enhanced ionic conductivities compared to the thermodynamically stabilized form. Co‐decomposition with Na₃SbS₄⋅9 H₂O enables an easy, fast, and scalable avenue to obtain the solid solution Na_4−x Sn_1−x Sb _x S₄ allowing methodical investigations of Sb‐substitution effects in these solid‐state electrolytes.

Introduction

The structures of thiostannate compounds include different anionic units like [SnS₄]⁴⁻, [Sn₂S₆]⁴⁻, [Sn₂S₅]²⁻, [Sn₂S₇]⁶⁻, [Sn₂S₈]²⁻, [Sn₃S₇]²⁻, [Sn₃S₈]⁴⁻, [Sn₄S₁₀]⁴⁻ or [Sn₅S₁₂]⁴⁻, which are charge compensated by metal cations.[ ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ ] The structural versatility of thiostannate anions is due to the variable coordination number as well as the variable oxidation state of Sn. In the above‐mentioned thiostannate anions Sn exhibits the coordination number 4 (tetrahedron), 5 (trigonal bipyramid), or 6 (octahedron). In addition, Sn occurs mainly in the oxidation state IV, but compounds with different oxidation states like II/IV or III/IV were also reported.[ ¹⁰ , ¹¹ , ¹² ] Thiostannates with metal cations as counter ions were mainly synthesized using the molten‐flux approach,[ ¹³ , ¹⁴ ] or via high‐temperature syntheses.[ ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ ] Thiostannates with protonated amine molecules, transition metal cations, or transition metal complexes as counter ions compensating the negative charge of the [Sn _x S _y ] ⁿ⁻ anions were mostly prepared under solvothermal conditions.[ ⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ ] Applying suitable precursors, the preparation of thiostannates at room temperature was also reported.[ ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ ] These examples demonstrate the broad variety of synthetic approaches established in the thiostannate chemistry.

Motivated by the intense research in the field of all‐solid‐state batteries (ASSBs), where sulfidic compounds came into the focus including thiophosphates,[ ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ ] or thioantimonate based solid electrolytes,[ ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ , ⁶² ] we searched for possible solid‐state electrolytes (SSEs) on the basis of chalcogeniodostannates. Indeed, first steps in this direction were reported for Li⁺ SSEs: Li⁺ ion conducting thio‐ and selenostannates were prepared by a high‐temperature route combined with extraction or thermal evaporation (T=320 °C) methods of intermediately synthesized solvates.[ ⁶³ , ⁶⁴ ] More recently, mixed quaternary Na⁺ superionic conductors based on the thiostannate chemistry were discovered, e.g., the series of compounds Na_4−x Sn_1−x Sb _x S₄[ ⁵⁰ , ⁵⁶ ] and Na₁₁Sn₂PS₁₂[ ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ ] show promising room temperature Na⁺ ion conductivities >0.5 mS cm⁻¹. Moreover, excellent cycling and rate stabilities were observed using Na_3.75Sn_0.75Sb_0.25S₄ or Na_10.8Sn_1.9PS_11.8 as SSEs in ASSB full cells.[ ⁴⁷ , ⁵⁶ ] In addition, such thiostannate based SSEs exhibit broad electrochemical windows, e.g., no redox reactions occur for Na₁₁Sn₂PS₁₂ vs. Na⁺/Na between 0.3–3.2 V.[ ⁴⁷ , ⁶⁵ ] The preparation of such SSEs is mainly performed applying classic high‐temperature solid‐state methods or sintering processes of appropriate precursors at high temperatures. The directed removal of crystal water molecules from hydrated solids is an elegant alternative synthetic approach. This was also applied for Na₃SbS₄⋅9 H₂O yielding dehydrated Na₃SbS₄ ^[52] with remarkable high ionic conductivity and an excellent electrochemical stability between 0–5.0 V vs. Na⁺/Na. ^[52] However, hydrated thiostannates comprising transition metal complexes or organic cations are often irreversibly destroyed during removal of water molecules, and thus suitable “pure” inorganic crystal water containing compounds have to be identified as possible candidates for Na⁺ ion conductors like already demonstrated for the lighter homologue Li. A prominent example is Na₄SnS₄⋅14 H₂O which was reported already decades ago. Here, we report on the thermal properties of this compound, the crystal structure of a new metastable polymorph of Na₄SnS₄, the co‐decomposition of Na₄SnS₄⋅14 H₂O and Na₃SbS₄⋅9 H₂O, the ionic conduction properties, ¹¹⁹Sn Mössbauer, and ²³Na as well as ¹¹⁹Sn magic‐angle spinning (MAS) NMR characterization of the materials.

