关键词:
Chemistry
摘要:
Part I The accuracy and precision of the results of any chemical analysis depends on the calibration graph and its associated systematic and random errors. Least squares regression generally treats all data with equal weights. A weighted least-squares fit is an improvement but requires knowledge of the imprecision in each point of the calibration graph. The imprecision is not easy to estimate with high confidence because of the large number of replicates needed. The imprecision depends on the types and magnitudes of the sources of noise. We characterized the noise sources in ICP-OES and UV/Vis and developed a model that effectively predicts the standard deviation of emission and absorption as a function of concentration. Once a model is fit to the data, calibration designs were studied. These designs ranged from one to three decades of response and concentration in order to optimize precision over the entire calibration space for ultraviolet-visible spectrochemical analyses. Different calibration strategies, composed of different concentrations and numbers of replicates, have been evaluated determine the calibration design that minimizes imprecision as measured by the average relative concentration error integrated over the entire calibration graph. A laboratory experiment utilizing potentiometric titrations was created to connect electrochemistry, stoichiometry, equilibria and reinforce acid-base titrations. Students performed a potentiometric titration to determine the initial analyte concentration and reactant concentrations at varying points in the titration in order to determine the solubility product constant of a solid species. Part II Advances in chemistry are highly dependent on the procedures published in peer-reviewed journals. Some chemistry journals require authors to address safety considerations in their manuscripts but others do not. In this study, we examined 726 chemistry journals from 28 publishers to determine if they require the author to mentio