A detailed theoretical study is presented of the noise sources present in selective spectral line modulation (SLM) atomic emission spectrometry. Shot and flicker noise terms associated with the signal and background radiation, the modulating-reservoir transmission, the modulating-reservoir atom emission, the modulating-atom absorption, and the readout system are considered and compared. The results of the study reveal that shot noise never limits the precision of an SLM experiment. Instead, precision at high signal levels is limited by flicker noise associated with the modulating-atom reservoir transmission, the modulating-atom absorption, and the primary emission signal. In contrast, precision near the detection limit is limited by flicker noise associated with the modulating-atom emission and the background emission from the source and modulating-atom reservoir. The relative magnitudes of these noise sources were found to vary considerably with experimental conditions. Experimental data on four elements (Bi, Mg, Cu, Ag) are presented and compared. General trends of the noise magnitudes with experimental conditions are discussed and findings are extended to other elements that can be determined by the SLM method.
