Spectrophotometric Determination of Magnesium by Titan Yellow

Ind. Eng. Chem. Anal. Ed. , 1942, 14 (11), pp 895–897. DOI: 10.1021/i560111a028. Publication Date: November 1942. ACS Legacy Archive. Cite this:Ind...
1 downloads 0 Views 444KB Size
ANALYTICAL EDITION

November 15, 1942

Acknowledgment The author wishes to thank L. J. Lassalle of Louisiana State University and R. S. Nelson and H. C. Leonard of the Gulf States Etilities Co. for permission to publish the material in this paper. The writer is indebted to Thos. E. Crossan and B. Gurney of the Gulf States Utilities Co., whose interest in the subject made this work possible. Thanks are due M. B. Sturgis, Agronomy Department, Louisiana State University, for the use of the Coleman spectrophotometer.

Literature Cited Alimarin, I. P., and Zverev, V. S., Trans. Inst. Ecqn. Mineral. ( U . S.S. R.), No. 63, (1934) ; Mikrochemie, 22, 89-100 (1937). Am. Public Health Assoc., “Standard Methods for the Examination of Water and Sewage”, 7th ed., New York, 1933. Ibdd., 8th ed., pp. 103-5, New York, 1936. Ammer, G., Wasser, 8, Pt. 11, 134-7 (1934). Bodnar, J., and Torok, T., 2. physiol. Chem., 261, 257-68 (1939). Cerny, M., Listy Cukrouar., 52, 269-73 (1934). Davies, J. G., Gomez, A. C . , and Boon, D., Intern. Sugar J . , 40, 105-6 (1938). Davydov, A. L., and Malinovskaya, 0. A., Zavodskaya Lab., 9, 964-7 (1940). Davydov, A. L., Reznik, B. E., and Valsberg, Z. M., Ibid., 8, 1033-8 (1939). Dienert, F., and Wandenbulcke, F., Compi. rend. 176, 1478-80 (1923); Bull. SOC. chim., (4) 33, 1131-40 (1923). Hadley, W. H., Analyst, 66, 486-9 (1941). Kahler, H. L., IND.ENQ.CHEM., ANAL.ED., 13, 536-9 (1941). King, E. J . , Biochem. J., 33, 944-54 (1939). King, E. J., J . Biol. Chem., 80, 25-31 (1928). King, E. J., and Lucas, C. A,, J . Am. Chem. Soc., 50, 2395-7 (1928). King, E. J., and Stantial, H., Biochem. J . , 27, 990-1001 (1933). Knudson, H. W., Juday, C., and Meloche, V. W., IND.ENG. CHEM., - ~ X \ L ,ED., 12, 270-3 (1940).

