Today, one of the main challenges of lithium-ion batteries is to have the ability to charge faster and to provide high power during longer periods, thus allowing a larger development of intermittent electric power sources [1]. The actual negative electrode materials (graphite and silicon mainly) cannot sustain such high rate capabilities. In this work, we are therefore investigated some materials of the plentiful family of niobium-based oxides. A presentation of old and interesting structures will be briefly reviewed. As an example, the case of KTiNbO5 will be discussed and proposed as efficient negative electrode materials for next generation of lithium ion batteries.The layered KTiNbO5 with a two-dimensional framework represents a playground to produce a large range of closely related phases.[2] The protonated HTiNbO5 analogue is obtained by ion exchange in acidic solution and it preserves a layered structure with a smaller interlayer distance. After dehydration of HTiNbO5 at 400°C, the so-obtained Ti2Nb2O9 phase displays a 2D arrangement with empty channels unlike tunnels of H(K)TiNbO5.In this work, all the synthesized phases were studied as negative electrode materials in lithium-ion batteries. Tested in 1M LiPF6 in EC/DMC between 1.0V and 3.0V vs Li/Li+, these phases have shown different electrochemical behaviors. When HTiNbO5 exhibits a typical plateau during the charge/discharge experiment corresponding to a biphasic phenomenon, lithiation of Ti2Nb2O9 is governed by a solid—solution mechanism. For a better understanding of the charge storage mechanism, we have combined electrochemical experiments with in situ XRD measurements. We have shown that multielectron redox and corner/edge sharing system of Ti/Nb octaedra are at the origin of an interesting capacity of more than 100 mAh.g-1 at a rate of 0.2 A.g-1. A good observed cyclability (< 500 cycles) is in accordance with in situ XRD results showing a reversible behavior of the structure during cycling. Another synthesis method, using Chimie Douce techniques, leads to nanoparticules for both HTiNbO5 and Ti2Nb2O9 and allows an increase of the electrochemical performance of these materials. References : [1] Choi, C. et al, Nature Rev. Mat. 5, 5-19 (2020).[2] A. D. Wadsley, Acta Cryst. 623, 17-41 (1964).