It is very important in the quantitative X-ray fluorescence (XRF) analysis and the mathematical correction for interelement effects to increase the accuracy of the theoretically calculated relative intensities (R~i). In this paper, the influences of several key factors on R~i were investigated. Effect of arbitrary change of the spectrometer geometry on R~i was studied. A comparison was made when using three different X-ray tube primary spectral distribution sources to calculate R~i of some analytical lines. It was observed that there were notable differences within the four commonly used compilations of mass absorption coefficients (MAC) or between calculated and experimental MAC values when the energy of the incident X-ray lines fell in the regions of L~Ⅱ-L~Ⅲ or M~Ⅳ-M~Ⅴ absorption edges of an absorber. A series of R~i were calculated based on a set of simulated fused samples and compared when using the four MAC compilations. The results showed MAC selection was more important for improving the accuracy of the calculated relative intensities. The comparison of the four different compilations of the excitation factors indicated that the average relative differences were 2.6% for K~α, 40.7% for L~α1 and 131.8% for M~α1.

Key words: X-RAY FLUORESCENCE SPECTROMETRY, THEORETICAL CALCULATION, INTENSITY, ABSORB COEFFICIENT, FUNDAMENTAL PARAMETER METHOD, SPECTRAL INTENSITY, INTENSITY DISTRIBUTION