895

(18) Korenman, I. M., and Kozhukhin, L. A., Zauodskaya Lab., 9, NO. 1. 43-5 (1940). (19) F u m h o l z , P., 2. anoTg. allgem. Chem., 212, 91-6 (1933). (20) Liebknecht, O., Gerb, L., and Bauer, E., J. angew. Chem., 44, 860-3 (1931). (21) Nagerova, E. I., and Petrova, A. D., Vsesoyuz. Nauch.-Issledouatel. Inst. Tsement. V N I T s . Sbornik Rabot, No. 17, 56-60. (22) Nemec, A., Lanic, J., and Koppova, A., 2. anal. Chem., 83, 4 2 8 4 5 (1931). (23) Oberhauser, Ferd, An. jac. 510s.educac., Univ. Chile, Secc. quim.. 2, NO. 2-3, 93-7 (1938). (24) Rees, 0. W., IND.ENG.CHEM.,ANAL.ED., 1, 200-1 (1929). (25) Rosenheim, A., Z.anorg. Chem., 4, 352-73 (1893). (26) Rosenheim, A . , and Bertheim, A.. Ibid., 34, 427-47 (1903). (27) Sakane, H., J . Oriental Med., 32, 95-8 (1940). (28) Schwartz, M.C., IND. ENG.CHEX.,ANAL.ED., 6, 364-7 (1934). (29) Schwartz, M.C., La. State Univ. Bull. 30, No. 14 (1938). (30) Schwartz, M.c.,and >Iorris, L. W., IND.ENG.CHEbf., ANAL. ED.,to be published. (31) Swank, H. W., and Mellon, M. G., Ibid., 6, 348-50 (1934). (32) Tananaev, N. A , and Shapovalenko, A. M., J. Applied Chem. (U.S . S. R.), 11, 3 5 2 4 (1938). (33) Thayer, L. A., IND.ENG.CHEM.,ANAL.ED., 2, 276-83 (1930). (34) Tschopp, Ernst, and Tsohopp, Emilio, Helv. Chim. Acta, 15. 793-809 (1932). (35) Urbach. C., Mikrochemie, 14, 189-218 (1934). (36) Urech, P., Helu. Chim. Acta, 22, 1023-36 (1939). (37) Vasil’ev, K. A., and Zakharov, E. L., Zavodskaya Lab., 10, 1435 (1941). (38) Weihrich, Robert, and Schwarz, Walter, Arch. Eisenhiittenw., 14, 501-3 (1941). (39) Weinland, K. F., and Zimmermann, K., 2. anorg. Chem., 108. 248-66 (1909). (40) Winkler, L W., Ibid., 27, 511-12 (1914). before t h e Division of Water, Sewage, a n d Sanitation Chemistry PRESENTED a t t h e 103rd Meeting of t h e AMERICAN CHEYICAL SOCIETY, Memphis, Tenn. Joint contribution from t h e Water Technology Laboratory, Engineering Experiment Station, Louisiana State University, and t h e Gulf State8 Utilities Co., Baton Rouge, L a .

Spectrophotometric Determination of Magnesium by Titan Yellow E. E. LUDWIG

AND

C. R. JOHNSON, University of Texas, Austin, Texas

F

OLLOWISG Kolthoff’s (6) introduction of Titan yellow as a reagent for magnesium, his method has been adapted to various determinations of this element ( 7 ) . A limited amount of spectrophotometric data is available (IO). The method has been recommended for the colorimetric determination of magnesium in water (6, 8, 9, IO), but there has been no attempt to develop a standard spectrophotometric procedure and to compare it with the official gravimetric method. The present research was undertaken for this purpose, since it seemed likely that there could be worked out a rapid spectrophotometric procedure, comparing favorably in accuracy with the official gravimetric method, and less subject to uncertainties than standard volumetric methods. The increasing availability of fine yet comparatively inexpensive spectrophotometers also made this work seem worth while.

Principle of the Method When magnesium hydroxide is precipitated in the presence of Titan yellow by sodium hydroxide the yellow color of the reagent changes to red or orange-red. With dilute magnesium solutions the lake which is formed remains dispersed for rather long periods, particularly in the presence of protective colloids such as starch, agar, or dextrin, and the suspension appears clear to the eye. I n the presence of calcium ion the red color is deeper, and the suspension has a minimum trans-

mittance a t a wave length of 525 millimicrons when compared spectrophotometrically with a suitable blank. Transmittance readings are highly reproducible and sufficiently constant to allow ample time for precise observation. The lake is apparently kept in suspension by electrostatic forces created by the calcium salt and the protective colloid, since suspensions prepared without either calcium sulfate or starch, or similar materials, flocculate readily or give transmittance readings which are not reproducible. Statements made in this connection by Hirschfelder and Serles (6) in criticizing Becka’s (2) procedure should not be interpreted as meaning that accurate analyses cannot be made in the presence of excess calcium.

