ANALYTICAL CHEMISTRY
1200 Tahle IV.
Recovery of Beryllium with Aluminum Hydroxide Carrier
r Be/MI. Added 0.05 0.10 0.20 0.30 1.0 5.0 10 40 50
r Be/MI. Found 0,042 0.005 0.19 0.35 0.32 0.6 5.0 4.4 7.0 9.4
13.4 45 56 58 56 41
Relative Error 15 5 5 17 7 40
0 12 40 6 34 13 12
16 12 18 Av. 16
Other Spectrographic Variables. The National Carbon spectroscopically pure graphite electrodes showed a slight increme in sensitivity over the regular grade, but not enough to warrant the difference in cost. The burning time wm selected as a reasonable compromise between speed'and complete volatilization of the beryllium. The concentration of the sulfuric acid used has a considerable effect on the spectrographic results. This was first pointed out to the authors by Landis (8) and confirmed in their laboratories. The exact concentration is less critical in the range of 3 to 5 N and the 1 to 7 acid concentration was selected for this reason. It should be noted that 0.5 mg. of phosphate ion on the electrode lowered the sensitivity to 0.05 microgram of beryllium. The use of aluminum as internal standard was recommended by Barnes, Piros, Bryson, and Wiener ( I ) ,who used AI 2321.6. The present authors have selected AI 2367 for its suitable density and ability to correct slight variat,ions in operating conditions. LITERATURE CITED
The chief interference seems to be the incomplete precipitation of aluminum (and possibly beryllium) caused by carry-over of organic matter from the oxine separation. However, this is readily detectable by the resultant decrease in the density of the aluminum 2367 line, and such results are discarded. Care in the operations reduces this error to negligible proportions, and, if necessgry, the solution and precipitate may be digested briefly in a water bath to aid coagulation. Selection of Spectroscopic Buffer Carrier. Several salts were tested as buffer carriers for the beryllium, using a crater 5 mm. deep packed with the salt to within 1 mm.of the lip. All samples were arced for 2 minutes a t 10 amperes. Sodium chloride was selected on the basis of sensitivity and freedom from background. The cup depth was selected to hold a sufficient quantity of salt to last for the full 2-minute escitation period.
(1) Barnes, E. C., Piros, W. E., Bryson, T. C., and Wiener, G. ANAL.CHEM.,21, 1281 (1949). (2) Cholak, Jacob, and Hubbard, D. M., Ibid., 20,73 (1948). (3) Ibid., p. 970.
W.,
(4) Eisenbud, Merril, Wants, R. C., Dustan, Cyril, Steadman, L. T., Harris, W. B., and Wolf, B. S.. J . Ind. HYU. .. TozicoE.. 31,282 (19491. ( 5 ) Feldman, Cyrus, ANAL.CHEM., 21, 1041 (1949). (6) Knowles. H. B., J . Research Natl. Bur. Standards, 15, 87 (1936). (7) Kolthoff, I. M.,and Sandell, E. B., J . Am. Chem. Soc.. 50, 1900 (1928). (8) Landis, Parks, Knolls Atomic Power Laboratories, personal communication. (9) Steadman. L. T., U. S. Atomic Energy Commission, AECD-1957 (1948). RECEIVEDJanuary 27. 1950. Presented a t the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Analytical Division, Pittsburgh Section, AMERICAN CHEMICAL SOCIETY, and Spectroscopy Society of Pittsburgh, February 16. 1950.
Micromethod for Estimation of Potassium by Paper Chromatography ERNEST BEERSTECHER, JR.
Biochemical Institute, tnicersity of Texas, and Clayton Foundation for Research, Austin, Tex.