Results and Discussion

The thermal properties of Na₄SnS₄⋅14 H₂O (I) were investigated using simultaneously thermogravimetry (TG) and differential thermoanalysis (DTA). Upon heating I to 300 °C, two steps are observed in the TG curve accompanied by intense and broad endothermic events in the DTA curve (T _p=66 and 142 °C) (Figure 1a). The mass loss of 42.4 % at 170 °C matches well with that calculated for the removal of all H₂O molecules (42.6 %), which is confirmed by ¹H MAS NMR (Figure S1). The X‐ray powder diffraction (XRPD) pattern of the product obtained at 300 °C (Figure S2) is dominated by reflections of a new compound with composition Na₄SnS₄ (II) with small amounts (≈4 wt %) of literature known Na₄SnS₄ (III), which was prepared at elevated temperatures. ^[8] A second thermal experiment was stopped at 170 °C and the XRPD pattern reveals only reflections of II (Figure 1b). A suitable structural model for II was refined via Rietveld method (Figure 1c, details in Supporting Information). The preparation of phase‐pure II is scalable, as confirmed by further syntheses using an inert gas furnace.

Figure 1.

a) TG, DTA and derivative TG (DTG) curves of I (heating rate: 4 K min⁻¹, N₂ atmosphere); b) XRPD pattern of a TG product of I recovered at 170 °C (4 K min⁻¹) compared to simulated (sim.) patterns of the known Na₄SnS₄ polymorph (III) and of I; c) final Rietveld plot revealing the new polymorph Na₄SnS₄ (II), the inset shows a magnification at high angles.

We hypothesize that II formed by heating of I with a rate of 4 K min⁻¹ up to 170 °C is kinetically stabilized. Hence, I was heated to 300 °C with a lower heating rate of 1 K min⁻¹ (Figure 2a) exhibiting significant differences in the TG curve compared to that recorded with 4 K min⁻¹, showing at least three not well resolved reaction steps accompanied by two broad endothermic events in the DTA curve (T _p=62 and 92 °C). The weight loss (41.4 %) is already finished at ≈105 °C in reasonably good agreement with the emission of all crystal water molecules. A second TG run was performed to 140 °C and independent from the final temperature (140 or 300 °C), the XRPD patterns show reflections of II and of III (Figure 2b). Quantitative analyses (Figure S3) reveal that the amount of III increases during heating from ≈54 wt % at 140 °C to ≈67 wt % at 300 °C, thus the kinetically stabilized polymorph II transforms into III.

Figure 2.

a) TG, DTA and DTG curves of I (heating rate: 1 K min⁻¹, N₂ atmosphere); b) XRPD patterns of TG products of I recovered at 140 and 300 °C (1 K min⁻¹) compared to simulated (sim.) patterns of the known (III) and the new (II) polymorph of Na₄SnS₄.

Further TG experiments were performed up to 620 °C and 700 °C (heating rate: 1 K min⁻¹) showing a broad exothermic event at T _p=563 °C (Figure S4). The XRPD patterns (Figure S5) are now dominated by reflections of III (≈95 wt % in TG product recovered at 700 °C, Figure S6). These investigations evidence that III is the thermodynamically stable polymorph over the entire temperature range.

