In this work, High-Performance Liquid Chromatography (HPLC) separation followed by Thermal Ionization Mass spectrometry (TIMS) measurements was conducted to measure Am, Cm and Nd isotope ratios in irradiated samples. The method validation was performed on a natural Nd standard (JNdi-1 reference material). The literature JNdi-1 reference values are calculated using conventional internal normalization. However, in nuclear science, such normalization is not possible as no isotope ratio can be considered as a reference. The results from measurements were compared to reference values using the Exponential Mass Fractionation Law, which is a powerful tool to detect drifts. The results show that the dominant bias comes from isotope fractionation which can be minimized using the TE method. The acquired data suggest the following absolute composition for the JNdi-1 reference standard: ¹⁴²Nd/¹⁴⁴Nd = 1.13950(47), ¹⁴³Nd/¹⁴⁴Nd = 0.51163(30), ¹⁴⁵Nd/¹⁴⁴Nd = 0.34876(17), ¹⁴⁶Nd/¹⁴⁴Nd = 0.72322(65), ¹⁴⁸Nd/¹⁴⁴Nd = 0.24264(46) and ¹⁵⁰Nd/¹⁴⁴Nd = 0.23789(68) (k = 2). These values suggest the need to revisit the agreed upon reference values when the Nd absolute isotopic composition is required. The conditions of HPLC separation were also optimized to obtain purified fractions of Nd, Am and Cm in one experiment using a cation exchange column. Separations were achieved on a LUNA SCX column and using 2-hydroxy-methylbutyric acid (HMB) as an eluent introduced in the gradient mode. A difference below 0.03% was observed between data obtained with separation and without separation. Compared to previous work that required three separations using two different HPLC conditions, improvements were made in terms of both duration and the amount of radioactive effluents. The methodology was then successfully applied to irradiated discs coming from MARIOS and DIAMINO analytical irradiation.