A
VARIETY of methods exists for the determination of small quantities of potassium. Relatively few laboratories are equipped with spectrographic equipment or flame photometers, however, and the latter instrument is frequently troublesome and inaccurate when applied to the analysis of comples samples (8). Polarographic analysis does not distinguish between potassium and sodium (IO). Chemical tests involving gravimetric, volumetric, turbidimetric, and colorimetric analyses of the perchlorate, cobaltinitrite, chloroplatinate, iodoplatinate, phosphotungstate, and dipicrylamine complex are frequently laborious ( 4 )and occasionally dangerous (perchlorate, 6),and require considerable skill (9). I n most of these cases, quantities of potassium esceeding 0.100 mg. are required, and the results obtained from complex mistures vary from *2 to * 10% (I,S,?'). In the course of studies on potassium metabalism there has been occasion to develop a technique for potassium determination which is somewhat simpler than many of the usual methods and requires the use of no equipment beyond that generally found in smaller laboratories. A total sample of less than 0.1 mg. of potassium is required, which for most biological fluids represents
less than 0.1 ml. For this reason it is felt that the technique has broad applicability in general, clinical, research, and classroom work. The method involves a rapid chromatographic separation on paper, development with a sensitive and specific reagent, and quantitation to eliminate errors caused by extraneous materials. By employing a modification of the technique described by Berry and Cain (a), each sample serves both as a standard and as an unknown at the same time. PROCEDURE
The sample solution to be analyzed is diluted or concentrated so as to contain approximately 0.1% potassium. Portions of this sample are placed on a 30 X 12 cm. sheet of Whatman KO.1 filter paper as follows: .Ten points are marked with a pencil 2 cm. from one edge of the paper and 2.5 em. apart from each other. To each of the first five of these spots are added from a suitable pipet 5.0 p l , of the sample, and all the spots are allowed to dry. A standard solution of potassium chloride containing 1.91 grams per liter (1.00 gram of potassium) is then prepared and distributed upon the spots in amounts as follows: 0, 5, 10, 15, 20, 0, 5, 10, 15, and 20 p l . On the sheet as finally prepared, the f i s t five spots contain 5 p l . of unknown plus 0, 5, 10, 15, and 20 micrograms of
V O L U M E 22, NO. 9, S E P T E M B E R 1 9 5 0
1201 ~
~~
A paper chroniatogaphic method for the quantitative estimation of potassium involves a rapid chromatographic separation on paper, development of the chromatogram with sodium lead cobaltous hexanitrite solution, and quantitation of the potassium spots by planimetry. A technique of chromatogram preparation is described which eliminates errors due to foreign materials in the sample. Using the method described, potassium in concentrations of O.l% may be determined with an error of less than 10% on a 0.1-ml. sample.
potassium, respectively. and the second five spots contain 10 uI. of unknown with the same amounts of added potassium as the first five. The filter paper is fixed in the form of a cylinder with staples, and set in a 1- or %liter beaker with the spots along the bottom of the cylinder and in such a maliner that the paper cylinder does not touch the sides of the beaker. Into the bottom of the beaker is introduced, to a depth of about 1 cm., a solution consisting of 80 pa& of ethyl alcohol and 20 parts of 0.1 A' hydrochloric. acid. The bwker is covered with a watch ghss and the solvent is allowed to ascend to the top of the filtrr paper cylinder. This ascent requires about an hour. The cylinder of filter paper is then removed from the beaker and dried in air. Sodium lead cobaltous hexanitrite ( 5 ) is used to develop the chromatogram. The reagent is conveniently prepared 1)y dirsolving 11.5 grams of lead nitrate and 15.0 grams of sodium niti,ite in 50 nil. of distilled water. Khen solution has occurred, 10.0 granis of c.ol)altous nitratc are added and the volume is madr up to 100 ml. with distilled water. This solution is allowed to st:tnd for 1 hour, diluted to 200 nil. with distilled water, and filterid just prior to use. The reagent is sprayed lightly over the chromatogram with a hand atomizer or pressure spray, after which a green spot a t R, 0.4 rapidly develops, owing to the presence of potassium. Ilnder the conditions of the test, no interferenrr occurs as a result of the presence of ammonia, c-alcium, or magnesium, which give positive reactions in crystallographic work employing this re:igent (IO). The chromatogram is then allowed to dry, and the spots are carefully circumscribed with a pencil. The ai'cas of the spots may then