Electrochemical cells based on intercalation and deintercalation of chalcogen anions

Abstract

An electroactive material suitable for electrochemical cell electrode wherein the electroactive material includes a chalcogen oligomer that can reversibly deintercalate/reintercalate an anion of the chalcogen, such as La.sub.2O.sub.2S.sub.2, and the electrochemical electrodes and cells containing the electroactive material.

Claims

1. An electroactive material for electrochemical cell electrode, said electroactive material comprising a chalcogen oligomer, wherein said electroactive material reversibly deintercalates an anion of said chalcogen by reduction and reintercalates said anion by oxidation, while maintaining its overall structure.

2. The electroactive material of claim 1 wherein said chalcogen anion is sulfur or an oligomer thereof of formula (III):
(S.sub.n).sup.x(III) where n and x are Integers, such that x equals to 1 or 2 and n is comprised between 1 and 6.

3. The electroactive material according to claim 1 wherein said chalcogen anion is S.sup.2.

4. The electroactive material according to claim 1 which is chosen from La.sub.2O.sub.2S.sub.2, SrS.sub.2, SrS.sub.3, BaS.sub.2, BaS.sub.3, Ba.sub.2S.sub.2F.sub.2, FeS.sub.2, NIS.sub.2, CoS.sub.2, MnS.sub.2, TIS.sub.3, VS.sub.4, PbS.sub.2, BIS.sub.2.

5. The electroactive material according to claim 1 wherein said active material is La.sub.2O.sub.2S.sub.2, which reversibly deintercalates and reintercalates S.sup.2 according to the following reaction:
La.sub.2O.sub.2S.sub.2+2e.sup.oA-La.sub.2O.sub.2S+S.sup.2 where oA designates the centered orthorhombic crystalline form.

6. The electroactive material according to claim 1 wherein said active material is a material of formula (I):
oA-La.sub.2O.sub.2S(I) where oA designates the centered orthorombic crystalline form.

7. The electroactive material according to claim 1 wherein said active material is a material of formula (II):
oA-La.sub.2O.sub.2S.sub.1.5(II) where oA designates the centered orthorhombic crystalline form.

8. A positive electrode comprising the electroactive material according to claim 1.

9. The positive electrode according to claim 8 comprising: a current; and a coating layer comprising said electroactive material.

10. The positive electrode according to claim 9, wherein the current collector is an aluminum sheet.

11. An electrochemical cell comprising: a positive electrode as defined in claim 8; a negative electrode; and an electrolyte layer sandwiched between the positive and the negative electrodes, wherein the electrolyte is a solid electrolyte comprising as conducting ion the chalcogen anion of the chalcogen oligomer of said electroactive material.

12. A battery comprising a plurality of electrochemical cells as defined in claim 11, wherein said electrochemical cells are electrically connected.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0055] FIG. 1 illustrates the principle of a sulfur-sulfur battery according to the invention, involving sulfur as the chacolgen and La.sub.2O.sub.2S.sub.2 as the electroactive material.

[0056] FIG. 2 illustrates (a) Structure of La.sub.2O.sub.2S.sub.2 reported by Ostorero et al. (SG: Cmca) (Acta Cryst. C46, 1376-1378 (1990)); (b) Conceptual scheme of SS bond cleavage under the donation of one electron per elemental metal M.sup.0 that triggers subsequently the deintercalation of half sulfur atom of the S.sub.2 dumbbell that possibly enables topochemical conversion of La.sub.2O.sub.2S.sub.2 into the new polymorph of La.sub.2O.sub.2S; (c) The two low-energy dynamically stable phases of oA-La.sub.2O.sub.2S predicted by USPEX.

