and 10.45%. The downgraded portion contained 1.07% aluminum and results showed 1-05and 1.04%. This definitely established the desired range for the new procedure. STRUCTURE
OF
SODIUM FLUOALUMINATE
At this stage, the method was valid in the range of 1 to 10% aluminum, but there was no conclusive proof that the compound formed was ll?;aF.4A1F3 and not 3SaF.AlF3. X-ray diffraction analysis of the final precipitate indicated that the crystal structure of the double salt formed did not compare with the monoclinic systems of either natural or synthetic cryolite. Hull-Davey x-ray charts indicated that the double salt had a tetragonal structure. From the charts, the lattice parameters of the double salt were determined. Then, from the formula: B
= a%
where a and c are the lattice parameters, the volume of a unit cell was found. With this volume known, it was possible to calculate the x-ray density of the double salt and compare it Ivith that of cryolite. P =
nd
j q
where p
=
n
=
A
=
V = .Ir =
x-ray density the number of atoms per unit cell the molecular weight of the compound the volume of the unit cell Avogadro's number
The calculated x-ray density for the double salt was 2.73 compared to that of cryolite which is 2.95. It was believed that this x-ray density could be
further verified by measuring the density of the double salt formed. When measured in duplicate (using the pycnometer method), the density of the double salt was found t o be 2.72 and 2.74. There were now two negative proofs and one positive proof that the double salt was l l N a F . 4 A l F ~and not 3NaF.AlF,. The negative proofs were: The crystal structure of the double salt was tetragonal and not monoclinic as is the case with cryolite, and there was a difference in densities. The positive proof was the very close agreement between 8-quinolinol and sodium fluoaluminate results of the various alloys analyzed. If results from the sodium fluoaluminate method of analysis had been calculated as 3KaF.dlF3 and not as 1ll\'aF.4B1F3, all of the aluminum percentages n-ould have been low. Another positive proof was needed to verify conipletely the structure of the double salt. Because of the lack of information available on the compound llKaF.4AlF3, the only alternative was to analyze the double salt chemically. Portions of the double salt were analyzed for sodium as sodium sulfate and for aluminum by the 8-quinolinol method. After the percentage of each of these two constituents was determined, that of fluorine was found by difference. By using the percentages of each element, their respective atomic weights, and the total molecular weight of the compound, it was established that there were 11 sodium atoms, four aluminum atoms, and by difference, 23 fluorine atoms present. This then, was the final positive proof needed to verify the structure of the compound. ANALYSIS
OF
COPPER AND ZINC-BASE ALLOYS
-4limited number of copper alloys
and zinc-base alloys containing aluminum were analyzed to test the method on alloys other than titanium. The only changes in procedure necessary were the inclusion of a minimum amount of nitric acid to completely dissolve the samples and the exclusion of the oxidizing step with hydrogen peroxide. Results are as follows: Aluminum, 5;_Sodium 8-
Type of hllay NBS No. 164
fluo-
Quinolinol aluminate results results
[bronze (Mn-A41)] 6 21
6.20
[bronze (Mn)]
0 97
0 99
[bronze (Si)] Aluminum-brass Zamak-3 (zinc base)
0.54 1.92 4.09
0.52 1.90
NBS S o . 62b
NBS No. 158
4.11
ACKNOWLEDGMENT
The author wishes to acknowledge the efforts of J. E. Taylor on x-ray analysis, J. H. Allwein on spectrographic analysis, and the constructive criticisms of J. J. Aldrich and R. C. Burnham, all of this laboratory. The author also wishes to thank Chase Brass and Copper Co., Inc., for permission to publish this work. LITERATURE CITED
(1) Dupuis, T., Duval, C., Anal. Chttn. Acta 3, 183-5 (1949). (2) Lel'chuk, Yu. L., Rutskaya, E. I., J . Bvvlied Cheni. IU.S.S.R.1 22. 490-503 (i9'49). (3) Tananaev, I. V., Lel'chuk, Tu. L., J . Anal. Chem. U.S.S.R. 2 , 93-10?