be measured with a grid system or ivith a polar planimeter, or quantitated by ewision and tveighing oit an analytical balance. The results reportrtl here wew ohtninod hy planinieti,y. \\'heu the areas of the first five spots are plotted against tiic amounts of added potassium, a straight line is obtained, as shown in Figure 1. A second straight line, parallel t o the first, is obtained by plotting the corresponding areas against the potassium added to the 10-pl. samples of unknown. It is apparent from Figure 1 that the difference between the intercept,s of these two parallel lines with the z-axis is equivalent to the potassium content of 5 pl, of unknown solution ( K in Figure 1). This follows because K corresponds to the amount of potassium that must be n d d c ~ lt o the S p l . aliquot of unknown to produce a spot of the
same size as the lO-pl. aliquot. It is generally best t o analyze the d a t a by this graphic method. By employing simple mathematical formulas, however, it is possible to calculate the potassium content of the sample from the equation: K(1,,g./,nI,) =
M here X , O = the mean area of the spots containing 10 p l . of the unknown, xj = the mean area of the spots containing 5 MI. of the unknown, Z(Iia.4) = the sum for the ten spots of the products of the area of each spot and the microliters of p o t w i u m standard added to that spot, arid A = the total area of all ten spots. Minor deviations in technique, however, may greatly increase the errors obtained by the use of this formula, so that a graphic solution is preferable except when experienced workers are conducting routine analyses. The accuracy of the procedure may be further increased by increasing the number of increments of Lnown potassium concentrations on the chromatogram, or by t l v t cbrmin:it ion i n rfbp1ic:itP,
.9 .~
c
9.8
z
e.7
a 4:
w a.5 4:
.4
.3 .2
t1
Table I. Typical Analytical Data Based on Determination of Potassium in Urine and Unknowns Sample Urine sample 1
r r i n e sample 2
Crinesample2 Cnknown Unknown Unknoxn Unknown Unknown
+ 0 .. 26 nmg. i g . IC/inl.b + K/ml. b + 01.Omg. Ii/ml.b
.4c (1 .70 mg. &'nil.) B (1.16 mg. K / n ~ l . ) C (0.67 mg. K:inl.)
D (2.07 mg. 1
+I2
tlf - 1
The technique described was originally designed for the determination of potassium in urine, and the analytical data in Table I were obtained from this material and from a series of unknowns such as are used in quantitative analysis courses. Tests on other materials have shown that the method works well in general analytical work, although slight changes in technique may be desirable in some cases. Thus, when large amounts of foreign materials interfere with the potassium spot, the test should be run on a larger sheet of filter paper, w that greater resolution of thr components of the sample is obtaiiied. DISCUSSION
The technique employed largely embodies the usual methods of paper chromatography. The modifications concerned with the
ANALYTICAL CHEMISTRY
1202 analysis of the data, however, involve several unique factors. When the same increments of potassium are added to the two series containing different amounts of unknown, and the areas of the resulting spots are measured, the rate of increase of the areas within the two series should be equal, and the plotted results should produce two parallel lines. It is therefore easy to devise a series of simultaneous equations which could be solved for the potassium content of the unknown. The equation presented is based on these considerations and provides a simplified method for calculating the result. The quantitation of results obtained by paper chromatography is extremely subject to errors due to the effects of foreign materials in the sample, and a method based on the principle of sample distribution employed here is essential for a suitable degree of accuracy, as pointed out by Berry and Cain ( 2 ) . The concentration of the extraneous materials ia a sample spot on paper may markedly influence the degree of spreading of any partirular constituent. Consequently, comparison cannot validly be made of the areas of urinary potassium spots with standard spots in which large concentrations of other materials are absent. I t is essential that the potassium standard be influenced in its “spreading” and migrating characteristics in the same manner as is the potassium in the unknown sample. This may be achieved by preparing the standard solution so that it contains representative amounts of the major constituents of urine. Such a procedure is analogous to that employed in the flame photometry of complex mixtures, where the reference standard is made up with foreign materials present so as to resemble the type of unknown under study. The use of the unknown itself as a base for the standard, however, presents R more accurate means of achieving this effect.
