ANALYST, DECEMBER 1986, VOL. 111 1375 Rapid Atomic Spectrometric Determination of Sodium, Potassium, Calcium and Magnesium in Powdered Milk by Direct Dispersion Miguel de la Guardia Departamento de Quimica Analitica, Facultad de Quimicas, Burjasot, Valencia, Spain Amparo Salvador Departamento de Didactica de las Ciencias Experimentales, Escuela Universitaria del Profesorado de Educacidn General Basica, Valencia, Spain and Pilar Bayarri and Rosaura Farre Departamento de Bromatologia, Toxicologia y Quimica Analitica, Facultad de Farmacia, Valencia, Spain A method for the determination of Na, K, Ca and Mg in powdered baby milk formulas is proposed. The procedure consists in weighing 0.1 g of milk, adding 10 ml of a solution of lanthanum (4% mlV) and dispersing the mixture in water to give a volume of 100 ml.The samples are then directly introduced into an air - acetylene flame and Na, Ca and Mg are determined by atomic absorption spectrometry using multi-element standards containing La. K is determined by atomic emission spectrometry. Keywords: Direct dispersion atomic spectrometry; powdered baby milks; sodium, potassium, calcium and magnesium determination The concentrations of alkali and alkaline earth elements present in powdered baby milk formulas must be kept within narrow limits in order to satisfy infants' nutritional require- ments and to avoid physical disorders. Several international organisations , including ESPGAN' and the Codex Alimen- tarius Mundi,2 recommend minimum and maximum limits for sodium, potassium, calcium and magnesium in powdered milk, and these are used to set legal limits on element concentrations. It is common practice, both in industry and in quality control laboratories, to determine these four elements when analysing dairy formulas. The determinations are often carried out in liquid and powdered baby milk formulas by the destruction of organic material, either by wet digestion3-5 or by dry ashing.6 Other methods that do not require this destruction process have been proposed for liquid milk, e.g., protein precipitation and separation ,3,7-9 or dilution followed by direct analysis4J0,11.No such alternatives have been suggested for powdered milks, probably because they are not totally soluble in water. However, methods that utilise either flame emission or atomic absorption spectrometry have been proposed for the direct determination of elements in solid samples in suspension other than powdered milk,l2-15 although L'vov16 has pointed out that the direct determination of solids by atomic absorption spectrometry is impossible because of the characteristics of conventional nebulisers. The aim of this study was to compare the results obtained when sodium, potassium, calcium and magnesium in pow- dered milk are determined using atomic absorption and flame emission spectrometry, after ashing or dispersion in water. A series of samples was analysed by two different analysts in different laboratories.Experimental Apparatus A Pye Unicam SP 1900 atomic absorption spectrometer equipped with Na, Ca and Mg hollow-cathode lamps was used for the test method and a Perkin-Elmer 2380 atomic absorp- tion spectrometer with Na, K, Ca and Mg hollow-cathode lamps was used for the reference method.Reagents Test method Sodium, potassium, calcium and magnesium standard solu- tions, 100 p.p.m. Prepared from NaCl (Probus), KCl (Merck), CaC03 (Probus) and Mg metal (Merck). Concentrated nitric acid. From Probus, d = 1.38. Lanthanum solution. Prepared from La203 (Phaxe), 4% NaCl, 2% mlV aqueous solution. Multi-element standards containing sodium (0.5-4 p.p.m.), potassium (0.5-10 p.p.m.), calcium (0.5-10 p.p.m.) and magnesium (0.02-1.5 p.p.m.) with 0.4% mlvlanthanum were prepared from these solutions. Genapol PF 10 surfactant. Condensed from ethylene oxide and propylene oxide (Hoechst). A-11 milk sample (powder).Certified by the International Atomic Energy Agency. Commercial samples of powdered milk. mlV, and from La203 (Merck), 5% mlV. Reference method Sodium, potassium, calcium and magnesium standard solu- tions, 100 p.p.m. Prepared from NaCl (Merck), KCI (Merck), CaC03 (UCB) and Mg metal (Panreac). Lanthanum solution, 5% mlV. Prepared from La203 (Merck). NaCl, 2% mlV aqueous solution. Concentrated nitric acid. From Probus, d = 1.38. The following standards were prepared from the stock solutions: sodium (0.5-4 p.p.m.), potassium (0.5-2 p.p.m., containing 2000 p.p.m. of sodium in the form of NaCl), calcium (1-10 p.p.m.), magnesium (0.2-1 p.p.m., containing 0.1% lanthanum). A-11 milk sample (powder). Certified by the International Atomic Energy Agency. Commercial samples of powdered milk.General Procedure Test method Tests were carried out to determine whether commercial surfactants encouraged the dispersion of the samples in water1376 ANALYST, DECEMBER 1986, VOL. 111 Table 1. Instrumental parameters Intensity of Angle of Flow-rate Flow-rate Wavelength/ lamp current/ Slit width/ burner/ Burner height/ of C2H2/ of air/ Element nm mA mm degrees cm min min Sodium . . . . . . 