Boron in Sodium Metal Determination of Microgram Amounts by Alcohol Extractions JOSEPH RYNASIEWICZ, MURIEL P. SLEEPER, and 1. W. R Y A N Knolls Atomic Power Laboratory, General Electric Co., Schenectady,
A
N. Y.
sodium hydroxide and 2 nil. of glycerol-methanol solution, and evaporate to dryness under a heat lamp. Continue the heating for a few minutes, char the residue carefully over a Meker burner, and complete the ignition with the full heat of the iMeker flame. Cool the platinum dish in a tray of crushed dry ice and add 2 ml. of the curcumin reagent down the sides of the dish. Then add 1 nil. of the oxalic acid-hydrochloric acid solution, allowing each drop to freeze before adding the next one. Remove the dish from the dry ice, heat it with the palm of the hand, and rotate the viscous niisture around the inside walls of the dish until most of the salt dissolves. Float the dish on a water bath at 60" i 1' C. for 10 minutes in excess of the complete evaporation period. After cooling, extract the residue with 70% ethyl alcohol, transfer to a 25-ml. volumetric flask, and make to volume with 70% ethyl alcohol. Shake well! centrifuge the suspension in a 15-ml. cent,rifuge tube for 10 minutes a t 1500 r.p.m., decant the supernatant solution into a 20-mm. cell, and measure the absorption a t 550 mg with a Lumetron colorimeter using air as 100% transmittance. The blank on the reagent should read about 80% transmittance. Calculate the boron present from a standard curve made by plotting parts per million of boron against per cent transmittance. The range of the standard curve is from 0.02 to 0.32 p.p.ni. of boron, or 0.5 to 8.0 y of boron in a 25-ml. volume. Trea.t the standard boron samples used for preparing the curve in the same manner as the unknown samples except for the alcohol extraction from sodium chloride. Run standard boron samples with each nelv batch of curcumin solution.
CONPREHGXSI\T review of the literature dealing with t,he quantitative analysis of boron, especially in trace aniounls, has been made by Russell ( 4 ) . Bewirk, Beaniish, and Rartlet ( 1 ) determined boron in calcium and magnesium metals by distilling methyl borate from the chlorides and then estimating t,he boron colorimetrically wit,h turmeric reagent. The authors have applied the procedure of Bewick and his coworkers to the dctermination of microgram amounts of boron in sodium and obtained resuks which werc about 20% low, but sufficiently accurate for most purposes. The lo^ values wcre ascribed t o incomplete distillation of methyl borate from concentrated sodium chloridc solutions, as reported by Russell ( 4 ) . The classical met,hyl borate distillation is time-consuming, and rrquires special apparatus and large quantities of methanol. As this laboratory was intereeted in the routine determination of boron in sodium metal, a more rapid procedure was desirable. Consequently, the following method was developed, whereby boron ie extracted from sodium chloride with ethyl alcohol and estiniatcd colorimetrically using curcumin (1, 3 ) . PROCEDURE
Apparatus. Quart,z or 728 Corning glass (low in boron and alkali resistant 1) was used for storage and distillation of all reagents. =Z 500-ml. quartz Erlenmeyer flask was used to dissolve the sample. Reagents. Distilled water tapped directly from a Barnstead still. Hydrochloric acid solution, approximately 4.5.V. Dissolve hydrogen chloride gas in distilled water in a boron-free glass 1liter bottle set in a bath of cool running water. Dilute for use. Hydrochloric acid solution, approximately 10%. LIethanol, distilled from sodium hydroxide. Sodium hydroxide, approximately 0.1N. Curcumin solution, 0.1%. in 95% ethyl alcohol. Decant the solution from any insoluble material. This solution is stable for about I week. Oxalic acid-hydrochloric acid solution. Dissolve ti grmis of oxalic arid in 80 ml. of water plus 20 ml. of concentrated hydrochloric acid. This solution is stable for about 2 weeks., Glycerol-mct.hano1 solution; 3% glycerol hy volume. Ethyl alcohol, io%, by volume. Dilute 95% ethyl alcohol distilled from sodium hydroxide. Standard boron