[0057] FIG. 3 illustrates (a) the partial intercalation of sulfur into oA-La.sub.2O.sub.2S and de-intercalation of sulfur from La.sub.2O.sub.2S.sub.2 that leads to an intermediate compound oA-La.sub.2O.sub.2S.sub.1.5; (b) intercalation experiments of sulphur into oA-La.sub.2O.sub.2S. Experimental XRD patterns of pure oA-La.sub.2O.sub.2S and the products of its mixture with sulfur (0.5 or 1 equiv. of S) after thermal treatments at 150 or 200 C. The new XRD peaks emerging after the thermal treatment with 0.5 S are marked by*; (c) Deinterciation of sulphur from La.sub.2O.sub.2S.sub.2. Experimental XRD patterns of La.sub.2O.sub.2S.sub.2 and of the product of its mixture with Rb.sup.0, Ag.sup.0 and Ni.sup.0 after the thermal treatments at 200 or 350 C. The XRD peaks assigned to by-products are marked as follows: =Ag.sub.2S (Fr Krist.-Cryst. Mater. 110, 136-144 (1958); .square-solid.=-NiS (J. Trahan, R. G. Goodrich, S. F. Watkins, Phys. Rev. B 2, 2859-2863 (1970)).

[0058] FIG. 4 represents the overview of the rich low temperature sulfur topochemistry in the LaOS system. The topochemical intercalation and deintercalation of sulfur in the oxychalcogenide compound La.sub.2O.sub.2S.sub.2 lead to the formation of two new metastable compounds.

[0059] FIG. 5 represents (a) EDX spectrum of the oA-La.sub.2O.sub.2S powder sample impregnated with the epoxy resin and (b) its backscattered electron image (BEI) as well as its elemental composition mapped for La and S.

DETAILED DESCRIPTION OF EMBODIMENTS

[0060] In FIG. 1, the cell of such a sulfur-sulfur battery is schematically represented during discharge.

[0061] Said cell comprises a positive electrode 1 (cathode), and a negative electrode 2 (anode). A sulfur anion conducting electrolyte 3 is sandwiched between electrodes land 2.

[0062] Both electrodes 1 and 2 are electrically connected by means of an electrical circuit including an ammeter 9.

[0063] As depicted in FIG. 1:

[0064] The positive electrode 1 comprises a current collector 4 and a layer of electroactive material 5. The layer 5 is at the interface between the conducting electrolyte 3 and the inner face of the current collector 4. Typically, the current collector 4 may be an aluminum sheet.

[0065] The negative electrode 2 comprises a current collector 6 and a layer 7 at the interface between the conducting electrolyte 3 and the inner face of the current collector 6.

[0066] Generally, the current collector 6 of the negative electrode is made of copper.

[0067] The layer 7 can consist of a sulfur composite or metal M which can react with sulfur anion according to the reaction: wM+S.sup.2.fwdarw.M.sub.wS+2e.sup.. It can also consist of another material capable of intercalating and deintercalating sulfur anions.

[0068] During a discharge, the positive electrode 1 attracts electrons from the electrical circuit so that a reductive cleavage occurs, such as in the case of La.sub.2O.sub.2S.sub.2:

La.sub.2O.sub.2S.sub.2+2e.sup.oA-La.sub.2O.sub.2S+S.sup.2

[0069] The S.sup.2 anion migrate through the electrolyte 8 from the positive electrode 1 towards the negative electrode 2 and are collected at the negative electrode 2 to undergo an oxidation, releasing electrons: S.sup.2, S.sup.0+2 e.sup. or wM+S.sup.2M.sub.wS+2e.sup.

[0070] The resulting electrons are then migrating back to the positive electrode 1 through the electrical circuit 9.

[0071] Although not represented in FIG. 1, the opposite reactions occur in charge, where the positive electrode becomes the anode (set of the oxidation) and the negative electrode becomes the cathode (seat of the reduction).

[0072] The following examples are given for illustrative purposes only.

