112 Analyst, February, 1967, Vol. 92, $9. 112-114 The Determination of Calcium in Biological Samples by X-Ray Fluorescence BY K. P. CHAMPION AND R. N. WHITTEM (Australian Atomic Energy Commission Research Establishment, Sutherland, New South Wales, Australia) An X-ray fluorescence method is described for the rapid determination of calcium in ashed biological samples. Samples are digested in nitric acid, and calcium is determined in the extract. The results are comparable with those obtained by a conventional titrimetric method. THE determination of calcium in biological samples, such as ashed milk, vegetation, oysters, etc., by an X-ray fluorescence technique is complicated by the presence of high levels of potassium and chlorine, both of which exhibit strong absorption effects for calcium K, radi- ation.Methods which might be used to overcome this difficulty include- Lithium borate fusion technique (as used by Rose, Adler and Flanaganl for siliceous rocks)-This has the disadvantage that the buffer used (lanthanum oxide) does not materially assist in overcoming the potassium and chlorine interference, so that each of these elements must be determined and corrections made to the calcium intensity. Thus the analysis becomes quite complex. Internal standard methods-The best internal standard line2 is tin L,, and this offers good correction for absorption effects. However, in the Philips spectrometer it is not sufficiently resolved from the calcium K, line to be useful over a reasonable range of calcium concentrations, unless the tin content is varied to suit each sample type.Further, particle-size effects3 are likely to be important in this wavelength region, so that thorough grinding and mixing of the internal standard material are essential. Aqueous solzdion methods4-These appear attractive as particle-size effects have no significance, and also because accurate dilutions can be made, thus reducing absorption effects to acceptable levels, provided, of course, that adequate sensitivity is available. The use of aqueous solutions was therefore investigated, and a method was developed for the rapid determination of calcium in the ash of milk, oyster shells and flesh, and of vegetable matter. EXPERIMENTAL APPARATUS- The apparatus and experimental conditions used are as follows. Spectrometer” .. X-ray tube . . Detector* .. Crystal . . . . Collimator . . Pulse height analyser Counting strategy Sample holder . . .. . . Philips PW1520 . . . . . . Sealed proportional counter, PW1965/10 . . . . Lithium fluoride .. . . 320-p spacing .. . . . . Fixed counts, generally 10,000 . . . . Chromium target, type 25718/61, operating a t 40 kV, 20 mA . . . . Set to 90 per cent. of Ca I<, counts Plastic holder PW1527/10, fitted with 6 - p thick “Melinex” window * If available, the all-vacuum instrument, PW1540, with flow proportional counter, offers greater sensitivity. SAMPLE PREPARATION- Weigh between 0.25 g and 1 g of sample (w,), depending on expected calcium content, and place it in a 100-ml beaker. Add 5 ml of analytical-reagent grade nitric acid (sp.gr. 1-42), and evaporate it on a hot-plate to about 1 ml.Allow to cool, add about 30 ml of water, cover the beaker, and boil the solution for 5 minutes. Cool and filter the solution into a lOO-ml,CHAMPION AND WHITTEM 113 weighed calibrated flask. Wash the beaker and filter-paper three times with 10 ml of hot 0-5 N nitric acid. Make up to the mark with water, cool, mix, and weigh (net weight = W,). PREPARATION OF STANDARDS- Dissolve 2.497 g of analytical-reagent grade calcium carbonate by adding dropwise dilute nitric acid (1 + 2) and dilute with water to 1 kg. This stock solution contains 0-1 per cent. w/w of calcium and lower concentration working standards can be made by dilution. In general, only one working standard is necessary, as linear calibrations in the range 0 to 0.1 per cent. w/w are obtained.PROCEDURE- Fit the sample holder with a "Melinex" window (6 p thick) that has been well rinsed to remove calcium contamination. Pour about 5 ml of sample solution into the cell and measure the time for 10,000 counts. If the counting time exceeds 200 seconds, errors may occur owing to the formation of bubbles on the window surface. If this happens, stop the count, refill the cell, and then resume the count. Measure the count-rate of a suitable Concentration of standard, and also the count-rate of water as a blank. Set the spectrometer to the calcium K, line at 3-360 A. CALCULATIONS- equation- Calculate the percentage of calcium in the original ash sample from the following ?%x - %, w, Percentage of calcium = C, - .- n S - nB w X where C , is the concentration of the standard expressed as per cent.w/w, W , is the weight of the sample solution, wx is the weight of sample taken, "rzx is the count-rate of the sample, 12, is the count-rate of the standard, and n, is the count-rate of the blank. The standard deviation of the result can be simply calculated from the count-rates and the number of counts accumulated for sample, standard and blank5 RESULTS AND DISCUSSION A selection of samples was analysed by the present method, and also by a conventional chemical method. Some difficulty was experienced in choosing a chemical method that would suit the range of samples, as co-precipitation of many elements frequently occurred in conventional oxalate methods. The A.O.A.C. method 20.027,6 modified so that the phos- phate separation was performed at pH 4.4 instead of pH 4, was adopted.This modification eliminated the co-precipitation of zinc from the oyster samples. The results are shown in Table I and indicate excellent agreement between the two methods. The result on the NBS limestone indicates the accuracy. TABLE 1 COMPARISON* OF X-RAY FLUORESCENCE AND CHEMICAL RESULTS X-ray fluorescence method, Sample per cent. w,/w Limestone NBS l a t . . . . 29.4 Grass ash 64/221 .. .. 2.S5 Grass ash 65/33 . . . . . . 6-02 Milk ash 65/76 . . . . . . 15.1 Oyster-flesh ash 65/29 . . . . 10.5 Milk ash 65/79 . . . . . . 15.5 Oyster-flesh ash 65/131 . . . . 5.s3 Chemical method, per cent. w/w 29.3 2.82 5.97 15.3 15.4 10.5 5-90 * Based on the means of four analyses.t Certificate analysis 29.52 per cent.114 CHAMPION AND WHITTEM The speed of analysis is worthy of comment. Sample preparation takes about 4 hours for a batch of 10 to 20 samples. Instrument time is about 2 minutes per sample (with the PW1540 instrument, counting times of about 10 seconds are adequate). It should be stressed that it is essential to make up the sample solutions on a weight basis, as their specific gravities generally lie in the range 1.02 to 1.05. Thus, if a weight per volume technique is used, systematic errors of 2 to 5 per cent. can occur. Recent work with the all-vacuum instrument PW1540 has shown that this method can be extended to the simultaneous determination of potassium, chlorine and sulphur. We gratefully acknowledge the help of Mrs. B. McAllister for the titrimetric analyses. REFERENCES 1. 2. 3. 4. 5. 6. Rose, H. J., Adler, I., and Flanagan, P. J., Appl. Spectrosc., 1963, 17, 81. von Hevesy, G., “Chemical Analysis by X-rays and its Applications,” McGraw-Hill, New York, Claisse, F., and Samson, C., Adv. in X-ray Analysis, 1962, 5, 335. Gunn, E. L., A.S.T.M. Special Technical Publication No. 349, 1964, p. 70. Mack, M., and Spielberg, N., Spectrochim. Acta, 1958, 12, 169. Horwitz, W., Editor, “Official Methods of Analysis of the Association of Official Agricultural Chem- ists, ” Ninth Edition, Association of Official Agricultural Chemists, Washington, D.C. ~ 1960, Method 20.027, p. 268. Received November 22nd, 1966 1932, p. 161.