solution containing 1 y of boron per nil. Solution A . Dissolve 4.4065 grams of pure borax (Na2B40i. lOH,O) in a 500-ml. volumetric flask and make to volume. This gives a solution containing 1 mg. of boron per ml. Solution B. Pipet 5 ml. of solution h into a 250-nil. volumetric flask and make to volume. Solution C. Pipet 10 ml. of solution B into a 200-m1. volumetric flask and make to volume. This contains 1 y of boron per ml. After being made to volume the solutions should be transferred and stored in quartz or low-boron glass. Solutions B and C should he freshly prepared. Method: Cut a weighed sodium sample (about 2 grams) int,o pea-size pieces. Dissolve the sodium, piece by piece, in about 50 ml. of water in a quartz Erlenmeyer flask purged wit,h a continuous stream of nitrogen. Cool the alkaline solution in an ice bath and nearly neutralize it with 4.5N hydrochloric acid using methyl red indicator. Transfer the solution to a 3.75-inch platinum dish and evaporate just to dryness under a heat lamp. Break up the aggregated sodium chloride crystals in the platinum dish with a quartz pestle and add 10% hydrochloric acid dropwise until the salt becomes slightly acid to methyl red. Extract the boron with 15 ml. of %yoethyl alcohol and decant the solution into a 50-ml. centrifuge tube. Repeat the extraction with t F o more portions of alcohol made acid with 1 drop of 10% hydrochloric acid. Centrifuge the combined extracts for about 5 minutes, transfer the solution to a 2.75-inch platinum dish, and neutralize with 0.LV sodium hydroxide. Add 5 ml. of 0.1N
RESULTS
Tablc I shows that as little as 0.5 p.p.m. of added boron was determined in sodium hydroxide solutions by the alcohol extraction procedure with less than a 20% relative error, comparable to the accuracy obtained by the arduous methyl borate distillation method. -4n accuracy of ~!=200/,was considered extremely satisfactory for these low-boron concentrations.
Table I. B .4dded, P.P.1\1.
Determination of Boron in Sodium"Using Ethyl ..ilcohol Extractions B Presenth, P.P.M.
P.p 111
B Found ';o of B added
% of B present 0.0 0.0 0.0
. .
0 5 n 5 0.5 0.6
1 5 1.5
2 4
R O
0.7 0 9 0.0 0.9 1.0 1.9 1.4 2.8 3.4
0.0
inn so so 1 no I 07
113 108 Q3
0.0 56
44 44
60 84 9n 93 82 85 77
3.4 2.9 97 4.9 5.3 4.1 84 72 i 16 A V . s; B found 2 0 . 9 p.p.111. QF, i i n * Boron added as NaQB4OI t o sodium hydroxide solutions containing equivalent of 2 grams of sodium. h Boron in sodium hydroxide plus added boron. 3.0
Several analyses were made to obtain a typical boron value for "as-received" Du Pont and Baker's analyzed sodium (Table 11). All values fell in the range 0.1 to 2.0 p.p.m. of boron by the alcohol extraction procedure and by the methyl borate distillation method. Sample 9060 contained approsimately 0.4 p.p.m. more boron according to the standard methyl borate distillation method than by the alcohol extraction procedure. Since the sodium hydroxide used for the boron recovery data (Table I ) was not analyzed by
935
ANALYTICAL CHEMISTRY
936 Table 11. Boron Content of As-Received Sodium Metal Sample 8879
KO,
9060 8693 8694
h‘o number
Distillation, Difference, Alcohol Extraction, P.P.M. B P.P.M. B P.P.M. B 0.11,0.06 0.19,0.19,0.30,0.11,0.22, 0.32 O.G9,0.57 0.41 0.11 0.22 3 . 2 . 1 . 9 . 1.0
the distillation method, it is possible that the alkali contained up to 0.4 p.p.m. of boron. Therefore, the data in Table I have been corrected to account for the extra 0.4 p.p.m. of boron in the sodium hydroxide. A more conservative appraisal of the data indicates that a minimum of 0.9 p.p.m. of boron can be detected. Above 0.9 p.p.m. of boron, the recovery was 72 f. 16%. DISCUSSION
Although 0.5 y (0.02 p.p.m.) of boron could be measured by the curcumin colorimetric procedure in the standard solutions, an increment of only 1 y of boron could be determined in concentrated sodium hydroxide solutions by the method described above. Apparently the boron was incompletely extracted, or interfering ions were extract-ed giving the low results.
Table 111. Effect of Sodium Chloride on Estimation of Boron by Curcumin Colorinietric Method 7
Boron P.p.m.
