JOURNAL OF CHEMISTRY Materials Unexpected substitution in the Li 123x Fe x NiPO 4 (0vxv0.15) solid solution. Weak ferromagnetic behaviour Aintzane Gon Äi, a Luis Lezama, a Marõ Âa Isabel Arriortua, b Gaston E. Barberis a,c and Teo ®lo Rojo* a a Dpto. Quõ Âmica Inorga Ânica, b Dpto. Mineralogõ Âa-Petrologõ Âa, F. Ciencias, UPV/EHU, Apdo. 644, Bilbao, Spain 48080. E-mail: qiproapt@lg.ehu.es c Instituto de Fõ Âsica Gleb Wataghin, UNICAMP, Campinas (SP), Brazil 13087-970 Received 30th September 1999, Accepted 22nd November 1999 Li 123x Fe x NiPO 4 (0vxv0.15) solid solution phases have been synthesized by solid state reaction. X-Ray studies on polycrystalline samples show that all phases are isostructural with the parent LiNiPO 4 compound, crystallizing in the orthorhombic system, space group Pnma. The evolution of the lattice parameters with the iron content follows Vegard's law. The IR spectra show split bands for the phosphate groups, in good agreement with the distortion observed in the crystal structure. The EPR spectra for all phases show a broad band centred at zero ®eld. The absence of a signal corresponding to Fe 3z ions suggests the presence of Fe III ± Ni II interactions strongly affected by the zero ®eld splitting of the nickel(II) ions. The magnetic behaviour of the Li 123x Fe x NiPO 4 (0vxv0.15) phases can be described as antiferromagnetic with the presence of weak ferromagnetism below the ordering temperature. The magnitude of the remanent ferromagnetic moment is indicative of the existence of iron clustering in the samples. The FeO 6 octahedra form ®nite chains in which the magnetic moments of the Fe 3z ions are antiferromagnetically aligned with a small canting angle. Introduction The potential of phosphate compounds for optical, catalytic, magnetic or ionic conductivity applications is a driving force in the emergence of materials science. 1 In this realm, the search for solid electrolytes for the design of new lithium batteries is of substantial interest. 2 Lithium phosphates are good candidates in this area since they present the required chemical and thermal stability for technological applications, 3 and in many cases, excellent ionic conductivity. 4±8 Moreover, the phosphate compounds show an enormous variety of structural types 9,10 which favours the study of new families. The presence of magnetic transition metal ions in lithium phosphate materials allows the study of magnetic behaviour in these phases. LiNiPO 4 crystallizes in a three-dimensional structure built by the stacking, along the [100] direction, of nickel phosphate layers. 11 Neutron diffraction studies have shown that the spin system associated with the Ni 2z ions in this compound undergoes a collinear antiferromagnetic ordering at T N ~19 K, with the characteristics of coupled two-dimensional square planes. 12,13 However, the crystal structure of LiNiPO 4 can be considered as tubular with channels along the [010] and [001] directions, which are occupied by the Li z ions. The characteristics of this crystal lattice suggest that the creation of vacancies in the Li positions will generate potential sites for ionic jumping, with expected enhancement of lithium mobility. One method to create such vacancies may be via substitution of Ni(II) ions [ionic radius (i.r.)~83.0 pm] by a non-magnetic ion, with higher ionic charge and similar or smaller ionic radius, such as Al III (i.r.~67.5 pm) or Ga III (i.r.~76.0 pm). However, attemps to carry out substitution with these ions have been unsuccessful. An attempt was thus made here to substitute Ni 2z for a 3d metal ion with chemical and geometrical similarities. A natural candidate was Fe 3z , and a strategy of synthesis was designed to obtain phases of the general formula Li 12x Ni 12x Fe x PO 4 . As high spin Fe III has S~5/2, its magnetic contribution to the lattice might be of interest. As observed with other ions, a mixture of phases was systematically obtained, with the presence of Li 3 Fe 2 (PO 4 ) 3 compound 14,15 in the ®nal product. An increasing proportion of this phase was observed with the amount of iron added. Modi®cation of the stoichiometry, in an attempt to eliminate this impurity in the reaction, led to Li 123x Fe x NiPO 4 phases (x~0.033, 0.067, 0.100, 0.133), in which each inserted Fe 3z ion unexpectedly displaces three lithium cations as result was con®rmed by neutron diffraction studies (see ref. 13). This article reports on the synthesis, structural study and spectroscopic and magnetic properties of the Li 123x Fe x NiPO 4 (0vxv0.15) solid solution, showing the presence of iron clustering in these phases and the existence of an interesting magnetic phenomenon of weak ferromagnetism. Experimental Synthesis The preparation of the Li 123x Fe x NiPO 4 (0vxv0.15) phases was carried out by solid state synthesis. Stoichiometric amounts of the starting materials Ni(NO 3 ) 2 ?6H 2 O, Fe(NO 3 ) 3 ?9H 2 O, (NH 4 )H 2 PO 4 and LiOH?H 2 O were mixed and homogenized in an agate mortar. The resultant mixtures were heated to 300 ³C for 1 h, in order to decompose the nitrates. After a second homogenization in the mortar, the samples were calcined at 800 ³C for 20 h, in air atmosphere. The Li, Ni, Fe and P contents in the obtained products were analyzed by ICP (inductively coupled plasma) analytical emission spectroscopy. The experimental results for each sample and the theoretical values corresponding to the proposed formulae are given in Table 1. Physical measurements Analytical measurements were carried out by ICP-AES analysis, with an ARL 3410zICP instrument with Minitorch equipment. IR spectra were measured with a Nicolet FT- J. Mater. Chem., 2000, 10, 423±428 423 This journal is # The Royal Society of Chemistry 2000 Published on 19 January 2000. Downloaded by UNIVERSIDAD ESTADUAL DE CAMPINAS on 30/06/2015 15:53:57. View Article Online / Journal Homepage / Table of Contents for this issue