3.1. Material and Synthesis
Attempts to synthesize new representatives of the composition of Li
3Ln[PS
4]
2 in analogy to the previously published Li
3La[PS
4]
2 [
20] following the reaction (2) led to the new compounds Li
4Ln[PS
4]
2Cl (
Ln = Pr, Nd and Sm) with one formally added LiCl equivalent.
The oven program used for their synthesis is explained in the experimental section and follows the results obtained from thermoanalytical measurements. It differs slightly from the procedure used to prepare Li3La[PS4]2 in that the reaction temperature was 150 °C less than for the lanthanum compound and ran significantly shorter although having more plateaus.
After the reaction process, single crystals were found and selected for Ln = Sm under the argon atmosphere of a glove box, which were solved for their crystal structure using SCXRD data. The structure refinement of Li4Sm[PS4]2Cl served as a starting point for the structure solution of the compounds incorporating Ln = Pr and Nd, where no suitable single crystals could be selected. However, powder X-ray diffraction measurements (PXRD) were utilized to confirm whether the structures are isotypic. The refinement of Li4Nd[PS4]2Cl is shown in detail below; however, due to unknown by-products, the refinement could not be accomplished reliably for Li4Pr[PS4]2Cl.
As expected from the stoichiometry of the starting materials and from the experience with the Li
3La[PS
4]
2 [
20] synthesis, lithium chloride was found as by-product. Additionally, small amounts of the ternary thiophosphates
Ln[PS
4] and some unidentified by-products could be observed in a few of the powder diffractograms.
To prove the stoichiometric formula resulting from the single-crystal X-ray diffraction experiments to be correct, especially concerning phosphorus, sulfur and chlorine in the compound, EDXS measurements were performed for the samarium-containing single crystal. The results are presented in
Table 1 and support the formula of Li
4Sm[PS
4]
2Cl very well. The observed oxygen content may be due to the short time when the sample was not protected by inert-gas conditions during its transfer. The vulnerability of the crystal towards ambient conditions could also be observed in the taken SEM picture (
Figure 1), as the surface of the sample looks rather rough. The moisture sensitivity was also confirmed by PXRD measurements. A sample of the bulk material was measured directly after storage under an argon atmosphere and for two consecutive days after being removed from the glove box. The loss of intensity of the reflection over a period of two days stemming from the target compound versus the stable intensities of lithium chloride hinted at a severe degradation of Li
4Sm[PS
4]
2Cl.
3.2. Structure Description from Single-Crystal XRD Data and Discussion of PXRD
The title compounds with the structured formula Li
4Ln[PS
4]
2Cl with
Ln = Pr, Nd and Sm crystallize isotypically in the monoclinic space group type
C2/
c with the lattice parameters shown in
Table 2 and
Table 3, allowing twelve formula units in the unit cell. A picture of the extended unit-cell content of Li
4Sm[PS
4]
2Cl as a representative of all three Li
4Ln[PS
4]
2Cl compounds viewed along the
c-axis and highlighting the complex [PS
4]
3− anions is presented in
Figure 2.
The single-crystal structure refinement could only be solved satisfyingly by introducing the two-fold twinning law (−0.5 −0.5 1; −0.5 −0.5 −1; 0.5 −0.5 0) into the calculation. Without this implementation, the structure solution appears to be also possible using the trigonal space group
R, due to generating symmetry on all three axes. The calculated solution in
R seemed chemically very plausible; however, the weighted
R2-value was far too high (w
R2 ≈ 0.43), as well as other refinement values (
Table 2). This supported the solution in the monoclinic space group
C2/
c, which led to the data presented in this article with the quality indices shown in
Table 2. The concept of the trigonal lattice, as well as an explanation for the tendency of crystals to form twins, becomes clear when observing the 2 × 2 × 2 unit cell from the viewing direction of [1 0
]. The threefold axis through all chloride anions suggests trigonal symmetry (
Figure 3).