Apparatus and Reagents A Coleman 10-5-30 spectrophotometer was used t o measure transmittances. Matched square cuvettes 1.308 cm. in depth were used t o hold the blanks and test solutions. Measurements were made at temperatures from 25 ’ t o 29 C. Several sets of reagents were prepared during the work. The chromogenic agent was a 0.05 per cent solution of Eastman’s Titan yellow. Calcium sulfate solution was made by stirring 5 grams of reagent grade calcium sulfate in contact with 1 liter of distilled water for about 4 hours and filtering the liquid. One per cent starch solution was prepared every few days from soluble starch. Normal sulfuric acid and 2 W sodium hydroxide were prepared in the usual manner. O

896

INDUSTRIAL AND ENGINEERING CHEMISTRY

TABLEI. CONCEKTRATION-TRANSMITTAKCE DATAFOR MAGNESIUM BY TITAN YELLOW Magnesium Mg./lOO mi. 0.1000 0.200 0.300 0.400 0.500 0.600

Transmittance

% 83.9 70.6 59.5 50.1 42.4 30.3

Magnesium Mg./100 ml. 0.700 0.800 0,900 1,000 1.100

Transmittance

% 31.8 28.8 28.8 25.3 23.8

Calibration Experiments Calibrations were made by the procedure described below for analyses. I n some cases the test solutions contained quantities of magnesium accurately measured by dilution of primary standard solutions of magnesium sulfate whose concentrations were established by several gravimetric analyses. Additional calibration points were also calculated from the transmittances of the test solutions prepared with differentsized portions of all the waters analyzed, the corresponding magnesium concentrations being established by the official gravimetric analyses. The calibration data in Table I are values read at even concentrations from the smooth curve drawn through the fifty-odd calibration points obtained by these two methods, to minimize the accidental errors graphically. It is convenient to construct the final calibration curve with these data on Coleman (2-202 concentrationtransmittance semilogarithmic paper.

Procedure After any colloidal sulfur or hydrous ferric oxide in the water to be analyzed has settled out, two or three different-sized samples are measured accurately into 100-ml. volumetric flasks. The samples should contain between 0.1 and 1.1 mg. of magnesium.

OF MAGNESIUM ANALYSES TABLE 11. SUMMARY

Water

Spectrophotometric Analyses Sample Transmittance M g found

M1. Tyler city Wac0 city Dallas city Austin city Corpus Christi city S s n Antonio city

Austin, Capitol fountain (artesian)

100.0 150.0 200.0 40.0 46.0 40.0 46.0 20.0 30.0 40.0 20.0 40.0 10.0 20.0 25.0 30.0

7' 73.0 62.6 58.7 79.7 73.2 72.9 70.0 62.4 53.1 44.5 66.2 45.6 76.6 58.9 51.5 46.6

P.p . m 1.8 1.8 1.6 3.3 3.9 4.5 4.5 13.6 12.3 11.8 11.9 11.4 15.4 15.3 16.4 14.8

20.0 35.7 30.7 25.0 29.6 30.8 Austin, Scout pool (arte10.0 50.5 39.6 sian) 20.0 28.8 40.0 25.0 24.6 41.7 Synthetic Arkansas city 10.0 43.6 48.4 Rosborough Springs b 1.00 69.0 213 2.00 48.2 212 34.6 212 3.00 5.00 24.3 213 S s n Antonio, Terrell wellsc 0,500 81.4 232 1.00 67.0 231 2.00 45.4 230 227 5.00 23.2 Sea a t Galvrston 0.250 64.0 1032 0.500 42.1 1010 0 750 29.4 1038 1037 1.00 24.7 Precipitated hydroua ferric oxide and sulfur on standing. b Heavy hydrous ferric oxide precipitate on standing. c Heavy sulfur precipitate on standing.