Transmittance, yo No NaCl added 0.58 Gram NaCl a d d z 78 53 23 16 7 ~~
It W&B found that 23 mg. of sodium chloride, together with the boron, were extracted with each milliliter of alcohol solution. The effect of this amount of salt on the estimation of boron by the curcumin procedure was insignificant, as reported in Table 111. There was some speculation that boron was present in sodium in a form which was insoluble in the slightly acid ethyl alcohol,
Therefore, a sodium hydroxide fusion was carried out to render all boron compounds (such as boron carbide and elemental boron) acid-soluble. Previous experience with an alkali fusion showed that boron carbide and amorphous boron were converted to an acid-soluble form which was completely recovered. The fused sodium hydroxide samples yielded no additional boron. The procedure described above assumes that no large amounts of interfering ions are extracted v i t h ethyl alcohol. Feigl (2) reported that the presence of iron(III), molybdenum, titanium, columbium, tantalum, and zirconium ions would give high results for boron by the curcumin method. Of the ions mentioned, only iron has been found in substantial amounts in “asreceived” sodium. Since sodium metal is a strong reducing agent, it was believed that the iron was present in the ferrous state, even after dissolution of the sample in water followed by neutralization with hydrochloric acid. A test for iron in Du Pont sodium after dissolving and neutralizing the sample indicated the presence of 10 p.p.m. of iron, half of which was iron(II1). The sodium hydroxide used to prepare the alkali-boron solutions (Table I ) contained 20 p.p.m. of iron. Though no data were obtained to show the quantitative effect of iron(II1) (or the other impurities in sodium) on boron recoveries, it is believed that the chief reason for the low results is inefficient extraction and not interferences from foreign ions in the concentrations usually found in sodium metal. Where large amounts of the interfering substances are present or where an abnormally high value for boron has been obtained, the methyl borate distillation procedure should be used. ACKNOWLEDGMENT
The authors wish to thank L. P. Pepkowitz for encouraging this work and for reviewing the manuscript. LITERATURE CITED
(1) Bewick, H. A., Beamish, F. E., and Bartlet, J. C., Natl. Research Council Can., Tech. Rept. MC-228 (1948). (2) Feigl, F., “Qualitative Analysis by Spot Tests,” p. 256, New York, Elsevier Publishing Co., 1946. (3) Naftel, J. A., IND.ENG.CHEM., .4xa~.ED.,11, 407-9 (1939). ( 4 ) Russell, J. J., Univ. Toronto, Rept. MC-47 (1944). The Knolls Atomic Power Laboratory is operated by the General Electric Co. for t h e Atomic Energy Commission T h e work reported here was carried out under contract No W-31-109Eng-.jZ
RECEIVED for review -4ugust 12, 1953. Accepted February 11, 1954
Fluorometric Determination of Cadmium NUR EVClM
and
LOUIS A. REBER
Department o f Chemistry, Philadelphia College o f Pharmacy and Science, Philadelphia 4, Pa,
A
USEFUL chelating agent, 2-(o-hydrouyphenyl)-benzoxaaole, has been suggested by Freiser (1). Walter and Freiser described gravimetric (2) and volumetric (3) methods for the determination of cadmium using this reagent. They also stated that the cadmium chelate is soluble in glacial acetic acid and the solution fluoresces with a bright blue color under ultraviolet light. The possibility of determining small amounts of cadmium, using fluorescence as a basis, has been investigated, and a suitable method is presented in this paper. It was found that 0.1 to 2 mg. of cadmium could be determined with an average error of 1 0 . 0 2 mg. APPARATUS AND MATERIALS
Reagent Solution. A solution was pre ared by dissolving 1 gram of 2-(o-hydroxyphenyl)-benzoxazole ?Eastman Kodak Co.) in 1 liter of ethyl alcohol (95%, U.S.P.).
Standard Cadmium Solutions. To test the precision and accuracy of the proposed method, a solution was prepared from 18.25 grams of cadmium sulfate (CdS04.5H20, reagent grade) dissolved in sufficient water to make 1 liter of solution. The cadmium content was determined b y the gravimetric procedure of Walter and Freiser ( 2 ) . By appropriate dilution, a standard cadmium solution was prepared containing approximately 0.04 mg. of cadmium per ml. Accurately measured volumes of this solution, containing known amounts of cadmium, were then analyzed by the proposed method. Solutions for Fluorometer Calibration. A quantity of the cadmium complex of 2-(o-hydroxyphenyl)-benaoxazole was precipitated by the method of Walter and Freiser (Z), washed, and dried for 2 hours a t 130’ C. The pure, dry material, 0.2371 gram, which is equivalent to 0.0500 gram of cadmium, was dissolved in sufficient glacial acetic acid to make 1 liter of solution. This solution could be kept for one week without any significant change. Appropriate volumes of solution were diluted as needed with glacial acetic acid to establish the calibration curve of t,he fluorometer.