The graphical results of the PXRD refinements for the neodymium-containing phase are presented in
Figure 4 for the trigonal setting and
Figure 5 for the monoclinic one. Both structure models led to acceptable residuals (trigonal structure:
R = 6.3%, monoclinic structure:
R = 5.4%). The graphical result of the fit using the monoclinic structure model is clearly better than the fit for the trigonal one, in particular in the range between 2
θ = 10 and 20°. Although this is partially attributed to the increased number of parameters used for the refinement (monoclinic model: 94 independent parameters, trigonal model: 75 independent parameters), we evaluate the monoclinic structure model as more suitable. We suggest that the monoclinic structure refinement as a superstructure reflects the actual crystal structure of Li
4RE[PS
4]
2Cl better than the trigonal setting. The crystallographic data are given in
Table 3.
The PXRD pattern of Li
4Pr[PS
4]
2Cl shows vast similarities to the neodymium analog; hence, we expected this phase to crystallize in the same monoclinic structure. A close inspection of the diffraction data revealed the presence of LiCl (
Figure 6, green tick marks) and of unindexed reflections (
Figure 6, grey line, e.g., at 2
θ = 7.5, 13.5 and 13.7°). As the diffractogram cannot be indexed, including these reflections, and as a symmetry reduction or using a reasonable supercell, does not lead to a proper description of these peaks, we assigned them to an unidentified by-product. Due to peak overlaps, we were not able to perform a fully weighted Rietveld [
28] refinement of Li
4Pr[PS
4]
2Cl, so we extracted the lattice parameters from a Le Bail fit [
27]. Thus, Li
4Pr[PS
4]
2Cl crystallizes in the space group
C2/
c with
a = 2104.5(6) pm,
b = 1581.3(3) pm,
c = 1322.3(5) pm and
β = 110.02(3)° for
Z = 12 (
Table 4). These axes would correspond to
a = 1581–1590 pm and
c = 2842 pm, if the structure were set up trigonally. The
a- and
b-axis differ slightly as these values are recalculated from the monoclinic results, showing once more that the monoclinic version is indeed more precise. The exact peak positions and integral peak intensities of the unknown by-product were extracted by single line fits (
Table 5). These peaks do not match any reference data of known lithium or praseodymium chlorides, thiophosphates, phosphates or mixed chloride thiophosphate phases.
In total, there are 27 different crystallographic sites within the monoclinic structure presented in
Table 6 for the single-crystal X-ray diffraction experiment of Li
4Sm[PS
4]
2Cl. For comparison, the 19 crystallographic sites of the trigonal representation are depicted in
Table 7. The crystallographic coordinates for Li
4Nd[PS
4]
2Cl from the powder X-ray diffraction experiments are shown in
Table 8.
There are two crystallographically different samarium sites. Both Sm1 and Sm2 are coordinated by eight sulfur atoms in the shape of bicapped trigonal prisms, with distances as presented in
Table 9.
This appears quite usual for lanthanoids; however, it is slightly different than in the ternary thiophosphates
Ln[PS
4] [
33] or in Li
3La[PS
4]
2 [
20], where the eightfold coordinated
Ln3+ cations are centering trigonal dodecahedra. The bond lengths are within the usual range of [
LnS
8]
13− polyhedra with
d(Sm–S) = 280–299 pm comparing well with those in Sm[PS
4] (285–294 pm) [
33]. The coordination polyhedron around Sm2 is slightly more distorted than the one for Sm1. These bicapped trigonal prisms of sulfur are constructed by [PS
4]
3− tetrahedra, whose ligands surround the three different phosphorus sites. The P–S distances are very much in the usual range for the complex [PS
4]
3− anions with bond lengths between 200 and 207 pm along with “tetrahedral” angles ranging from 106 to 112°.
Table 6.
