Kristen Erk
Instrumental Analysis
Dr. Vinson

                                 An Analysis of a Rapid and Reliable Analytical Method
                                                 Published in the Early 1990's
 

        In 1991, an article was published on a new method that had been developed for the determination of chlorine, phosphorous, and sulfur in flours of grains and legumes.  The method used wavelength dispersive x-ray fluorescence spectrometry (WDXRFS) to analyze wheat, barley, maize, rice, field bean, and soybean flours.  It had proven to be especially useful for the analysis of chlorine and sulfur because these substances could not be determined directly by atomic absorption techniques.  In addition to being accurate and precise, this new method also had several benefits over the old, commonly used techniques.1
        The accuracy of this method was affected by problems from matrix heterogeneity, particle size, and surface roughness.  Although these problems were minimized because agricultural materials are formed primarily by light elements, the main problem was still the nature of the matrix.  One solution to this problem was to obtain a standard additions graph on a given sample and use the line from this graph to calibrate that type of sample and some other samples with similar matrixes.1
         One of the disadvantages of the classical wet methods for the determination of chlorine, phosphorous, and sulfur in flours was that they had poor precision if the analyte concentrations were low (precision usually decreases at low concentrations).  This problem was overcome by reverting to the WDXRFS method, as it was shown to yield good percent precision values when mixed synthetic standards were not used.  As part of the experiment, two studies were done to show that when mixed synthetic standards were used with the new WDXRFS method, low precision levels were obtained.  In the first study, a pellet of each sample was analyzed eight times in order to determine the instrument stability.  In the second study, separate analyses were done on eight pellets of each sample in order to determine the sample precision.  Excellent precision was obtained for chlorine, phosphorous, and sulfate in both studies.  Therefore, it was shown that when mixed synthetic standards were not used in the new WDXRFS method, good percent precision values were obtained.1
        The standards for this new experimental method were prepared by mixing solutions containing chloride, phosphate, and sulfate with 2 g of “exactly weighed” dried powdered sample.  Clearly, accuracy and precision were tended to in preparing these standards, thereby improving both the accuracy and the precision of this new experimental method.1
         After the standard solutions were prepared for the experiment, standard addition calibration curves were obtained for the chloride, phosphate, and sulfate determinations.  The superb correlation coefficients (ranging from 0.994 to 1.000) obtained from the calibration graphs indicated the reliability of this new analytical process.  Through a study which added known amounts of chloride, phosphate, and sulfate to a fixed weight of sample and then determined the total amount of analyte present, the WDXRFS method was shown not only to be reliable, but also to have an excellent recovery; the recoveries were calculated and found to be almost 100%.1
        As already mentioned, most of the previously used techniques for the determination of chlorine, phosphorous, and sulfur in various flours were wet methods and were not suitable when analyte concentrations were too low.  In other words, these previously used methods were not sensitive enough.  The hypothesis that WDXRFS could provide better sensitivity at low concentrations of analytes proved to be true.1
         Another benefit of WDXRFS was that contamination problems arising from acids were overcome because sample manipulation was minimal in this new technique.  The doubly distilled water that was used in making the reagents for this new method increased the selectivity of the method.  All of the solutions used in this experiment were prepared from analytical reagent grade chemicals (<99.9% pure), which increased the selectivity of the experiment with respect to preparation of the samples.  In addition, to prevent moisture absorption, all of the pellets were stored in a desiccator prior to and in between analyses (a study using nine month old pellets and freshly prepared pellets showed that the pellets remained stable over time).  Therefore, the desiccator also improved this new method by preventing any chemical change of the pellets from occurring over time.1
        In accord with the proposed hypothesis, an additional benefit to this new WDXRFS method was that the total analysis for the quantitation of chlorine, phosphorous, and sulfur required less than twenty minutes per sample.  Most of the usual procedures for such routine determinations were quite time consuming due to a lengthy acid decomposition step.  The WDXRFS technique was advantageous compared to wet chemical analyses because it eliminated the sample dissolution step, thereby shortening the analysis time required by making it possible for the elements to be detected in solid samples.  Time was also saved with the new WDXRFS method because it kept sample manipulation to a minimum.  In addition, after the x-ray program was completed, the fluores cence readings were fitted using a least square regression method.  The equations obtained for each flour were used to analyze different batches of similar sample types, which also contributed to a reduction in the time required for the analysis.1
 All x-ray measurements performed in this new technique were done using a Philips PW 1400 wavelength dispersive x-ray spectrometer.  The spectrometer was interfaced to a Digital PDP 11/23 computer.  The Philips PW 1400 wavelength dispersive x-ray spectrometer is no longer manufactured by the Philips company.  When it was being manufactured in the early 1990's, however, it sold for about $100,000.  Today, the price for a used instrument of this type ranges from $25,000 to $30,000.  The instrument that has replaced the Philips PW 1400 is the Philips PW 2404 wavelength dispersive x-ray fluorescence spectrometer.  The price for this new and improved instrument ranges from $180,000 to $220,000, depending on its configuration and the number of crystals it uses in analyses.2   The price for a Digital PDP 11/23 computer, which the spectrometer was interfaced to in this experimental apparatus, was unavailable.  It, too, was no doubt quite a costly instrument.
        The use of WDXRFS, therefore, as a new technique for the determination of chlorine, phosphorous, and sulfur in flours of grain and legumes, had many benefits over previously used methods.  The new technique and experimental method had been shown to be more accurate and precise, especially at low concentrations of analyte, and therefore, more sensitive.  In addition, it was shown that the method yielded excellent product recoveries, had a higher selectivity than techniques used previously, and was not significantly time consuming.  The major disadvantage of this new technique seemed to be the cost of the instrument needed to perform the analyses.  Although the price of the instrument used for such analyses has risen since the early 1990's, even $100,000 (plus the cost of the interfaced computer) is quite a chunk of change!  Some scientists, however, deemed the instrument worth its price because it had many advantages over popular techniques that were used previously.  Compared to the price of the new, technologically advanced instrument that is used today, $100,000 is not so unreasonable!
 

                                                          REFERENCES

1)     Ruiz, T. P.; Cordoba, M. H.; Gonzalez, R. O.  J. Assoc. Off. Anal. Chem.

        1991, 4, 625.

2)     PHILIPS Analytical Instuments.

        http://www-eu.analytical.philips.com/wwa/ (accessed Apr. 15, 1999).

3)     Skoog, D. A.; Holler, F. J.; Nieman, T. A.  Principles of Instrumental

        Analysis, 5th ed.; Harcourt Brace & Company: Philadelphia, 1998;

        Appendix 1.