The new Na₄SnS₄ polymorph (II) crystallizes in the tetragonal space group (SG) $I{4}_1/acd$ (no. 142) with a=13.8413(2), c=27.4232(4) Å, V=5253.80(12) ų and Z=24 (Table S1 and S2). ^[66] The overall structural features are similar to those reported for Na₁₁Sn₂PS₁₂,[ ⁴⁸ , ⁴⁹ , ⁵⁰ ] and Na₁₁Sn₂SbS₁₂.[ ⁵⁰ , ⁵⁶ ] The structure of II (Figure 3) comprises two unique Sn atoms located on special positions, three crystallographically independent S atoms on general positions, four Na atoms on special positions, and one Na atom on a general position (Table S2). A sixth “interstitial” Na site in loose coordination to eight S atoms (“Na6 void” in Figure S7), which is partially occupied in Na₁₁Sn₂MS₁₂ (M=Sb and P),[ ⁴⁷ , ⁴⁸ , ⁵⁰ , ⁶⁷ ] is not detected for II. One has to keep in mind that reliable information about low‐level occupation of light elements can hardly be extracted from in‐house XRPD data. Both Sn⁴⁺ cations are tetrahedrally coordinated by S²⁻ anions and II can be described as Na₁₂[SnS₄]₂{SnS₄} with Sn1 located in [SnS₄] (Table S6, dark blue in Figure 3) and Sn2 in {SnS₄} (light blue in Figure 3), respectively. According to literature, P and Sb are located on a special position in Na₁₁Sn₂PS₁₂ and Na₁₁Sn₂SbS₁₂ and substitution of Sn⁴⁺ by >1/3 P⁵⁺ or Sb⁵⁺ leads to formation of additional phases (e.g. Na₃PS₄ or Na₃SbS₄)[ ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵⁶ ] or results in another crystal structure (e.g. Na₁₀SnP₂S₁₂). ^[68] Using a general notation Na_12−3y−3x [SnS₄]₂{Sn_1−3y−3x P_3y Sb_3x S₄}, corresponding to the formula Na_4−y−x Sn_1−y−x P _y Sb _x S₄, II can be described as the new pure thiostannate (x=0, y=0) of this structural family.

Figure 3.

Structure of II: Unit cell a) along [001], b) along [100] (note: Na−S bonds are hidden), c) Na⋅⋅⋅Na contacts <3.6 Å.

Na−S distances are observed at 2.7–3.3 Å (Table S6) and a distance gap occurs between 3.3 and 3.7 Å (Figure S10) for II, indicating that ionic bonding interactions are present <3.3 Å and very weak interactions in an enlarged environment. Taking the sum of ionic radii as a limit for ionic bonds, the Na⁺ cations are in three different coordination environments (Table S6): Na1 is in a distorted trigonal‐bipyramidal S²⁻ coordination (Na−S between 2.849–3.189 Å), and a sixth S²⁻ anion is located at Na1⋅⋅⋅S1: 3.810 Å. Na2 is tetrahedrally surrounded by S²⁻ (Na−S: 2.793–2.924 Å) with additional weak interactions to two S²⁻ (Na2⋅⋅⋅S2: 3.696 Å). Na3, Na4, and Na5 are in a highly distorted octahedral geometry of S²⁻ (Na−S: 2.719–3.270 Å). Considering also weak interactions <4 Å, all Na⁺ cations are located in highly distorted NaS₆ octahedra. Noticeably, the longest Na⋅⋅⋅S distance in these NaS₆ polyhedra is distinctly larger for II (Na1−S1: 3.810 Å), than those reported for Na₁₁Sn₂MS₁₂ (M=Sb and P, Na⋅⋅⋅S ≤3.572 Å and ≤3.361 Å, respectively). ^[50]

A three‐dimensional network is generated by edge‐ and corner‐sharing of NaS _x polyhedra, and a detailed description of interconnections is given in the Supporting Information (see Figure S7). This network exhibits nearly linear and zig‐zag Na⁺ cation chains along the c‐axis and in the ab‐plane (Figure 3a and b) with interatomic Na⋅⋅⋅Na distances of 3.297–3.593 Å (Figure 3c, Table S6). Each Na⁺ is in almost square planar environment of four Na⁺, leading to a distorted NbO‐topology substructure prominent for, e.g., Na₁₁Sn₂PS₁₂.[ ⁴⁸ , ⁶⁷ ]

The literature known Na₄SnS₄ polymorph (III) crystallizes in the tetragonal SG $P\bar{4}{2}_{1}c$ (no. 114) with a=7.84679(3) and c=6.95798(4) Å, V=428.42(1) ų and Z=2 (Tables S1, S5), close to values already reported.[ ⁸ , ⁶⁹ ] In this structure (Figure 4, 2×2×4 unit cells for comparability with II) all Sn⁴⁺ cations are located on one special position and are tetrahedrally coordinated by S²⁻ (Sn−S: 2.380 Å). The Na⁺ cations are located on one general position and are surrounded by five S²⁻ (Na−S: 2.809–3.113 Å, Table S7). These NaS₅ units are linked by sharing common edges and joined to the SnS₄ tetrahedra by edge‐sharing. In contrast to II, several Na⁺ cations are located in the voids between SnS₄ units. The Na⋅⋅⋅Na contacts in III (Figure 4c) are similar in the ab‐plane (3.425–3.534 Å) to those in II (3.463–3.525 Å), but significantly longer along the c‐axis (dotted lines in Figure 4c) (III: 3.669 Å vs. II: 3.293 and 3.595 Å, see Figure S11).