Atomic coordinates and equivalent isotropic displacement parameters for Li4Sm[PS4]2Cl (monoclinic, C2/c).
Table 6.
Atomic coordinates and equivalent isotropic displacement parameters for Li4Sm[PS4]2Cl (monoclinic, C2/c).
Atom | Wyckoff site | x/a | y/b | z/c | Ueq */pm2 |
---|
Sm1 | 4e | 1/2 | 0.41098 (8) | 1/4 | 0.0201 (4) |
Sm2 | 8f | 0.33338 (5) | 0.08096 (6) | 0.42856 (6) | 0.0186 (3) |
P1 | 8f | 0.1558 (3) | 0.0280 (3) | 0.3298 (3) | 0.0198 (11) |
P2 | 8f | 0.4330 (3) | 0.2513 (3) | 0.3808 (3) | 0.0204 (11) |
P3 | 8f | 0.1494 (3) | 0.4707 (3) | 0.2997 (3) | 0.0205 (12) |
Cl1 | 4e | 0 | 0.2506 (6) | 1/4 | 0.076 (3) |
Cl2 | 8f | 0.1877 (3) | 0.2405 (4) | 0.0696 (3) | 0.0413 (14) |
S1 | 8f | 0.3969 (3) | 0.4938 (3) | 0.3252 (3) | 0.0284 (13) |
S2 | 8f | 0.0906 (3) | 0.0161 (3) | 0.4135 (3) | 0.0277 (13) |
S3 | 8f | 0.1912 (3) | 0.1490 (3) | 0.3394 (3) | 0.0243 (12) |
S4 | 8f | 0.2629 (3) | 0.4468 (3) | 0.1049 (3) | 0.0287 (12) |
S5 | 8f | 0.4657 (3) | 0.1312 (3) | 0.4187 (3) | 0.0297 (13) |
S6 | 8f | 0.4948 (2) | 0.3369 (3) | 0.0384 (3) | 0.0236 (11) |
S7 | 8f | 0.4083 (3) | 0.2711 (3) | 0.2176 (3) | 0.0333 (15) |
S8 | 8f | 0.3479 (3) | 0.2629 (3) | 0.4239 (3) | 0.0252 (12) |
S9 | 8f | 0.2077 (3) | 0.3768 (3) | 0.3852 (3) | 0.0338 (14) |
S10 | 8f | 0.1011 (3) | 0.4356 (3) | 0.1436 (3) | 0.0255 (12) |
S11 | 8f | 0.2918 (3) | 0.0745 (3) | 0.2011 (3) | 0.0281 (13) |
S12 | 8f | 0.4188 (3) | 0.0008 (3) | 0.1243 (3) | 0.0266 (13) |
Li1 | 4e | 0 | 0.403 (3) | 1/4 | 0.048 (13) |
Li2 | 4e | 1/2 | 0.114 (3) | 1/4 | 0.056 (14) |
Li3 † | 8f | 0.076 (7) | 0.284 (7) | 0.079 (8) | 0.11 (4) |
Li4 † | 8f | 0.105 (7) | 0.161 (8) | 0.132 (9) | 0.12 (4) |
Li5 | 8f | 0.305 (2) | 0.199 (3) | 0.093 (3) | 0.063 (11) |
Li6 | 8f | 0.079 (2) | 0.177 (3) | 0.401 (3) | 0.068 (11) |
Li7 | 8f | 0.178 (3) | 0.075 (4) | 0.064 (5) | 0.12 (2) |
Li8 | 8f | 0.300 (5) | 0.367 (5) | 0.329 (6) | 0.13 (2) |
Table 7.
Atomic coordinates and equivalent isotropic displacement parameters for Li4Sm[PS4]2Cl (trigonal, R).
Table 7.
Atomic coordinates and equivalent isotropic displacement parameters for Li4Sm[PS4]2Cl (trigonal, R).