Examples

[0073] La.sub.2O.sub.2S.sub.2 was used as the precursor to test the topochemical reduction for chalcogenides (FIG. 2a). Its structure consists of fluorite-type 2/[La.sub.2O.sub.2].sup.2+ infinite layers separated from each other by isolated (S.sub.2).sup.2 sulfur dimers aligned in parallel to these 2D blocks. The de-insertion of one sulfur atom per dimer should lead a priori to a La.sub.2O.sub.2S compound (FIG. 2b) whose structure should be inherited from the layered structure of the precursor La.sub.2O.sub.2S.sub.2. First, the low-energy structures of La.sub.2O.sub.2S compound using a designed crystal structure prediction (CSP) methodology was investigated. The combination of USPEX structure searching evolutionary algorithm with first-principles calculations makes it possible to locate two polymorphs, namely hP and oA crystal structures that are respectively stable and metastable (see FIG. 2c). Both phases are dynamically stable, justifying their respective location at global and local minima on the potential energy surface of La.sub.2O.sub.2S. The most stable candidate exhibits a hexagonal layered structure with 2/[La.sub.2O.sub.2]fluorite-type (111) slab alternating with sulfur atoms in octahedral environment of lanthanum. Interestingly, this is the exact structure of the La.sub.2O.sub.2S compound reported in the literature (Acta Cryst. B29, 2647-2648 (1973)), commonly prepared at high temperature (800-1200 C.). In the following this structure will be noted hP-La.sub.2O.sub.2S according to the Pearson notation (h for hexagonal and P for primitive cell). USPEX predicted also the structure of a unknown metastable polymorph with an enthalpy only slightly higher. This structure displays also a layered feature but is built upon the stacking of 2/[La.sub.2O.sub.2] fluorine-type (001) slabs (full reminiscence of the La.sub.2O.sub.2S.sub.2 structure) alternating with sulfur atoms in prismatic environments. In the same way with hP-La.sub.2O.sub.2S, this metastable polymorph with orthorhombic Amm2 space group is named hereafter oA-La.sub.2O.sub.2S. The thermal and kinetic stability of these two structures were further confirmed by ab initio molecular dynamics (AIMD) simulation in which both hP- and oA-La.sub.2O.sub.2S retained their main structural framework after 10 ps at temperatures up to 600 K. Consequently, the theoretical calculations clearly anticipate the possible existence of metastable oA-La.sub.2O.sub.2S besides the already known phase hP-La.sub.2O.sub.2S.

[0074] The topochemical deinsertion of sulfur in the layered precursor La.sub.2O.sub.2S.sub.2 was subsequently attempted by reaction with alkali metal Rb.sup.0 at low-temperature in evacuated and sealed pyrex tubes. Once Rb excess (and its salts) was washed out by dry ethanol (see synthetic procedure in SI), the powder X-ray diffraction (XRD) patterns were collected on products synthesised at 200 C. and 350 C. Both of them turn out to be very similar and did not bring to light any known phases. Moreover, the EDX analyses of the bulk product powder clearly revealed the absence of rubidium and a molar ratio La/S of 2.0(2)/1.0(1) (See FIG. 5). These results indicate the formation of a sulfur-deficient La.sub.2O.sub.2S phase without incorporation of Rb in the structure. The hP-La.sub.2O.sub.2S XRD peaks were not detected at all in the X-ray pattern but the existence of the polymorph oA-La.sub.2O.sub.2S predicted by USPEX could be readily established via a Rietveld refinement with goodness of fit X.sup.2=1.33 and Bragg reliability factor R(obs)=1.67%: see Table below:

TABLE-US-00001 TABLE 1 Crystallographic parameters determined from Rietveld refinement of oA- La.sub.2O.sub.2S powder. Crystallographic data Chemical formula La.sub.2O.sub.2S Molar mass (g mol.sup.1) 341.87 Symmetry Orthorhombic Color White Space group Amm2 (No. 38) a () 4.1489(1) b () 3.9750(9) c () 12.728(0) Volume (.sup.3) 209.9(1) Z 2 Density (g cm.sup.3) 5.4088 Anisotropic strain (.sup.2).sup.2 S.sub.400 = 11.8(9); S.sub.040 = 8.66(0); S.sub.004 = 0.0485(9); S.sub.220 = 2.82(0); S.sub.202 = 0.830(7); S.sub.022 = 0.636(8); March-Dollase parameter P.sub.md = 0.943(5) (Preferred orientation along <100>) Structural refinement Profile reliability factor R.sub.p = 6.38% Weighted profile reliability factor R.sub.wp = 8.73% Bragg reliability factors R(obs) = R(all) = 1.67% Weighted Bragg profile reliability factors R.sub.w(obs) = R.sub.w(all) = 2.29% Goodness of fit x.sup.2 = 1.33 Atomic positions and isotropic thermal parameters Atom x y z U.sub.iso (.sup.2) La1 0 0 0.6442(4) 0.0054(8) La2 0.5 0.5 0.8379(7) 0.0040(1) O1 0 0.5 0.7350(1) 0.001.sup.b O2 0.5 0 0.7169(3) 0.001.sup.b S1 0 0 0.9664(7) 0.0068(4) Site-occupancy factors of all atoms are fixed to full occupancy. .sup.bThese atomic displacement factors are fixed to 0.001.

[0075] Scanning Transmission Electron Microscopy (STEM) also support the conclusion that the newly synthesized phase is oA-La.sub.2O.sub.2S. The stacking of 2/[La.sub.2O.sub.2] infinite sheets with the fluorine-type (100) slab structure is clearly visible on the High Angle Annular Dark Field (HAADF) STEM image. In contrast, the fluorite-type (111) slabs characteristic of the stable polymorph hP-La.sub.2O.sub.2S could not be found in the experiment STEM image. The EDX spectrum of a nanosized single crystal was, similarly to the EDX analysis of the bulk powder, consistent with the composition of La.sub.2O.sub.2S. The structural arrangement of the new oA-La.sub.2O.sub.2S compound is directly inherited from the one of the La.sub.2O.sub.2S.sub.2. This observation definitely supports the topochemical nature of the deintercalation process. The sulfur deintercalation process does not modify at all the integrity of the 2/[La.sub.2O.sub.2] slab but entails a shift of one 2/[La.sub.2O.sub.2] layer over two along the % (b+c) direction of the pristine La.sub.2O.sub.2S.sub.2 structure (SG: Cmca). Raman spectroscopy confirmed the complete loss of the sulfur dimers along the topochemical reduction: the band associated to the SS stretching modes located at 487 and 498 cm.sup. in La.sub.2O.sub.2S.sub.2 have totally disappeared after the deintercalation of one sulfur from La.sub.2O.sub.2S.sub.2 confirming the conclusion made from the XRD pattern that the reaction of hP-La.sub.2O.sub.2S towards oA-La.sub.2O.sub.2S was complete. Finally, the diffuse-reflectance spectra also support the cleavage of (S.sub.2).sup.2 dimers. The absorption thresholds move from 2.50 eV in La.sub.2O.sub.2S.sub.2, a value characteristic of a *-* electronic transition of isolated pairs, to 3.88 eV in oA-La.sub.2O.sub.2S, a value slightly lower than the one observed in the hP-La.sub.2O.sub.2S (4.13 eV). Thus, it was concluded that during the reaction with La.sub.2O.sub.2S.sub.2 with elemental rubidium, the alkali metal activates a redox reaction with (S.sub.2).sup.2 dimers that trigger the fracture of the SS bonds. However, contrary to Cu.sup.0 nanoparticles that intercalate into the La.sub.2O.sub.2S.sub.2 host lattice (Angew. Chem. Int. Ed. 57, 13618-13623 (2018)) Rb.sup.0 leads to the topochemical de-insertion of sulfur to afford the oA-La.sub.2O.sub.2S metastable phase. The choice of reducing agents is the decisive factor on the consequence of the reaction. No reaction occurred when La.sub.2O.sub.2S.sub.2 was treated at 200-300 C. under reducing atmosphere, i.e. 5% H.sub.2/Ar flow. At 350 C. the reduction finally happened, but it ended up with the thermodynamically stable hP-La.sub.2O.sub.2S, where the original fluorite (100) slab was deformed into the fluorite (111) slab. This result highlights the contrast between the common reducing agent such as H.sub.2 and the more powerful reducing agent Rb.sup.0, which favored even at the same reaction temperature (350 C.) topochemical reduction to oA-La.sub.2O.sub.2S.