Figure 4.

Structure of III: 2×2×4 unit cells a) along [001], b) along [100] (note: Na−S bonds are hidden), c) Na⋅⋅⋅Na connections (dotted lines represent connections >3.6 Å).

Applying the same fast and moderate temperature approach as for II (x=0), samples containing Sb (IIa, x=0.01, IIb x=0.10) were prepared by co‐decomposition of stoichiometric amounts of Na₄SnS₄⋅14 H₂O and Na₃SbS₄⋅9 H₂O at 170 °C (details in Supporting Information) to obtain Na_4−x Sn_1−x Sb _x S₄ with low Sb contents.

Rietveld refinements (Figure S8) reveal that this easy, scalable, and cost‐effective synthetic approach yields phase‐pure products IIa and IIb crystallizing isostructurally to II in $I{4}_1/acd$ with lattice parameters a=13.8399(3) and 13.8284(3) Å, c=27.4273(8) and 27.4679(8) Å, V=5253.51(22) and 5252.55(24) ų, respectively (Tables S1, S3 and S4), ^[66] in good agreement with values reported for, e.g., Na_3.95Sn_0.95Sb_0.05S₄ (a=13.8362 Å, c=27.4499 Å, V=5255 ų). ^[56] The lattice parameters for IIa are between those for II and IIb (Figure S9), evidencing that even the substitution of only x=0.01 was successful. Some structural features should be discussed for this solid solution with increasing Sb content: i) expansion of the unit cell along the c‐axis and contraction in the ab‐plane (Figure S9); ii) a narrower gap between near and far Na−S distances assuming highly distorted NaS₆ octahedra with the longest distance in all NaS₆ polyhedra to be 3.810, 3.754, and 3.666 Å for x=0, 0.01, and 0.10 (Figure S10 and Table S6); iii) a more uniform distribution of distances along the c‐axis in the NbO‐like Na⁺ cation sublattice (longest‐shortest Na⋅⋅⋅Na separation: 0.30, 0.23, and only 0.04 Å for x=0, 0.01, and 0.10, Figure S11, Table S6); iv) only small changes (<0.03 Å) of Na⋅⋅⋅Na separations in the ab‐plane (Figure S11, Table S6).

The local environments of Sn⁴⁺ and Na⁺ cations in II, IIa, and IIb were further investigated by solid‐state ¹¹⁹Sn and ²³Na MAS NMR (Figure 5a and b, also Figure S12) and ¹¹⁹Sn Mössbauer spectroscopy (Figure 5c). Parameters derived from these experiments are given in Table 1. The ¹¹⁹Sn MAS NMR spectrum (Figure 5a) for II exhibits two peaks at 84.7(1) (p1) and 75.6(1) ppm (p2) with an intensity ratio of A_p1:A_p2≈2 : 1. Hence, p1 is assigned to Sn⁴⁺ cations on the crystallographic Sn1 site and p2 to those located on the Sn2 site. The chemical shifts are comparable to values reported for SnS₄ tetrahedra in the solid polymorph III ^[70] and for [SnS₄]⁴⁻ units in the liquid‐state.[ ⁷¹ , ⁷² ] With increasing Sb content (IIa and IIb), a shift of p1 and p2 to lower values, broader signals, and a higher degree of peak asymmetry is observed. Additionally, a third peak (p1′) close to p1 appears. These observations point towards an increased shielding around Sn⁴⁺ cations and local distortion of the polyhedra caused by Sb⁵⁺ cations in the nearby environment. An intensity ratio of A_p1+p1′: A_p2≈2 : 1 is determined for all samples, leading to the conclusion, that p1′ represents Sn⁴⁺ cations on the Sn1 site which are in close proximity to Sb⁵⁺ cations, thus being in a different local environment than Sn⁴⁺ cations close to other Sn⁴⁺ cations only. In all ²³Na MAS NMR spectra (Figure 5b), two peaks at ≈12.8 (p3) and ≈6.9 ppm (p4) are detected. No ²³Na MAS NMR data is reported for Na₁₁Sn₂SbS₁₂, but a single and sharp ²³Na MAS NMR signal is observed for Na₁₁Sn₂PS₁₂ ^[48] and Na_11.1Sn_2.1P_0.9Se₁₂, ^[43] despite six crystallographically distinct Na sites, resulting from fast ionic motion in the NMR time frame. We assume a comparable situation for Na_4−x Sn_1−x Sb _x S₄. A distinctly lower peak intensity of p4 in IIa and IIb compared to II is a good hint that more Na⁺ cations participate in the ion motion if Sn⁴⁺ is substituted by Sb⁵⁺.