Atom | Wyckoff site | x/a | y/b | z/c | Ueq */pm2 |
---|
Sm | 18f | 0.33662 (18) | 0.33405 (17) | 0.08330 (8) | 0.0309 (7) |
P1 | 18f | 0.3743 (7) | 0.1807 (6) | 0.0036 (3) | 0.0204 (19) |
P2 | 18f | 0.5131 (7) | 0.0391 (8) | 0.1629 (3) | 0.023 (2) |
Cl1 | 6c | 0 | 0 | 0.0689 (6) | 0.028 (3) |
Cl2 | 6c | 0 | 0 | 0.2447 (6) | 0.045 (5) |
Cl3 | 6c | 0 | 0 | 0.4374 (5) | 0.038 (4) |
S1 | 18f | 0.3287 (7) | 0.4771 (7) | 0.0168 (3) | 0.026 (2) |
S2 | 18f | 0.0951 (10) | 0.4181 (10) | 0.0075 (4) | 0.043 (3) |
S3 | 18f | 0.3496 (7) | 0.1588 (7) | 0.0746 (4) | 0.026 (2) |
S4 | 18f | 0.1578 (8) | 0.2475 (9) | 0.0322 (5) | 0.049 (4) |
S5 | 18f | 0.0117 (8) | 0.2500 (12) | 0.1593 (4) | 0.045 (3) |
S6 | 18f | 0.1854 (6) | 0.1897 (7) | 0.1503 (3) | 0.0231 (19) |
S7 | 18f | 0.2406 (12) | 0.4196 (10) | 0.1340 (6) | 0.056 (4) |
S8 | 18f | 0.4929 (7) | 0.0140 (7) | 0.0911 (3) | 0.024 (2) |
Li1 | 18f | 0.515 (4) | 0.186 (4) | 0.091 (2) | 0.016 (11) |
Li2 | 18f | 0.165 (6) | 0.023 (6) | 0.065 (3) | 0.036 (17) |
Li3 | 18f | 0.030 (4) | 0.334 (4) | 0.0916 (17) | 0.008 (9) |
Li4 | 18f | 0.171 (5) | 0.550 (5) | 0.067 (3) | 0.032 (15) |
Table 8.
Atomic coordinates of monoclinic Li4Nd[PS4]2Cl stemming from the powder X-ray diffraction data.
Table 8.
Atomic coordinates of monoclinic Li4Nd[PS4]2Cl stemming from the powder X-ray diffraction data.
Atom | Wyckoff site | x/a | y/b | z/c |
---|
Nd1 | 4e | 1/2 | 0.4146 (8) | 1/4 |
Nd2 | 8f | 0.3342 (4) | 0.0813 (1) | 0.4287 (6) |
P1 | 8f | 0.1587 (19) | 0.024 (2) | 0.339 (2) |
P2 | 8f | 0.4365 (17) | 0.255 (3) | 0.393 (3) |
P3 | 8f | 0.15663 (18) | 0.481 (2) | 0.310 (3) |
Cl1 | 4e | 0 | 0.229 (2) | 1/4 |
Cl2 | 8f | 0.1833 (12) | 0.245 (2) | 0.055 (2) |
S1 | 8f | 0.200 (7) | 0.149 (5) | 0.345 (9) |
S2 | 8f | 0.112 (5) | 0.015 (7) | 0.184 (4) |
S3 | 8f | 0.096 (4) | 0.014 (6) | 0.424 (6) |
S4 | 8f | 0.263 (3) | 0.430 (5) | 0.103 (7) |
S5 | 8f | 0.497 (4) | 0.339 (6) | 0.028 (5) |
S6 | 8f | 0.360 (3) | 0.266 (6) | 0.445 (5) |
S7 | 8f | 0.415 (6) | 0.278 (6) | 0.224 (5) |
S8 | 8f | 0.471 (4) | 0.127 (4) | 0.430 (6) |
S9 | 8f | 0.119 (5) | 0.440 (7) | 0.162 (4) |
S10 | 8f | 0.296 (5) | 0.088 (5) | 0.200 (7) |
S11 | 8f | 0.220 (5) | 0.380 (6) | 0.400 (6) |
S12 | 8f | 0.412 (4) | 0.017 (7) | 0.125 (7) |
Li1 | 4e | 0 | 0.402 | 1/4 |
Li2 | 4e | 1/2 | 0.120 | 1/4 |
Li3 † | 8f | 0.079 | 0.275 | 0.097 |
Li4 † | 8f | 0.104 | 0.168 | 0.120 |
Li5 | 8f | 0.303 | 0.198 | 0.094 |
Li6 | 8f | 0.076 | 0.176 | 0.403 |
Li7 | 8f | 0.174 | 0.078 | 0.071 |
Li8 | 8f | 0.306 | 0.382 | 0.321 |
Table 9.
