Probe—detektor-anordnungen zur Kernstrahlungsmessung für Analytische Zwecke

Abstract

The “analytical power” of detection systems is investigated, considering typical arrangements in analytical terms. From N atoms of interest in a sample Z counts are drawn: Z=R N (R registration yield). Crude data ZM contain also background U generated by interferences: ZM=Z+U. Identification and separation of Z is accompanied by errors ~σ_MU^1/2 due to imperfect distinguishability and to statistical error (σ_M model inherent factor). With “selectivity”=σ_M ^- (R/R_interference)^1/2 the figure of merit is G=selectivity efficiency/interference)^1/2. Sample background dominating, “interference” or both “interference” and “efficiency” have to be cancelled. Present semiconductor spectrometry of low energy γ-quanta and of large rates is not very far from its limits. Further progress will be obtained by enhancement of high energy efficiency and reduction of natural background.