[0076] At this stage, it was hypothesized that the topotactic de-intercalation of La.sub.2O.sub.2S.sub.2 may be reversible or not at low temperature. To test this possibility, a portion of oA-La.sub.2O.sub.2S was mixed with one equivalent sulfur and heated at 200 C. (FIG. 3a). The product was analyzed by means of XRD as shown in FIG. 3b. The original La.sub.2O.sub.2S.sub.2 material could be fully recovered with no sub-product confirming the reversible character of the temperature-assisted intercalation/de-intercalation processes based on the formation/rupture of sulfur dimers within the La.sub.2O.sub.2S/La.sub.2O.sub.2S.sub.2 layered oxychalcogenides. To gain more insight about the intercalation of sulfur, the reactivity of oA-La.sub.2O.sub.2S towards only half equivalent of sulfur at low temperature was also tested. The XRD pattern of the product obtained from intercalation of 0.5 S in oA-La.sub.2O.sub.2S at 200 C. (see FIG. 3b) clearly evidences the conversion of oA-La.sub.2O.sub.2S into an unknown intermediate phase along with a small amount of La.sub.2O.sub.2S.sub.2. The XRD pattern of the intermediate phase was similar to that of oA-La.sub.2O.sub.2S but shifted to lower diffraction angles, suggesting the existence of an intercalated oA-La.sub.2O.sub.2S.sub.x phase (1<x<2.0). The same XRD pattern was observed in the attempt to de-intercalate 0.5 S from La.sub.2O.sub.2S.sub.2 using 1.0 equiv. of Rb.sup.0, 1.0 equiv. of Ag and 0.5 equiv. of Ni.sup.0 (FIG. 3c). Diffraction pattern of oA-La.sub.2O.sub.2S.sub.1.5 could be refined with the same space group as oA-La.sub.2O.sub.2S (Amm2) and cell parameters of 8.4 , 4.0 and 12.8 without any superstructure peak. This clearly proved the existence of an intermediate phase with a strong reminiscence of the oA-La.sub.2O.sub.2S structure. One of the reasonable assumptions is that this new intermediate phase replaced one half of monoatomic S.sup.2 with dimeric (S.sub.2).sup.2 anions retaining the main structural framework of oA-La.sub.2O.sub.2S. This partial dimerization should lead to the expected oA-La.sub.2O.sub.2S.sub.1.5 composition. Indeed, both intercalation of 0.5 S and de-intercalation of 0.5 S using metal species gave the similar Raman spectra that featured the single intense peak at 413-417 cm.sup. while Raman peaks from oA-La.sub.2O.sub.2S nor La.sub.2O.sub.2S.sub.2 were absent. Since an intense peak around 400-500 cm.sup. is characteristic of SS stretching mode (Angew. Chem. Int. Ed. Engl. 14, 655-720 (1975)) these Raman spectra support the formation of oA-La.sub.2O.sub.2S.sub.1.5 through the partial cleavage of SS bonds.