Figure 5.

a) ¹¹⁹Sn MAS NMR, b) ²³Na MAS NMR, and c) ¹¹⁹Sn Mössbauer spectra for II, IIa, and IIb (Na_4−x Sn_1−x Sb _x S₄, x=0, 0.01, 0.10).

Table 1.

Results of ¹¹⁹Sn and ²³Na MAS NMR and ¹¹⁹Sn Mössbauer spectroscopy for II, IIa, and IIb (Na_4−x Sn_1−x Sb _x S₄, x=0, 0.01, 0.10): chemical shift δ, isomer shift IS, and line width Γ.

	119Sn MAS NMR			23Na MAS NMR		119Sn Mössbauer
	δ [ppm]			δ [ppm]		IS [mm s−1]	Γ [mm s−1]
peaks	p1	p1′	p2	p3	p4
II	84.7(1)	–	75.6(1)	12.8(1)	6.9(1)	1.17(1)	0.87(1)
IIa	84.6(1)	81.4(1)	75.5(1)	12.8(1)	6.9(1)	1.14(1)	0.88(1)
IIb	84.3(1)	81 to 83	74.5(1)	12.7(1)	≈6.9	1.16(1)	0.93(1)

All ¹¹⁹Sn Mössbauer spectra (Figure 5c) can be well fitted with a singlet with isomer shifts in the range 1.15(1) mm s⁻¹ and a full width at half maximum of about 0.9 mm s⁻¹. Similar values have been reported for III ^[6] and Li₁₀SnP₂S₁₂. ^[73] This peak is thus assigned to Sn⁴⁺ cations in the tetrahedral S²⁻ environments. The different Sn environments in the crystal structure are too similar to be discriminated by Mössbauer spectroscopy. Nonetheless, the spectra show a broadening of the line for the sample with the highest Sb content (IIb) indicating a local structural distortion in good agreement with the ¹¹⁹Sn MAS NMR results.

The structural features observed for II, IIa, and IIb promise a potential application as SSEs in ASSBs. Hence, the Na⁺ cation conductivities were determined by electrochemical impedance spectroscopy (EIS) between −10 to +70 °C. Representative results are shown in Figure 6. In the Nyquist plots recorded at 0 °C (Figure 6a) for II, IIa, and IIb, single depressed semicircles and sloping lines at small frequencies are observed. The deconvolution of contributions from bulk and grain boundaries is not possible and all AC impedance spectra were fitted with an equivalent circuit as shown in Figure 6b (details in Supporting Information). The Na⁺ cation conductivities σ and capacitances C at different temperatures are shown in Table 2. Values for C correspond to averaged bulk and grain boundary transport behavior if compared to empirical characteristics reported in ref. [74] (C _bulk≈0.001 nF, C _{grain boundary}≈4 nF). The EIS experiments demonstrate good (II) to superior (IIb) Na⁺ cation motion in the whole temperature regime with σ covering several orders of magnitude. The substitution of Sn⁴⁺ by Sb⁵⁺ has a highly beneficial impact on the Na⁺ cation motion and σ of II is already enhanced by one order of magnitude for x=0.01 (e.g. σ(II)_30°C=3.0 μS cm⁻¹, σ(IIa)_30°C=26 μS cm⁻¹) despite it is isostructural. For x=0.10, σ is >40 times larger than for II (σ(IIb)_30°C=0.1 mS cm⁻¹) enabling superionic conduction properties comparable to findings for high‐temperature products of Na_4−x Sn_1−x Sb _x S₄ (e.g. σ _30°C≈0.1 mS cm⁻¹ for Na_3.95Sn_0.95Sb_0.05S₄). ^[56] In contrast, the AC resistance for III is too high for a reasonable determination of σ at T<25 °C. Even at T≥25 °C, only approximate values for σ can be calculated (inset in Figure 6b), similar to those reported for III. ^[69] σ is at least two orders of magnitude lower in III (σ(III)_30°C≈0.02 μS cm⁻¹) compared to II.

