J. MATER. CHEM., 1991, 1(6), 1023-1025 Structure and Conductivity of an Li,SiO,-Li,SO, Solid Solution Phase M. A. K. L. Dissanayake" and Anthony R. Westb "Department of Physics, University of Peradeniya, Peradeniya, Sri Lanka bDepartment of Chemistry, University of Aberdeen, Meston Walk, Aberdeen A59 2UE, UK In the system Li4Si04-Li,S04, phase-diagram studies show the existence of a narrow range of stable y solid solutions, Li4-2pw(Si, -xS,)04: 0.30<x< 0.045. These have high Li+ ion conductivity with maximum values for xz0.40 of 1.32 x S cm-' at 25 "C, rising to 8.5 x S cm-' at 300 "C and an activation energy of 0.80 eV. The y solid solutions are structurally related to orthorhombic y-Li,PO, and contain interstitial Li' ions; for x= 0.40, a=6.187(2) A, b=10.621(3) A, c=5.008(3) A.Keywords: Lithium ion conductor; Lithium silicate ; Lithiiim sulphate ; Solid electrolyte A number of solid solutions based on Li4Si04 are reported to have high lithium ion conductivity. These include: (i) stoichiometric solid solutions with Li,GeO, and Li4Ti04 whose conductivity is increased only moderately at intermedi- ate compositions;',2 (ii) interstitial Li+ solid solutions with Li,A104 and Li,Ga04 which have greatly increased conduc- tivity;,~~and (iii) vacancy solid solutions with Li3X04, X= P,As,V, and Liz ,5(A1,Ga)o .,SO4, which also have greatly increased c~nductivity.~-'~. ?-Tetrahedral structure phases such as Li3X04: X=P,As,V and Li2ZnGe04 also have high Li ion conductivity when they form solid solutions contain- + ing interstitial Li+ The literature on the system Li4Si04-Li2S04 is rather confusing.Shannon et a!., were the first to report high Li' ion conductivity and attributed it to the creation of Li' vacancies in the Li4Si04 structure; they obtained a conduc- tivity of 1 x S cm-' at 300 "C for composition Li, .4(Sio ,7So,3)04. Burmakin and Zhidovinova reported a change-over from the monoclinic Li,Si04 structure to an orthorhombic structure at 30-35% Li2S04, but without any sudden change in electrical properties. ' ' Neutron diffraction studies on composition Li, .4(Si0 .,So.3)04 showed it to be orthorhoinbic and based on the y-Li3P04 structure, but containing interstitial Li+ ions.16 Given these rather conflict- ing results and the fact that the structures of Li,Si04 and Li3P04 are significantly different, we have made a combined phase diagram, X-ray diffraction and conductivity study in order to resolve these differences, and the results are reported here.Experimental Lithium orthosilicate, Li,Si04 was prepared by solid-state reaction of Si02 (high-purity crushed quartz crystal) and Li2C03 (BDH, AnalaR). Completeness of reaction was checked by X-ray powder diffraction using a Shimadzu model XD-7A X-ray diffractometer with Cu-Ka, radiation of wave- length 1.54188,. Mixtures of Li,Si04 and Li2S04 (BDH, AnalaR) in various proportions were heated at 600-700 "C for 12 h in gold-foil boats. Pellets of the reacted mixtures were cold-pressed at 2 tons cm-2 and sintered at 900-1000 "C for 12-24 h.Electrodes made from Engelhard liquid-gold paste were attached to the pellets, whose temperature was gradually raised to 800 "C in order to decompose the organometallic paste and harden the electrodes. The pellets were subsequently placed inside a high-temperature sample holder" which was inserted into a Heraeus tube furnace controlled by a Euro- therm 810 temperature controller, and the conductivity was determined by a.c. impedance measurements over the range 10 Hz-10 MHz with a Hewlett-Packard 4192A LCR meter and HP86B microcomputer. Measurements were made at 25 "C intervals up to 