Gravimetric Analyses, M g Found P. p . m 2 0

4.4 4.9 11.6 11.6 15.8

29.6 39.1 49.0 211

231

1027

Vol. 14, No. 11

For waters requiring samples less than 10 ml. a microburet readable to 0.001 ml. may be used. The largest sample that can be used is about 49 ml. and if the water contains less than 2 p. p. m. of magnesium the samples should be acidified slightly with hydrochloric acid and concentrated by evaporation, preferably in platinum, since inconsistencies are sometimes observed when ordinary glassware or porcelain is used. To each volumetric flask are then added in order 1 ml. of 1 N sulfuric acid, 10 ml. of 1 per cent starch solution, 20 ml. of saturated calcium sulfate solution, 10.0 ml. of 0.05 per cent Titan yellow solution, and 10 ml. of 2 N sodium hydroxlde. The solution is made to exactly 100 ml. with distilled water, poured into a 250-ml. glass-stoppered Pyrex Erlenmeyer flask, and shaken vigorously for 5 minutes. A portion of the suspension is then transferred to a cuvette of the spectrophotometer and as soon as air bubbles have disappeared the transmittance relative to a blank is determined at a wave length of 525 millimicrons. The blank is prepared in exactly the same manner as the test solution, except that the unknown water sample is not added The blank is good only for any one day's determinations. The milligrams of magnesium for the samples taken are read from the calibration graph, converted to p. p. m., and averaged. The complete analysis requires about 20 minutes.

TABLE 111. AKALYSES OF COSCEXTRATED MAGKESIUM CHLORIDE SOLUTION -Spectrophotometric Aliquot Transmittance

M1.

%

0.500 0,750 1.50

64.1 51.4 29.4

Grsvimetrir Analyses, N g C b Found

AndyEes--MgCh found

Wt. %

Wt.

31.4 31.6 31.8

7%

32.1 31.6

..

ON MAQNESIUM ANALYSES TABLE IV. EFFECTOF IRON

Iron Added P. p . m. San Antonio city (10-ml. samples) 0 1 5 Austin Scout pool (10-ml. 0 samiles) 1 5 Water

M g Found 11g Found (Spectroscopic) (Gravimetric)

P. p .

m

15.4 15.0 14.4 39.6 39.7 41.1

P. p . m 15 8

39.I

Summary of Results The above procedure has been compared with the official gravimetric method for various Texas waters covering a wide range of magnesium, calcium, and iron concentrations. No analyses were made for the latter two elements, since in most cases sufficient data are available elsewhere (3). The spectrophotometric analyses were made without removal of iron and calcium. I n the gravimetric analyses iron and calcium were removed in the usual manner with ammonium hydroxide and ammonium oxalate, and magnesium was determined by the official method (1). Results of the analyses are summarized in Table 11. I n Table I11 similar data are given for analyses of a highly concentrated magnesium chloride solution, such as is obtained in the commercial preparation of magnesium metal. This solution contained 0.12 per cent of calcium, 0.11 per cent of sulfate, and 90 p. p. m. of manganese, and had a density of 1.281 a t 29" C. A 5.00-ml. sample was diluted to a liter and aliquots were taken for analysis as indicated in the table. Over the wide range of concentrations represented by the data in Tables I1 and I11 the agreement between the results obtained by the official gravimetric and spectrophotometric methods is excellent, particularly after they are rounded off according to standard procedure. It thus appears that for most practical purposes the Titan yellow method is sufficiently specific and accurate for magnesium in the presence of potentially interfering substances normally found in municipal water supplies. Moreover, interferences have been

November 15, 1942

ANALYTICAL EDITION

studied thoroughly (4, 7-10) and may be taken into account if necessary. The effect of varying the iron concentration was studied for two water samples containing relatively small and large amounts of magnesium, respectively, and the results are summarized in Table IV. Iron in natural waters rarely reaches concentrations which mould interfere seriously in the spectrophotometric procedure. Although the amount of sulfide in the mater samples was not determined, some of the ground waters \%-ereknown to contain considerable amounts, since they yielded colloidal sulfur on standing exposed to air. The resulting error is negligible if the sulfur is allorred to settle out.

