1 The limit of detection ( c min ) is the most important performance characteristic of test procedures with visual indication. The analysts who develop these procedures evaluate this property in different manners, and the values of c min have not been specified in the majority of procedures. In the list of test systems manufactured by the Russian ecoanalytical association Ekoanalitika in 2004, the limits of detection were specified for only 4 of 55 systems.Previously [1,2], an algorithm was proposed for the reliable evaluation of c min . This algorithm included special steps for checking the adequacy of the experimental material. To test hypotheses on the type of distribution of analyte detection frequencies P ( c ) in the region of unreliable response, numerical methods and the following set of criteria was proposed: χ 2 and Kolmogorov-Smirnov ( λ ) tests, the asymmetry ratio ( ), and the coefficient of excess ( γ 2 ).In this work, we used this approach for evaluating the limits of detection in 12 test systems, which were either described in the literature or developed within the framework of this study. The aim of this work was to find the dependence of c min and other properties of test systems with visual indication on the following factors: the material of the support, the type of the analytical reaction, the conditions of color observation in the test system, and the number of observations. Substrate paper for rapid tests, polyurethane foam, methylsilicic acid, and gelatin film were used as support materials in the test procedures. The complexation of metal ions with inorganic and organic ligands, the formation of associates and double salts, chemisorp-1 Presented at the 2nd All-Russia Symposium "Test Methods in Chemical Analysis" (Saratov, 2004).Ã tion, and redox reactions were used as analytical reactions.Indicator paper for the test determination of cobalt(II) based on its reaction with 2-nitroso-1-naphthol in the presence of diphenylguanidine and NiSO 4 (these reagents enhance the contrast of a color change) is well known [3]. In this study, we extended this reaction to other materials (polyurethane foam and methylsilicic acid) and estimated the detection limit for this reaction in solution.Butylrhodamine was chosen as an analytical reagent for the test determination of 2,4-dichlorophenoxyacetic acid (2,4-D). Butylrhodamine, which was proposed as a reagent for the extraction-photometric determination of 2,4-D [4], forms a bright pink associate with 2,4-D.
EXPERIMENTALAll solutions were prepared using twice-distilled water and reagents and chemicals of analytical grade or better.
Reagent indicator papers.The following five types of reagent indicator paper (RIP) strips were used: RIB-Cadmium-Test [1, 5, 6], RIB-Iron(II)-Test [1, 5, 6], RIB-Iron(III)-Test [5, 6], RIB-Nitrite-Test [5], and RIB-Metal-Test II for the determination of total heavy metals (Cd 2+ , Co 2+ , Cu 2+ , Fe 3+ , Hg 2+ , Ni 2+ , and Zn 2+ ) [1, 2, 5]. Without preconcentration, a RIB strip was immersed in solution for 1-2 s and removed; t...