Recommended Inorganic Chemicals for Calibration
John R. Moody, Robert R. Greenberg, Kenneth W. Pratt, and Theodore C. Rains Inorganic Analytical Research Division Center for Analytical Chemistry National Institute of Standards and Technology (NIST) (Formerly National Bureau of Standards) Gaithersburg, MD 20899
All analytical techniques depend on the use of calibration chemicals to re late analyte concentration to an instru mental parameter. A fundamental component in the preparation of cali bration solutions is the weighing of a pure chemical or metal before prepar ing a solution standard. The analyst must be assured that the purity, stoichiometry, and assay of the chemical are known. These terms have different meanings, and each has an important influence. This R E P O R T is intended to assist the analyst in the selection and use of chemical standards for instrumental calibration. Purity, stoichiometry, and preparation of solutions for different purposes are discussed, and a critical evaluation of the best materials avail able for each element is presented for use in preparing solutions or calibra tion standards. Information on the
chemical form, source, purity, drying, and appropriate precautions is given in Table I. In some cases, multiple sources or chemical forms are available. Cer tain radioactive elements, the transuranic elements, and the noble gases are not considered. There is a subtle difference between the two types of calibration solutions— those for assay standards and those for matrix matching—commonly used in the laboratory. Assay solutions can be less expensive to prepare than matrixmatched calibration solutions; the lat ter often require the highest available purity for the matrix component in or der to avoid uncertainty in the analyte concentration. Because high-purity materials may be expensive, the ana lyst must balance the cost against at tainable accuracy. Occasionally, there is no way to prepare a calibration solu tion with the required accuracy. By re viewing the principles discussed in this REPORT, analysts can better evaluate the accuracy of their calibration solu tions.
Interpretation of purity claims There are two ways to establish the ele mental assay for a compound. The first way is to obtain the assay of each ele ment directly by a suitable method. Usually these assays are not much
This article not subject to U.S. copyright Published 1988 American Chemical Society
more accurate than ~0.1%. Accurate assays of elements in compounds are not easily obtained. The second meth od is to determine the assay of one or more elements and infer the concentra tion of the last element by subtraction from 100%. The uncertainty in the con centration of the last element is in creased because of the uncertainty from each component element. Sub traction from 100% also does not ac count for species that are not deter mined (e.g., H2O, CO2, and S 0 2 ) . Therefore, the actual assay of an ele ment in a high-purity compound can be lower than that calculated by subtract-
REPORT ing all other constituents from 100%. For metals, the easiest route is to use a spectroscopic or other multielement technique to determine the impurities in the metal. The results are expressed as "total metallic impurities" or "total impurities" and usually are given as an upper limit (e.g., 3) are available at much higher levels of purity than are usually required. Except for cost, there is no reason not to use the purest mate rials. However, there is little benefit in preparing a standard solution that is several orders of magnitude more accu rate than needed. If calibration solu tions are a factor of 10 more accurate than the method for which they are to be used, the uncertainty from the cali bration chemical will be negligible. Matrix matching Some analytical methods require that the analyte standard be prepared in a sample matrix form that simulates the sample insofar as major constituents are concerned. This can complicate matters considerably, because a trace amount of analyte can result from im purities in the major constituents. For example, trying to prepare a 1 /tg/g by weight Cr standard in a 10% by weight Fe solution could be extremely difficult because Cr contamination in the Fe matrix could easily exceed 1 /tig/g by weight. Here, the purity of the matrix determines the achievable accuracy in the analyte element. Because this is a general problem with matrix-matching solutions, materials (oxides, carbon ates, or metals) are identified in the table that are particularly suited as matrix-matching materials. Some of these may not be exactly stoichiomet ric and thus are not recommended for use as assay materials for analyte cali brations. For matrix matching, purity (as mea sured by analyte impurity level) is the single most important factor. Whereas
Dissolution of calibration materials
some uncertainty in the concentration of the matrix elements can be tolerat ed, a significant impurity level for an analyte element is unacceptable. One can predict the necessary purity of the matrix element from the quality of an alyte element data that is required. For example, preparation of a 10 jtg/g V standard in a 10,000 jtg/g Fe matrix with an uncertainty of ±0.1 yug/g V re quires an Fe matrix containing