To enhance the kinetics of poorly conducting cathode materials for Li batteries, the authors have proposed a number of strategies based on crushing the active material into nanopowder and embedding the powder into a carbon-based web or coating. Using the well-elaborated example of LiFePO4, we demonstrate that the same goal can be achieved with a different approach where the active material remains in a form of large (I 20 gm) single crystals. Instead of crushing the material, we make it porous-with average pore size around 50 nm and pore surface area of 25 m(2)/g. The walls of the pores (but also the outer surfaces of crystals) are covered with ca. 1-nm-thick carbon film. Most surprisingly, such a unique nanoarchitecture can be prepared using a simple sol-gel based procedure including a single heat treatment. The crucial part is the selection of appropriate carbon precursor, For example, citric acid decomposes quite vigorously into gases and solid carbon at temperatures up to ca. 450 degrees C. This range matches exactly the first solidification of LiFePO4. Thus, the evolving gases can create an interconnected web of pores while the solid parts (carbon) are deposited simultaneously on the walls of pores. We further show that a carbon content of less than 3% is already sufficient for surpassing the percolation threshold with respect to surface conductivity of carbon. Using more carbon can decrease the rate performance so a fine balance is required in this respect. Most importantly, carbonization at a temperature of slightly less than 700 degrees C is sufficient to achieve a composite conductivity of the order of 10(-2) S cm(-2)-more than sufficient for good cathode kinetics. In the end, we show new evidence that the phase that is responsible for high conductivity of LiFePO4-C composites is indeed the carbon phase. (c) 2006 Elsevier B.V. All rights reserved.