600 "C on both heating and cooling. The signal applied to the sample was 20 mV, and the tempera- ture measured with a chromel-alumel thermocouple placed close to the pellet.For phase-diagram determination, approximate melting temperatures were determined by observing the appearance of pellets during stepwise heating in a muffle furnace. Since it is difficult to observe both the onset and completion of melting by this method, a single 'average' temperature was obtained for each composition which should lie somewhere between the solidus and liquidus temperatures. Phase-tran- sition and melting temperatures for some compositions were also determined using a Stanton Redcroft DTA 675, heating rate 5 "C min-'. Results and Discussion An approximate phase diagram for the system Li,SiO,- Li2S04, constructed using a combination of DTA results, pellet melting studies and room-temperature X-ray diffraction on samples reacted at high temperature, is shown in Fig.1. It contains a narrow range of solid solutions, labelled y, of formula Li, -2x(Sil-xS,)04: 0.30 <x<0.45, which are stable at all temperatures up to melting at 860-1060 "C. There was no evidence of any significant solid solution in either of the end-member phases, Li4Si04 and Li2S04. The y solid solutions appear to belong to the family of y phases, such as the Li3 +,(PI -,Si,)04 solid solutions, in accordance with the neutron diffraction results of composition x=0.3O.l6 Most y solid solutions are based on a parent end- member, such as Li3P04 or Li2ZnGe04, which have an overall cation:anion ratio of unity. In the present y solid solutions, there appears to be no such stoichiometric parent phase although the hypothetical composition x=0.5, outside the experimental solid solution range, would have a cat-ion:anion ratio of unity.Indexed powder diffraction data for one y composition have been tabulated.? The orthorhombic lattice parameters: a=6.187(2) A, b= 10.621(3)A, c= S.OOS(3)8, for x=0.40, are similar to those reported earlier, t Available from the authors on request; to be submitted to the JCPDS File. 1024 x in Li4_2x(Si,-,S,)04 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 liquid 1200p--.. i 1000 800 9.t-600 400 200 Li4sio4 10 20 30 40 50 60 70 80 90 Li2s04 mol Yo Fig. 1 Approximate phase diagram for the system Li,SiO,-Li,SO,.0, samples were single phase/two phase, respectively by X-ray powder diffraction; x ,DTA transition temperatures, heating at 8 "C min-I; 0,approximate melting temperature of pellets a=6.1701 A, b= 10.6550A, c=5.0175 A, for composition x =0.30.16 A.c. impedance data were used, in the form of complex- impedance plane plots, to extract net conductivity values. Generally, two semicircles were seen, one with a capacitance of ca. 2 pF attributable to the bulk resistance, and a second, with a capacitance of 10-20 pF, attributed to grain boundaries of constriction resistance type. l8 At lower frequencies, an inclined spike with an associated capacitance of ca. 5 pF was seen, supporting the idea that conduction is by Li' ions which are polarised at the blocking Au electrodes.Impedance data at three selected temperatures, showing this spike inclined at 65-70" to the horizontal, are given in Fig. 2. Conductivity Arrhenius plots are given in Fig. 3 for three y solid solution compositions; the three sets are fairly similar although x =0.40 has a slightly higher conductivity, 1.32 x 52-' cm-' at 25 "C, rising to 8.5 x lop3R-' cm-' at 300 "C, with an activation energy of 0.80 eV. Conductivity data for the two-phase mixtures, Li4Si04 +y and y +Li2S04, are much lower (Fig. 3) reflecting the fact that