Figure 6.

Nyquist plots recorded at a) 0 °C and b) 30 °C; c) Arrhenius plots for II, IIa, IIb (Na_4−x Sn_1−x Sb _x S₄, x=0, 0.01, 0.10), and III (Na₄SnS₄). An equivalent circuit for these AC impedance spectra is shown in (b).

Table 2.

Ionic conductivities σ and capacitances C at various temperatures for II, IIa, IIb (Na_4−x Sn_1−x Sb _x S₄, x=0, 0.01, 0.10), and III (Na₄SnS₄).

T [°C]	σ [μS cm−1]				C [nF]
	II	IIa	IIb	III [a]	II	IIa [b]	IIb [b]	III [a]
70	24.4(6)	124(19)	626(14)	0.25(3)	0.05(1)	–	–	0.06(1)
60	15.4(2)	90.7(77)	443(3)	0.15(2)	0.05(1)	–	–	0.06(1)
50	9.33(4)	62.1(30)	297(2)	0.08(1)	0.06(1)	–	–	0.06(1)
40	5.45(3)	40.8(14)	196(5)	0.04(1)	0.06(1)	–	–	0.06(1)
30	3.02(3)	25.5(6)	123(3)	0.02(1)	0.06(1)	0.08(1)	–	0.06(1)
25	2.21(3)	19.9(4)	95.3(11)	0.01(1)	0.06(1)	0.08(1)	–	0.07(1)
20	1.60(3)	15.3(2)	73.6(6)	–	0.07(1)	0.08(1)	0.14(1)	–
10	0.79(2)	8.76(1)	42.1(2)	–	0.07(1)	0.09(1)	0.14(2)	–
0	0.37(1)	4.72(1)	23.2(1)	–	0.07(1)	0.09(1)	0.16(1)	–
−10	0.17(1)	2.37(1)	12.1(2)	–	0.07(1)	0.10(1)	0.19(1)	–

[a] EIS spectra <25 °C are too noisy for reliable calculations. [b] C(IIa) and C(IIb) cannot be extracted at ≥40 and ≥25 °C, respectively (indistinct semicircles).

The temperature‐dependent σ follow the Arrhenius law for all samples (Figure 6c). The activation energy E _a is 0.65 eV for III comparable to literature data (0.58 eV) ^[69] and larger than for II (0.51 eV). Substitution of Sn⁴⁺ by Sb⁵⁺ reduces E _a to 0.41 eV for IIa and IIb close to data reported for superionic conductors, e.g., Na₁₁Sn₂PS₁₂ (0.39 eV, ^[48] 0.25 eV ^[49] ), Na₁₀SnP₂S₁₂ (0.36 eV ^[68] ), Na₁₁Sn₂SbS₁₂ (0.34 eV ^[50] ), Na₃PS₄ (0.38 eV, 0.35 eV), ^[45] and Na₃SbS₄ (0.25 eV, ^[51] 0.22 eV ^[52] ).

The results from EIS highlight the pronounced impact of structural differences, in particular the different three‐dimensional Na⁺ cation topology on the ion motion properties comparing II and III. Moreover, the significantly enhanced Na⁺ cation mobility with increasing Sb content in Na_4−x Sn_1−x Sb _x S₄ needs to be discussed. One reason is certainly the lower total Na occupation inevitably coming along with the substitution, hence statistically more vacant sites are present for Na⁺ ion hopping. Another reason is the more regular distribution of Na⋅⋅⋅Na connections with increasing Sb content (Figure S11), being beneficial for the ionic conduction along the c‐axis. One bottleneck debated for the ionic conduction in Na_4−y−x Sn_1−y−x P _y Sb _x S₄ is the Na⁺ cation hopping through a Na6 void (see Figure S7) limited by the shortest S⋅⋅⋅S distances between shared faces of Na6S₈ and Na1S₆ or Na2S₆ (Figure S13). ^[50] The S⋅⋅⋅S barrier for the Na1→Na6→Na1 pathway is almost the same (±0.01 Å) and increases only slightly (+0.06 Å) for the Na2→Na6→Na2 pathway if Sn⁴⁺ is replaced by Sb⁵⁺ (Table S8). These small changes in contradiction to the tremendously different ionic conductivity in this series confirm the previous observations that these pathways are less important for the ionic conduction in this structural family. ^[50]