(1947). (4) Tananaev, I. Y., Talipov, I. G., Zaz.odskaya Lab. 8 , 23-7 (1939). (5) Yaltov, V. S.. J . Gen. Cheni. l..S.S.R. 7, 2439-41 (1937). RECEIVED for reviex September 22, 1958. Accepted January 30, 1959. Division of Analytical Chemistry, 134th Meeting, ACS, Chicago, Ill., September 1958.
Determination of Glycine in Glycine Potassium Tri oxa Ia toc hromate(lll) GEORGE H. SPAULDING Department of Chemistry, Morgan State College, Baltimore,
b In a buffer solution, pH 9.1, the glycine content of glycine potassium trioxalatochromate(111) may b e quantitatively converted to copper glycinate. The solution of the copper glycinate has a purplish hue due to the presence of potassium trioxalatochromate(lll). This compound interferes with the determination of the copper in copper
Md.
glycinate b y the iodometric procedure. It may be removed completely b y passing the solution through an ion exchange column with Amberlite IR-400(OH).The determination of glycine in potassium trioxalatochromate(lll) b y this method i s much more rapid than b y the Kjeldahl method and it is of comparable accuracy.
E
Eotassiuni trioxalatochroniate(III), [Cr(C20i)3]Ks. (KTCr) reacts 11-ith glycine in a 2 s hydrochloric acid solution t o form a glycine complex. Glycine is the only amino arid known to form a n insoluble conipl~xn i t h this salt ( 1 ) . The selective affinity of KTCr for glycinp enables one to separate it from other amino XCESS
VOL. 31, NO. 6, JUNE 1959
1109
acids in protein hydrolysates and other complex mixtures. The glycine complex may be readily purified by recrystallization from a dilute hydrochloric acid solution to which absolute alcohol has been added. Green, iridescent needles are obtained. To facilitate the study of this unique and interesting reaction, a method more rapid than the Kjeldahl procedure for the determination of the glycine content of the complex is desirable. A modification (3) of the method of Pope and Stevens ( 2 ) for the determination of nitrogen in protein hydrolysates by iodometric determination of the soluble copper salts of amino acids is used. A borate-buffered solution of the glycine complex is mixed with a freshly prepared copper phosphate suspension and centrifuged. The clear, supernatant solution is passed through an ion exchange column in order to remove the potassium trioxalatochromate~III) which imparts a purple color to the solution, thus interfering with the subsequent iodometric procedure. The copper glycinate passes through the column unchanged, appearing as a clear blue effluent. The copper of this solution reacts in an acetic acid solution with an excess of potassium iodide to yield an eqivalent amount of free iodine. The iodine is titrated with a standard sodium thiosulfate solution. The procedure has been shown to be precise and accurate through determinations of the glycine content of solutions of the glycine trioxalatochromate(II1) complex standardized by the Kjeldahl method. APPARATUS AND REAGENTS
IONEXCHASGE COLUMN.Use a 10ml. buret, with a funnel top of an inside diameter of 7 mm. Place a wad of glass no01 a t the bottom of the column so as to retain all the resin particles. Charge the column with Amberlite IR-400-(OH) anion exchange resin, 200 to 400 mesh, prepared in the usual manner. An amount of resin which will give a bed about 60 mm. long is used. Wash the resin bed with distilled water until the effluent fails to color the phenolphthalein. Just prior to use, pass 5 ml. of the borate buffer solution through at a rate of about 1 drop per second. Wash the bed with distilled water until the effluent fails to color the phenolphthalein. The resin should be replaced after three determinations. APPARATUS FOR DEAD STOP END POINT,the same as that described by Vernimont and Hopkinson (4).