589.6 6 0.15 30 0.8 1 5 Potassium . . . . . . 766.5 - 0.15 90 0.8 1 5 Calcium . . . . . . 422.7 6 0.11 0 1 0.8 5 Magnesium . . . . 285.2 3 0.15 10 0.6 0.9 5 Table 2. Effect of the addition of surfactants on the determination of sodium, potassium, calcium and magnesium in powdered milks With surfactant Without surfactant Element X S t x* SS Sodium .. . . 0.26 0.006 0.27 0.006 Potassium . . 0.73 0.006 0.77 0.006 Calcium . . . . 0.48 - 0.49 0.006 Magnesium . . 0.45 0.006 0.46 0.006 * k = Mean value of the measured concentration in three independent analyses expressed as grams of element per 100 g of milk powder. t s = Standard deviation for three independent assays. between the slopes of the calibration graphs obtained with multi-element and with single-element standards of +8.6% for sodium, +0.9’/0 for potassium, +1.1% for calcium and -4.3% for magnesium. These differences can be reduced, especially for sodium and calcium, by reaching a compromise between the best sensitivity and the best agreement between absorbance values for each element in single and multi- element standard solutions. The selection of the optimum experimental conditions is based on this criterion.The burner was placed at an appropriate angle in order to reduce the sensitivity of the readings and increase the dynamic range, thereby avoiding excessive dilution of the samples and making it possible to measure the four elements in the same sample aliquot. Table 1 shows the instrumental conditions selected. Table 3. Results obtained for the analysis of a certified sample of the International Atomic Energy Agency’s A-1 1 powdered milk reference sample Certified Reference Element value, YO Test method, YO method, YO Sodium . . 0.442 k 0.033 0.42 f 0.03 0.44 Potassium . . 1.72 f 0.10 1.70 f 0.04 1.75 Calcium . . 1.29 rt 0.08 1.27 k 0.04 1.31 Magnesium 0.110 rt 0.008 0.110 f 0.004 0.10 and the conditions for using the multi-element standards were also studied.In view of the results obtained, the following procedure was used in further work. After weighing 0.1 g of milk powder, 10 ml of a 4% m/V solution of lanthanum were added and the mixture was dispersed in water to a total volume of 100 ml. The samples were then introduced directly into the air - C2H2 flame, and the absorption of sodium, calcium and magnesium and the emission of potassium under instrumental conditions estab- lished previously were measured using the multi-element standards containing sodium, potassium, calcium and mag- nesium in appropriate proportions and 0.4% m/V of lan- thanum. Reference method Weigh 1-2 g of a milk powder in a porcelain crucible, heat with a Bunsen burner flame until ignition and ash in a muffle furnace at 500°C for 24 h. Dissolve the residue using the minimum volume of concentrated nitric acid and dilute to 100 ml.Take 5 mi from this solution and dilute it to 100 ml, adding NaCl for the K determination, or La for the Mg determina- tion. Results and Discussion Instrumental Parameters Single-element standards of each of the elements under study were prepared in the presence of 0.4% lanthanum as an ionisation buffer and were compared with standards at the same concentration that contained all the elements con- sidered. Under the instrumental conditions that give the maximum sensitivity for each element, we found relative differences Use of Surfactant In order to achieve a better dispersion of the milk samples in water, and to guarantee stability during the assay, we added surfactants to the mixture.A non-ionic surfactant, Genapol PF 10, was chosen because it does not contain any of the elements under consideration. The sodium, potassium, calcium and magnesium content in a sample of powdered milk was determined using the method described under General Procedure, and the results were compared with those obtained by adding 0.75% of the surfactant to samples and standards. As can be seen from Table 2, the values obtained with the two procedures are similar, and it can therefore be concluded that it is not necessary to add surfactants in order to guarantee thorough dispersion of the samples. Analysis of Certified Sample In order to evaluate the precision of the two procedures used in this work, a determination of the concentration of the four elements in a certified sample of the International Atomic Energy Agency’s A-11 powdered milk was carried out.The results obtained are summarised in Table 3, from which it can be seen that the two procedures give analogous results and are in agreement with the reference values if the precision of the determination is taken into