Additionally, the paddle‐wheel mechanism is long‐debated for superionic conductors, e.g., explaining an enhanced Na⁺ cation conduction property for Na₁₁Sn₂PS₁₂ vs. Na₁₁Sn₂SbS₁₂ by less hindered PS₄ rotational dynamics compared to those of SbS₄ units. ^[75] The results of NMR and Mössbauer experiments indicate larger SnS₄ distortions and improved Na⁺ cation mobility with increasing Sb content, in agreement with the EIS results. However, calculations and modeling are now necessary to understand the SnS₄ dynamics, its relation to SbS₄ incorporation and Na⁺ mobility.

To this end, we performed quantum calculations to elucidate the role of [SnS₄]⁴⁻ and [SbS₄]³⁻ tetrahedrons for coordinating interstitial Na⁺ ions. As a proxy model featuring Na⁺ coordination by the edges of two MS₄ motifs, we characterized Na⁺ ion incorporation between two charge neutral Na₄[SnS₄] species. This was compared to Na⁺ incorporation between adjacent Na₄[SnS₄] and Na₃[SbS₄] tetrahedrons. Strikingly, we find latter scenario (i.e. the [Na₄[SnS₄]‐Na‐[SbS₄]Na₃]⁺ hetero‐pair) to be preferred by 0.3 eV, hence suggesting significant lowering of the interstitial formation energy. This may be rationalized by the larger formal charge of the Sb⁵⁺ species that gives rise to the more compact SbS₄ tetrahedron as compared to SnS₄—with the Sb−S and Sn−S distances being 2.53 and 2.44 Å, respectively (see Supporting Information for details of the quantum calculations). Thus, SbS₄ incorporation in Na_4−x Sn_1−x Sb _x S₄ offers additional space to the neighboring SnS₄ tetrahedrons—which is suggested to not only facilitate rotation in support of Na⁺ hopping, but also helps Frenkel defect formation (producing Na vacancies as the suggested mobile species).

Conclusion

Here we demonstrated that the new, kinetically stabilized Na₄SnS₄ polymorph is obtained via directed removal of crystal water from Na₄SnS₄⋅14 H₂O at 170 °C. This process at relatively moderate temperature is very easy, cost‐effective, and scalable, and also provides an economic preparation method for the solid solution Na_4−x Sn_1−x Sb _x S₄ by co‐decomposition with Na₃SbS₄⋅9 H₂O. In contrast to recent claims in literature, we here report that this solid solution exhibits no lower Sb limit using the presented syntheses, but slow heating to higher temperatures inevitable causes structural transformation of the new Na₄SnS₄ polymorph into the known, thermodynamically stabilized form. These findings enabled a systematic study at the lower Sb substitution boundary for the first time. Combined XRPD, ²³Na and ¹¹⁹Sn NMR, ¹¹⁹Sn Mössbauer, and EIS experiments shed light on the local and crystallographic structure for x=0, 0.01, and 0.10 and their relation to the Na⁺ cation mobility. Compared to the known Na₄SnS₄ polymorph, the room temperature Na⁺ conductivity of the new form is considerably improved by two orders of magnitude to σ _25°C=2×10⁻³ mS cm⁻¹. Incorporation of minor Sb⁵⁺ amounts (x=0.01) in the new polymorph already enhances the Na⁺ conductivity significantly and further substitution to x=0.10 enables room temperature superionic conduction properties (σ _25°C =0.1 mS cm⁻¹, E _a=0.41 eV). The synthetic approach presented here opens new avenues to syntheses of SSE by thermal co‐decomposition using other precursor mixtures to access compounds of this thiostannate family with methodical low‐level cation and anion doping, e.g., by incorporation of P⁵⁺, Sb⁵⁺, Cl⁻, and Br⁻ in Na₄SnS₄, ^[76] which was difficult until now because of the temperature‐dependent polymorphism.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

F. Hartmann, A. Benkada, S. Indris, M. Poschmann, H. Lühmann, P. Duchstein, D. Zahn, W. Bensch, Angew. Chem. Int. Ed. 2022, 61, e202202182; Angew. Chem. 2022, 134, e202202182.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.