[0077] To solve the crystal structure of this novel phase, precession electron diffraction tomography (PEDT) analyses were performed. This emerging technique can reduce dynamical diffraction effects during data collection and enables complex structures to be solved ab initio using single nanocrystals. PEDT data were therefore collected on several nanocrystals of the novel phase. All data sets were analyzed using the computer programs PETS2.0 (Acta Crystallographica, B75, 512-522 (2019).), Superflip (Journal of Applied Crystallography, 40, 451-456. (2007)) and Jana2006 (Zeitschritt fr Kristallographie, 229, 345-352. (2014)). The reconstruction of the reciprocal lattice planes hk0, h0l and 0kl was observed which are consistent with an orthorhombic unit cell a=8.348 , b=3.961 and c=12.645 (V=418.1 .sup.3) and a non-centrosymmetric space group Amm2. The structure was subsequently solved and refined using the Jana2006 Program on the basis of electron diffraction data. The structural analysis converged with electron Bragg reliability factor R(obs)=10.1%, revealing a layered structure with a composition oA-La.sub.2O.sub.2S.sub.1.5. This obtained new structure consists of 2/[La.sub.2O.sub.2] fluorine-type (001) infinite slab alternating with sulfur layers containing one third and two thirds of sulfur anions as S.sup.2 and (S.sub.2).sup.2 species, respectively. Using this oA-La.sub.2O.sub.2S.sub.1.5 structure model, both powder XRD patterns from sulfur intercalation and de-intercalation, i.e. from oA-La.sub.2O.sub.2S+0.5 S and La.sub.2O.sub.2S.sub.2+0.5 Ni reaction mixtures (see FIG. 3), were successfully refined. Large strain parameters had to be considered to reach satisfactory fitting. This can be interpreted as the signature of a stacking disorder occurring, as expected, during the intercalation or de-intercalation processes in relation with the 2D structure of the host lattice and possible existence of different stages. The structural analysis was based on data collected on the best crystallized crystals. However, in most of the PEDT data, the stacking faults lead to diffuse scattering features along [001]. The experimental contrast in the HAADF-STEM image asserts the stacking of 2/[La.sub.2O.sub.2] fluorine-type (001) infinite slabs. A similar structure was predicted independently by the evolutionary algorithm USPEX for this specific oA-La.sub.2O.sub.2S.sub.1.5 composition. The structure predicted to be the most stable accorded well with the experimental structure obtained by the PEDT analysis. The 2.sup.nd and 3.sup.rd most stable structures displayed 1D slabs and 2D hexagonal (fluorine-type (111)) slabs as their [La.sub.2O.sub.2] units and these slabs constituted intergrowth structures with (quasi-)2D arrays of sulfur dimers/atoms. However, none of them could be found in the experiments that were conducted.

[0078] This work demonstrates the de-intercalation and re-intercalation of sulfur in a layered oxychalcogenide compound using an original topochemical approach. Alkali or transition metals may be used as reducing agent to trigger the reduction of the chalcogenide oligomers and breaking the chalcogen-chalcogen bond. In the case of La.sub.2O.sub.2S.sub.2 the low temperature deintercalation of sulfur atoms proceeds in a two steps to form two new metastable phases oA-La.sub.2O.sub.2S.sub.1.5 and oA-La.sub.2O.sub.2S that retain the layered feature of the precursor. As illustrated in FIG. 4 this is a fully reversible topotactic process as the sulfur atoms may be re-intercalated at low temperature to form back the precursor La.sub.2O.sub.2S.sub.2.

TABLE-US-00002 TABLE 2 Summary of crystallographic parameters of La.sub.2O.sub.2S.sub.x series (1 x 2.0) La.sub.2O.sub.2S.sub.2 hP-La.sub.2O.sub.2S.sub.2 oA-La.sub.2O.sub.2S oA-La.sub.2O.sub.2S.sub.1.5 Source Ostorer Morosin This study This study et al. et al. (XRD) (PEDT) (XRD) (XRD) Space group Cmca P-3m1 Amm2 Amm2 a () 13.215(2) 4.049(1) 4.148(9) 8.348 b () 5.943(1) 3.975(1) 3.961 c () 5.938(1) 6.939(2) 12.728(0) 12.645 S-S distance 2.103 4.049 3.975 2.011 () Ostero et al Acta Cryst. C46, 1376-1378 (1990) Morosin et al Angew. Chem. Int. Ed. Engl. 14, 655-720 (1975)