Selected interatomic distances (d/pm) for Li4Sm[PS4]2Cl (monoclinic, C2/c). For further details on the remaining distances as well as the angles between the atoms, see the entries in the ICSD.
Table 9.
Selected interatomic distances (d/pm) for Li4Sm[PS4]2Cl (monoclinic, C2/c). For further details on the remaining distances as well as the angles between the atoms, see the entries in the ICSD.
Atoms | d/pm | Atoms | d/pm |
---|
Sm1–S2 | 285.5 (4) | Sm2–S11 | 280.4 (4) |
Sm1–S2 | 285.5 (4) | Sm2–S4 | 285.1 (5) |
Sm1–S7 | 286.0 (5) | Sm2–S12 | 288.0 (4) |
Sm1–S7 | 286.0 (5) | Sm2–S8 | 289.3 (5) |
Sm1–S1 | 296.1 (5) | Sm2–S5 | 292.0 (6) |
Sm1–S1 | 296.1 (5) | Sm2–S9 | 292.5 (5) |
Sm1–S6 | 297.5 (4) | Sm2–S10 | 298.7 (5) |
Sm1–S6 | 297.5 (4) | Sm2–S3 | 299.5 (5) |
P1–S1 | 202.3 (5) | P2–S5 | 201.9 (6) |
P1–S2 | 202.7 (7) | P2–S6 | 203.3 (6) |
P1–S3 | 203.8 (6) | P2–S7 | 204.3 (5) |
P1–S4 | 206.8 (7) | P2–S8 | 205.0 (7) |
P3–S9 | 200.2 (6) | | |
P3–S10 | 202.3 (5) | | |
P3–S11 | 205.0 (7) | | |
P3–S12 | 205.2 (7) | | |
Cl1–Li6 | 239 (4) | Cl2–Li5 | 240 (4) |
Cl1–Li6 | 239 (4) | Cl2–Li5 | 244 (5) |
Cl1–Li1 | 241 (5) | Cl2–Li3 | 248 (13) |
| | Cl2–Li4 | 249 (14) |
| | Cl2–Li7 | 262 (6) |
In the trigonal setting, there is only one Sm
3+ site present, as expected, also coordinated by eight sulfur atoms with
d(Sm–S) = 284–300 pm, which belong to only two crystallographically distinct [PS
4]
3− tetrahedra. The three resulting samarium surroundings, two from the monoclinic refinement and one from the trigonal one, are depicted in
Figure 7 for comparison, but all show the edge-grafting of four [PS
4]
3− units at the Sm
3+ centers.
In the monoclinic structure solution, the [(Sm2)S
8]
13− polyhedra, linked via [(P3)S
4]
3− anions, developed infinite strands along the
c-axis. These strands are connected via [(P1)S
4]
3−, [(Sm1)S
8]
13− and [(P2)S
4]
3− polyhedra along [110], forming the three-dimensional network
(
Figure 8) of the Li
4Sm[PS
4]
2Cl structure.
