摘要:

Near-infrared spectrophotometric analysis is a rapid technique that typically uses the reflectance of a solid sample at several wavelengths to determine the sample's composition. A computerized modeling process is generally used to correct for background and sample-matrix interferences. The modeling process employs a training set of samples to, in effect, “teach” the computer to recognize relationships between minute spectral features and sample composition. The contents of the training-set samples must be determined initially by some other reference method before applying the near-IR technique. The model developed from near-IR spectra and reference values gives the sample composition using a number of linear equations. Each of these equations expresses a particular component concentration as a weighted sum of the signals observed at a number of near-IR wavelengths. Instruments used for near-IR spectrophotometry can be as simple as a filter photometer or a grating monochromator. The broad spectral peaks and highly correlated wavelength vectors generally limit the number of wavelengths used in the model. Little or no sample preparation is required by near-IR methods, and many solid samples can be directly analyzed. Near-IR spectrophotometry has found application in agriculture, industry, biology, medicine, and even satellite remote sensing.

ISSN:1040-8347    

DOI:10.1080/10408349108051642

出版商:Taylor & Francis Group    

年代:1991

数据来源: Taylor

摘要:

In spite of the great improvement in selectivity and sensitivity of modern analytical instrumentation, the classical preconcentration and separation techniques, such as liquid-liquid extraction, ion-exchange chromatography, precipitation, etc., are still widely used to improve analytical measurements, especially in connection with trace and ultratrace analysis. Liquid-liquid extraction is among the most frequently used conventional methods for sample pretreatment preceding measurements of the analyte concentratioin. It offers a potential for selectivity and sensitivity improvement in a large number of determinations.

ISSN:1040-8347    

DOI:10.1080/10408349108051643

出版商:Taylor & Francis Group    

年代:1991

数据来源: Taylor