1. Synthetic Procedures

[0079] The initial precursor La.sub.2O.sub.2S.sub.2 was synthesized following the procedure described in Angew. Chem. Int. Ed. 2018, 57, 13618-13623

oA-La.sub.2O.sub.2S: Topochemical de-intercalation of S.sup.2 anions by Rb

[0080] Prior to the preparation, all of experimental glassware and utensils were dried in the oven (T=80 C.). Under argon atmosphere, La.sub.2O.sub.2S.sub.2 and Rb (Aldrich, 98+%) were weighed in 1:2 molar ratio and introduced into the silica tube whose bottom was protected by carbon coating. All these preparations were done under argon atmosphere. Then the pyrex tube was evacuated (10.sup.3 torr) and sealed. The sealed mixture was heated to 200 C. at a rate of 20 C. h.sup.1 and annealed for 2 h. Finally the sealed mixture was gradually cooled in a furnace to give the pale greyish-blue powder. The surplus Rb deposited on the opposite side of the silica tube. The silica tube was opened under argon atmosphere and all the content was quenched with the excess amount of ethanol (Attention: under the ambient atmosphere, Rb ignites upon the contact with ethanol). The colorless precipitate was contaminated by the tiny flakes of carbon, which were separated by repetitive decantation with mechanical agitation. The precipitate was then washed with ethanol, water, and acetone, followed by dryness in vacuo to afford the colorless powder of oA-La.sub.2O.sub.2S. The product was stable under the ambient atmosphere. The same reaction performed at 350 C. also gave the identical results: pure oA-La.sub.2O.sub.2S without any trace of hP-La.sub.2O.sub.2S nor any other impurities.

Intercalation of Sulfur Anions into oA-La.sub.2O.sub.2S

[0081] The colorless powder of oA-La.sub.2O.sub.2S (ca. 200 mg) was combined with S flakes (Aldrich, 99.99+%) in oA-La.sub.2O.sub.2S: S=1: 0.5 molar ratio and ground on an agate mortar under argon atmosphere. Then the mixture was pelletized and sealed in an evacuated (10.sup.3 torr) silica tube. The sealed mixture was heated to 150-200 C. at a rate of 100 C. h.sup.1 and annealed for 4-48 h (See FIG. 3b for the result), followed by gradual cooling in a furnace to afford the pale yellow pellet. When the sulfur was not completely consumed, the residual sulfur was deposited on the opposite side of the silica tube. To complete the intercalation, the obtained pellet was ground with additional 0.5 equiv. of S under argon atmosphere. The mixture was again subject to the thermal treatment at 200 C. in the evacuated silica tube. After 160 h of annealing, the mixture was fully converted into the pale yellow pellet of the pure La.sub.2O.sub.2S.sub.2.

oA-La.sub.404S.sub.3: General Procedure for topochemical de-intercalation of S.sup.2 anions by various metals

[0082] The detailed synthetic conditions (i.e. Stoichiometry, duration of annealing, forms of metal sources) for respective metal species were noted below. To 1.0 equiv. of La.sub.2O.sub.2S.sub.2 (ca. 150-250 mg), 0.5-2.0 equiv. of metal elements was added and ground together under argon atmosphere until the powder becomes greyish and sticky on an agate mortar. Then the mixture was pelletized and sealed in an evacuated (10.sup.3 torr) silica tube. The sealed mixture was heated to 350 C. at a rate of 300 C. h.sup.1 and annealed for 2-4 h. Finally the sealed mixture was gradually cooled in a furnace to afford the mixture containing oA-La.sub.4O.sub.4S.sub.3 (See FIG. 3c for its XRD).

The reaction with Ag:
1.0 equiv. of Ag powder (Aldrich, 2-3.5 m, 299.9%) was added. Annealing: 4 h Little excess (1.1 equiv.) of Ag nor prolonged/repeated thermal treatments did not lead further consumption of La.sub.2O.sub.2S.sub.2.
The reaction with Ni:
0.5 equiv. of Ni nanopowder (Aldrich, <100 nm, 99%) was added. Annealing: 4 h Prolonged and repeated thermal treatments did not improve the yield of oA-La.sub.4O.sub.4S.sub.3 but ended up with the partial decomposition into hP-La.sub.2O.sub.2S