The accurate and precise determination of Li isotopic composition by MC-ICP-MS suffers from the poor performance of traditional column chromatography. Previously established chromatographic processes cannot completely remove Na in complex geological samples, which is currently interpreted to be a result of Na breakthrough. In this study, Na breakthrough during single-column purification was found to differ between simply artificial Na-containing sample solutions, where a little Na residue was found, and silicate rocks, where a large amount of breakthrough occurred. A revised two-step column purification for Li using 0.5 and 0.3 mol l(-1) HCl as eluents was designed to remove the Na. This modified method achieves high-efficiency Li purification from Na and consequently avoiding high Na/Li ratio interference for subsequent MC-ICP-MS analyses. The proposed method was validated by the analysis of a series of reference materials, including Li2CO3 (IRMM-016, -0.10 parts per thousand), basalt (BCR-2: 2.68 parts per thousand; BHVO-2: 4.39 parts per thousand), andesite (AGV-2: 6.46 parts per thousand; RGM-2: 2.59 parts per thousand), granodiorite (GSP-2: -0.87 parts per thousand) and seawater (CASS-5, 30.88 parts per thousand). This work reports early Na appearance prior to the elution curves in chromatography and emphasises its influence for subsequent Li isotope measurement. Based on the findings, the established two-step method would be more secure than single-column chemistry for Li purification.