Moreover, there are two different chlorine sites, with Cl1 having three lithium atoms in closer proximity (
d((Cl1)–Li) = 239–241 pm) and Cl2 with five of them. However, around (Cl2)
−, one of the Li
+ cations is further away; thus, describing the coordination sphere as a 4 + 1 environment might seem more plausible. The distances between (Cl2)
− and Li
+ are
d((Cl2)–Li) = 240–249 pm
plus 261 pm). The (Cl2)
− anions form interesting dimers (
Figure 8), where it almost seems like the Li
+ cations meander around the two chloride anions. The Cl
− anions are bridged by two (Li5)
+ cations and end in three terminal lithium ions on either side. Finally, the (Cl1)
− anions separate these dimers from each other.
The chlorine positions themselves, however, are quite interesting as they appear to be residing within channel-like pores made of [PS
4]
3− tetrahedra, quite similar to the Li
6[PS
4]SCl [
6] case. These channels can best be seen, when viewing the monoclinic unit cell along [1 0
] direction (
Figure 3).
There are eight crystallographically distinct Li+-cation positions within the crystal structure of Li4Sm[PS4]2Cl. (Li1)+, which is surrounded by four sulfur atoms and one Cl− anion, as well as (Li6)+ and (Li7)+; however, the shape of the [LiS4Cl]8− polyhedra differ strongly from each other. Two of the bonds from (Li1)+ to S2− are rather long (295 pm), and the coordination polyhedron can therefore be described as [(Li1)S2+2Cl]8−. This is not the case for (Li6)+, where there are also longer bonds (292 and 276 pm), but there is no clear cutoff as there is for (Li1)+. The shape of the polyhedron surrounding (Li7)+ is more closely related to the (Li6)+-centered polyhedron, but now the angles are slightly different. The distances between the central lithium and the adjacent sulfur range from 244 to 289 pm. For all three lithium cations, (Li1)+, (Li6)+ and (Li7)+, the distances to the chloride ligand range between 236 and 260 pm. The coordination surrounding (Li2)+ is rather unique as there are six sulfur atoms surrounding it in the shape of a heavily distorted octahedron. However, there are four shorter distances (256–262 pm) and two much longer ones (2 × 307 pm). On the other hand, (Li8)+ is coordinated by four sulfur atoms, though strongly distorted, so it somehow bridges the relation between (Li2)+ and (Li6)+ as it does not show the two extra sulfur ligands at rational distances and the four coordinating sulfur atoms are 210–304 pm apart. Two chlorine and two sulfur ligands coordinate the (Li5)+ cation in the shape of a distorted tetrahedron. The bond lengths between lithium and chlorine range from 238 to 245 pm, and the distance between the (Li5)+ and sulfur is 249 pm. (Li3)+ is coordinated by one chlorine and three sulfur ligands. This [LiS3Cl]6− polyhedron is a motif also present for (Li4)+. The shape of these coordination environments is a chair-like geometry showing for both (Li3)+ and (Li4)+, very similar with distances ranging from 247 to 286 pm with the Li–Cl bond being the shortest one in the case of (Li3)+.
Only four crystallographically distinct sites in the trigonal refinement host Li
+ cations achieve charge neutrality (
Table 8). If some of them might suffer from only partial occupation, two more show up in the difference
Fourier map of the final refinement. Both refer to the special positions 9
e (
1/
2, 0, 0) and 9
f (
1/
2, 0,
1/
2), which could represent pseudo-octahedral migration pathway stops with six distances from lithium to the involved sulfur atoms between 243 and 258 pm for each one of them. Attempts for their inclusion in the structure refinements did not reduce the residuals and resulted in unstable values for their calculated occupation factors and
Ueq parameters. Since locating lithium atoms from X-ray diffraction experiments is already challenging, there remains, of course, no physical evidence to support this claim. However, it further strengthens the argumentation for the monoclinic description being more feasible.