the end- members Li4Si04 and Li,S04 have only modest Li' ion 100 200 300 400 z '/Q Fig. 2 Complex impedance plots for three temperatures for Li3 .2(si0 .SSO .4)O4 J.MATER. CHEM., 1991, VOL. 1 2 1 70 I E 7 -1 c Y2 -2 b Y P,2 -3 -4 -5 -6 -0 1.5 2.0 2.5 3.0 3.5 4.0 103K/T Fig. 3 Conductivity Arrhenius plots for Li4-2x(Sil -.S,)O, compo-sitions conductivity. The conductivity data in Fig. 3 were fully revers- ible on cooling. The data at lowest x (0.10) show a change in slope at ca. 290 "C. These data are probably dominated by the majority phase in this sample, Li,SiO,, for which a change in slope at 290 "C has been reported and discussed.' These results give conductivity values similar to, but slightly lower than, those of Shannon et aL3 The difference is probably not significant and reflects the fact that our pellets were not fully densified. Thus, the conductivity data in Fig.3 are net pellet conductivities and the true bulk values are likely to be somewhat higher. The highest conductivities, around x=0.3-0.4, correspond to Li contents of Li3,4 to Li3.2. The value Li3,4 is close to the optimum observed in the systems, Li4Si04-Li3X04: X =P,As,V, i.e. Li3 ,4(Si0 ,4Xo &)4.8.9.12 Shannon et aL3 believed their highly conducting compo- sition, x=0.30, to be a vacancy solid solution based on Li4Si04, whereas our results, in agreement with Fitch et a1.,16 indicate the structure to be rather an Li' interstitial solid solution based on the y-Li3P04 structure. This misassignment is readily understandable since there is considerable similarity between the X-ray powder patterns of Li4Si04 and the y phase. Thus it appears that Burmakin and Zhidovinova,' while recognising the changeover from monoclinic to ortho- rhombic symmetry with increasing x, were unable to recognise the two-phase region of Li,Si04+y at low x values.We thank the British Council for financial support, the International Programmes in the Physical Sciences, Uppsala University, Sweden for providing equipment and training for the Solid Electrolytes research at Peradeniya, P. W. S. K. Bandaranayake and C. N. Wijayasekera for research assist- ance, R. P. Gunawardane for helpful discussions and SERC for research support (A. R. W.). References 1 A. R. West, J. Appl. Electrochem., 1973, 3, 327, 2 I. M. Hodge, M. D. Ingram and A. R. West, J. Am. Ceram. SOC., 1976, 59, 360. 3 R. D.Shannon, B. E. Taylor, A. D. English and T. Berzins, Electrochim. Acta, 1977, 22, 783. 4 K. Jackowska and A. R. West, J. Mater. Sci., 1983, 18, 2380. J. MATER. CHEM., 1991, VOL. 1 1025 5 P. Quintana, F. Velasco and A. R. West, Solid State Ionics, 1989, 12 A. R. Rodger, J. Kuwano and A. R. West, Solid State Ionics, 34, 149. 1985, 15, 185. 6 7 Y. Saito, T. Asai, K. Ado, H. Kageyama and 0. Nakamura, Solid State lonics, 1990, 4/41, 34. Y. W. Hu, I. D. Raistrick and R. A. Huggins, J. Electrochem. SOC., 1977, 124, 1240. 13 14 15 16 H. Y.-P. Hong, Muter. Res. Bull., 1978, 13, 117. W. H. Baur, Inorg. Nucl. Chem. Lett., 1980, 16, 525. P. G. Bruce and A. R. West, J. Solid State Chem., 1982, 44, 354. A. N. Fitch, B. E. F. Fender and J. Talbot, J. Solid State Chem., 8 9 A. Khorassani and A. R. West, Solid State Ionics, 1982, 7, 1. A. Khorassani and A. R. West, J. Solid State Chem., 1984, 53, 17 1984, 55, 14. M. A. K. L. Dissanayake and M. A. Careem, J. Phys. E., 1988, 369. 21, 1203. 10 A. Garcia, G. Torres-Treviiio and A. R. West, Solid State Zonics, 18 P. G. Bruce and A. R. West, J. Electrochem. SOC., 1983,130, 662. 1990, 40/41, 13. 11 E. I. Burmakin and S. V. Zhidovinova, Russ. J. Znorg. Chem., 1980, 25, 1108. Paper 11025538; Received 30th May, 1991