Sodium lead cobaltous hexanitrite has been previously used for the detection of potassium by microcrystallographic means ( 5 ) , under which circumstances ammonium, cesium, lithium, rubidium, thallium, arsenic, antimony, calcium, chromium, iron, magnesium, silver, and tin are reported to interfere with the results. Under the conditions of the test described here, none of these materials interferes, although some difficulty may result when these elements are present in very high amounts. Using the method described, potassium in concentrations of about 0.1% may be determined with an error of less than 10% on a sample of 0.075 ml. of solution. Although the accuracy of the present method is thus not so great as that of some other methods, it seems adequate for many analyses in which greater precision is not required, and where limitations exist on equipment, sample size, and time. LITERATURE CITED (1) Adams, M. F., and St. John, J. L., IND.ENO.CHEM.,ANAL.ED., 17,435-6 (1945). (2) Berry, H. K . , and Cain, L., Arch. Biochem., 24, 179-89 (1949). (3)Cotton, R. H., IND.ENO. CHEM.,ANAL.ED.,17, 734-8 (1945). (4) Folch, J., and Lauren, M.. J . Biol. Chem., 169, 539-49 (1947). (5) Frediani, H. A., and Gamble, L., Mikrochemie, 29, 22-43 (1941). (6) Kailmann, S., IND.ENG.CHEM.,ANAL.ED., 18, 678-80 (1946). (7) Keiley. 0. J., Hunter, A. S., and Sterges, A. J., Ibid., 18, 319-22 (1946). (8) Parks, T.D., Johnson, H. O., and Lykken, L.,Ibid., 20, 822-5 (1948). (9) Tinsley, J., Analyst, 74, 167-78 (1949). (10) Weaver, J. R., and Lykken, L., ANAL. CHEM.,19, 372-6 (1947).
RECEIVED December 27, 1940.
Road-Table Discusebn
FLAME PHOTOMETRY Digest of stenographic report of round-table discussion held by Division of’ Analytical Chemistry, 117th Meeting, A.C.S., Houston, Tex., March 1950 Moderator: W. G . SCHRENK, Department of Chemistry, Agricultural Experiment Station, Manhattan, Kan.
F
LAME photometry qffers excellent possibilities for the development of analytical procedures for those metallic ions which can be excited by the relatively low excitation levels available in the flame. The procedure is not new and has been used for years as a qualitative test for ions such as potassium and sodium. Interest in quantitative procedures, however, has increased rapidly in the past few years. This has been due, first, to the development and availability of instruments, designed for this purpose (I, 2 , 4 , B ) , and secondly, to the need for simpler and more accurate methods for the determination of the ions of the alkali metals. Particular attention has been given to the determination of sodium and potassium because of their importance in biological systems. Other elements, however, can be determined by this procedure. Of particular interest a t the discussion were calcium and magnesium. Although flame methods appear promising with regard to the determinations of these elements, any new technique involves certain difficulties not found in established methods. As a result discussion centered around two generral topiqs concerning this method of analysis. First, were factors hfluencing instrument stability and precision and, second, were the preparation of samples and calibration techniques.
FACTORS INFLUEYCING INSTRUiMEYT REPRODUCJBILITY
Included among the factors which govern the reproducibility of any given instrument are such items as the atomizer system, air supply, gas supply, and the stability of the electronic amplifier circuit, including the photocell. Discussion brought out the point that atomizer construction is critical, Atomizers of apparently identical construction require separate calibration and therefore are not entirely interchangeable. The atomizer described by Weichselbaum and Varney (6) was mentioned, but no performance data were available other than those described by them in their paper. Several people mentioned the need for extreme cleanliness in handling the atomizer. It was pointed out that if the atomizer does not drain properly erratic results will occur. Some workers clean the atomizer with distilled water and alcohol after each determination, followed by periodic cleanings with other materials. Others routinely clean after a definite number of determinations. It was also pointed out that the type of sample plays a role in determining the number of analyses which can be made before cleaning is required. Biological samples, in general, tend to reduce the number of times the atomizer can be used before thorough