account. Comparison of Test Method with Reference Method The content of sodium, potassium, calcium and magnesium in powdered maternal milk and in other powdered milks was determined. The data given in Table 4 indicate that the results obtained from the two methods are similar, neither method yielding systematically lower values than the other. Table 5 shows the results obtained expressed in mg kcal-1.When these data are compared with the maximum and minimum concentrations established by legislation in several different countries (Table 6) and with the limits recommended by ESPGAN and the Codex Alimentarius Mundi (Table 7), it is observed that the values obtained by either procedure in most of the samples are within the legally permitted limits.ANALYST, DECEMBER 1986, VOL. 111 1377 Table 4. Comparison of determinations of sodium, potassium, calcium and magnesium by direct analysis (test method) and by analysis following incineration of samples (reference method). Results are expressed as the means of three determinations k standard deviation. Samples 5 and 6 correspond to dietetic milks that are not subject to the same laws as maternal milk. These data are given to show that the procedures are applicable to different milk samples Sodium Potassium Calcium Magnesium Reference Test Reference Test Reference Test Reference 1 2.70 f 0.06 2.60 k 0.04 7.7 k 0.1 7.40 k 0.08 4.90 IL 0.06 5.00 k 0.08 0.460 f 0.006 0.420 f 0.007 2 2.90 3.0 k 0.2 8.40 f 0.06 8.00 k 0.07 6.1k0.2 6.10+0.05 0.560k0.006 0.53k0.02 3 2.50 k 0.06 2.6 f 0.3 6.60 k 0.06 6.2 t 0.3 3.8 3.9 k 0.2 0.450 0.41 t 0.02 4 2.60k0.06 2.60k0.01 8.60k0.06 8.0 5.80 k 0.06 6.13 0.530 k 0.006 0.47 t 0.007 5 9.7 k 0.3 11.76 t 0.04 22 * 2 20.1 k 0.6 8 .2 f 0.2 9.5 + 0.08 2.96 k 0.03 3.27 i- 0.06 6 8.8 k 0.1 11.3 t 0.1 20.1 +- 0.5 18.1 + 0.3 8.10 _t 0.09 9.4 k 0.08 2.854 0.004 3.20 i- 0.01 Sample Test Table 5. Comparison of test and reference methods (results expressed in mg per 100 kcal) Sodium Potassium Calcium Magnesium Sample Test Reference Test Reference Test Reference Test Reference 1 54 52 154.0 148.0 98 100 9.2 8.4 2 60.4 62.5 175.0 166.6 127.1 127.1 11.6 11.0 3 48.1 50.0 126.9 119.2 73.1 75.0 8.6 7.9 4 57.5 57.5 190.3 177.0 128.3 135.6 11.7 10.4 ~ _____ _ _ _ _ ~ Table 6.Legislation: permitted sodium, potassium, calcium and magnesium concentrations in milk formulas for infants expressed in mg per 100 kcal USA Canada Net herlands Thailand Yugoslavia Spain Min. Max. Min. Max. Min. Max. Min. Max. Min. Max. 60 20.0 60.0 20.0 60.0 20.0 60.0 20.0 60.0 20.0 60.0 - 80.0 200.0 80.0 200.0 75.0 150.0 80.0 200.0 80.0 200.0 Element Min. Max. Sodium . . - Potassium . . - Calcium . . - - 50.0 - 50.0 - 50.0 140.0 50.0 - 50.0 - 6.0 - 6.0 - 6.0 12.0 6.0 - 6.0 - Magnesium .. - - Table 7. Recommendations for sodium, potassium, calcium and magnesium content (mg per 100 kcal) in an adapted baby formula. N.s. -= not specified Codex Alimentarius Element ESPGAN Min . Max. Sodium . . . . Max.40-48 20 60 Calcium . . . . Min. 60 50 N.s. Magnesium . . . . Min.6 6 N.s. Conclusions The determination of sodium, potassium, calcium and mag- nesium in powdered milk by atomic spectrometric analysis of dispersed samples gives results that are analogous to those obtained when the samples are first incinerated. The method has the advantage of requiring less manipulation of the samples and is less time consuming than the standard methods also studied. We therefore propose that this method could be used routinely to control the alkali and alkaline earth metal content of powdered milk.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Codex Alimentarius Commission, “Recommended Interna- tional Standards for Foods for Infants and Children,” Joint FAOIWHO Food Standards Programme, FAOIWHO, Rome, 1976, CACIRS 72/74. Maurer, J., 2. Lebensm. Unters. Forsch., 1977, 165, 1. Razifard, R., Lait, 1972, 52, 567. Tanner, J. T., J. Assoc. Off. Anal. Chem., 1982, 65, 1488. Murthy, G. K., J. Dairy Sci., 1967, 50, 313. Juarez, M., Martinez-Castro, I., Ramos, M., and Martin- Alvarez, P. J., Milchwissenschaft, 1979, 34, 149. Brooks, I. B., Luster, G . A. and Easterly, D. G., At. Absorpt. Newsl., 1970, 9, 93. Juarez, M., and Martinez-Castro, I., Rev. Agroquim. Tecnol. Aliment. 1979, 19, 45. Rebmann, V. H . , and Hoth, H. J., Milchwissenschaft, 1971,26, 411. Arpadjan, S., and Stojanova, D., Fresenius Z. Anal. Chem., 1980, 302, 206. Willis, J. B., Anal. Chem., 1975, 47, 1752. O’Reilly, J. E., and Hicks, D. G., Anal. Chem., 1979,51, 1905. Van Loon, J . C., Anal. Chem., 1980,512, 955A. Fietkav, R . , Wichman, M. D., and Fry, R . C . , Appl. Spectrosc., 1984, 38, 118. L’vov, B. V., Talanta, 1976, 23, 109. References 1. ESPGAN, Committee on Nutrition, Actu Pediutr. Scand., 1977, Suppl. No. 262. Paper A611 79 Received June 3rd, 1986 Accepted July 14th, 1986