11 10
ANALYTICAL CHEMISTRY
COPPERSULFATE SUSPENSION.To 40 ml. of a 0.18iM sodium phosphate solution, add 20 ml. of a 0.16M cupric chloride solution with swirling. Centrifuge the mixture for 5 minutes. Remove the supernatant solution and replace it with an equal volume of borax buffer, and resuspend the copper sulfate. Centrifuge the suspension and wash the copper sulfate with borax buffer. Finally, suspend the copper phosphate in 100 ml. of borax buffer, then add 6 grams of solid sodium chloride. BORAXBUFFER,P H 9.1. Dissolve 28.6 grams of sodium tetraborate, NazB406.10H20,in approximately 800 ml. of distilled water, then dilute to 1 liter. POT.4SSIUM
Table I.
Determination of Glycine in Glycine Potassium Trioxalatochromate(111)
KTCr.G, Mg. 37 2 64 3
74 3
Glycine Content,zKjeldahl Iodometric 4 8
8 0
9 2
4 4 4 8 8 8 9 9 9
7 6
7 2 0 0
1 3 1
TRIOXALATOCHROJIATE
(111) GLYCINATE SOLUTION. Solutions water. Allow the mixture t o stand in containing 15 to 36 mg. of the salt per the dark for 3 minutes, then titrate the ml. of solution were prepared. The liberated iodine with 0.005N sodium glycine content of the solution was dethiosulfate, using the dead stop-end termined by the Kjeldahl method. point method. POTA4SSIUN TRIOXALATOCHROMATE Blank determinations according to (111). To a solution of 23 grams of pothe procedure of Schroeder, Kay, and tassium oxalate monohydrate and 55 Mills (3) should be determined and grams of oxalic acid dihydrate in 800 ml. applied. of water heated to about 50" C., add 19 grams of powdered potassium dichroRESULTS mate in small portions with stirring. The results in Table I indicate that After the reaction subsides, evaporate the solution on a water bath to a small the modified iodometric procedure of volume. Upon cooling, microscopic, Pope and Stevens is applicable to the deep green crystals appear. Wash the quantitative determination of glycine crystals with absolute alcohol, then dry in the glycine complex of potassium in air. trioxalatochromate(II1). GLYCINE TRIOXALATOCHROMATE (111). Dissolve 1 gram of glycine and 8 DISCUSSION OF RESULTS grams of KTCr in 50 ml. of 0.5N hydrochloric acid. To this solution, add The results of analyses by the modislowly with stirring, 35 ml. of absolute fied iodometric procedure are in good ethyl alcohol. Allow the mixture to agreement with those obtained by the stand for a t least 2 hours a t about 20" C. determination of nitrogen by the KjelFilter of€ the crystals by means of a dah1 method. The iodometric proBuchner funnel and wash with several portions of a mixture of concentrated cedure is recommended, because of its hydrochloric acid and absolute ethyl simplicity and speed. Less than 30 alcohol 1 to 3. Air-dry the crystals. minutes are required for a determination as compared to 45 minutes to an hour by the Kjeldahl method. PROCEDURE The passage of the borate buffer soluPlace a 2-ml. aliquot of a solution of tion through the column prior to its use the glycine complex of potassium triis important. When this procedure was oxalatochromate(II1) in a 15-ml. centriomitted, reproducible values could not fuge tube. Add 2 ml. of the borate be obtained. buffer solution, followed by 2 ml. of the copper phosphate suspension. Allow LITERATURE CITED the mixture to stand for 5 minutes, shaking occasionally, then centrifuge for 5 (1) Bergmann, M., J . Biol. Chem. 109, minutes. Pour the centrifugate into the 317-24 (1935). ion exchange column and pass through (2) Pope, C. C., Stevens, M. F., Biochem. at a rate of about one drop per second. J. 33, 1070 (1939). Wash the residue in the centrifuge tube (3) Schroeder, TV. A., Kay, L. M.,Mills, R. S., Ax.4~.CHEM.22, 790-3 (1950). with small portions of water and pass the (4) Wernimont, G., Hopkinson, F. J., washings through the column. Finally, IND.ENG.CHEM.,ANAL.ED. 12, 308 wash the column with 50 ml. of water. (1940). Acidify the total effluent with 5 ml. of glacial acetic acid, then add 3 grams of RECEIVEDfor review October 6, 1958. potassium iodide dissolved in 5 ml. of Accepted February 5, 1959.