Summary Various natural and treated waters and concentrated magnesium solutions may be analyzed rapidly and accurately for magnesium by a spectrophotometric adaptation of Kolthoff’s Titan yellow method. For most practical purposes the

897

interferences likely to be encountered do not appreciably affect the results. The procedure is recommended for control use.

Literature Cited (1) Am. Public Health Assoc., “Standard Methods for the Examination of Water and Sewage”, 8th ed., p. 79, Lancaster, Penna., Lancaster Press, 1936. (2) Becka, Jan, Biochem. Z.,233, 118 (1931). (3) Collins, W. D., Lamar, W. L., and Lohi, E. W., Geol. Survey

Water-Supply Paper 658 (1934). (4) Eilers, H., Chem. Weekblad, 24, 448 (1927). (5) Hirschfelder, A. D., and Serles, E. R., J. Bid. Chem., 104, 635 (1934). (6) Kolthoff, I. XI., Biochem. Z., 185, 344 (1927); Chem. Weekblad. 24, 254 (1927). (7) Mellan, I., “Organic Reagents in Inorganic Analysis”, pp, 19. 212, 447, Philadelphia, P. Blakiston’s Son Co., 1941. (8) Muller-Xeugliick, H. H., Gluckauf, 77, 34 (1941). (9) Schmidt, R., and Gad, G., Kleine Mitt. Jfitglied, VeT. Wasser-. Boden- u. Lufthyg., 13, 326 (1937); J . Am. Water Works Assoc., 30, 173 (1938). (10) Urbach, C., and B a d , R., Mikrochemie, 14, 343 (1934).

Determination of Rotenone Improvements in the Gravimetric Method S. I. GERTLER ’ Bureau of Entomology and Plant Quarantine, L. S. Department of Agriculture, Washington, D. C.

F

OR the determination of rotenone in plant materials the

method adopted as official by the Association of Official Agricultural Chemists ( 1 ) is most widely used. The procedure involves room-temperature extraction with chloroform, crystallization as a solvate from carbon tetrachloride solution. and purification of the solvate by allowing it to stand in alcohol. A variation developed by Jones (2), and until recently extensively used by the writer of this paper, depends upon the isolation of the rotenone as a dichloroacetic acid solvate. This variation has been found satisfactory in this laboratory, but has met with some objection from others because of the difficulty of obtaining the solvate free from interfering gummy material. I n order t o get a shorter and simpler procedure the official method has been varied with respect t o the mode of extraction and purification of the solvate.

Procedure for Powdered Roots Although parts of the official method are duplicated more or less, for the sake of continuity and ease of following the method through, the procedure will be completely outlined. Weigh a sample of finely powdered root of such size that the quantity of extract finally obtained and used for analysis wdl contain at least 1 gram of rotenone, This will not be possible in the case of roots of exceptionally low rotenone content, and the method is modified accordingly, as noted later. Transfer the powder to a 1-liter round-bottomed flask, preferably one having a standard glass joint. Add 10 grams of decolorizing carbon and then an accurately measured volume of chloroform a t some definite (near room) temperature in the ratioof 1Oml. toeach gram of root used. Connect the flask to an efficient reflux condenserthat is, one shown to be able to conserve all the chloroform-and boil the contents for 1 hour. If there is any question as to the efficiency of the condenser, the flask and contents should be weighed before heating and any loss in weight made up by the addition of fresh chloroform. As soon as the condenser has drained, remove the flask, immediately stopper it, and cool it first under the tap and then in the refrigerator for 30 minutes. Filter the contents rapidly through a fluted filter, keeping the funnel covered with a watch glass to avoid evaporation of the solvent. Bring the filtrate back to the original temperature by immersing the flask in warm water and shaking until the chosen tempera-

ture is reached. Measure the volume of the filtrate in a graduate to the nearest milliliter, and calculate the proportion of the sample that it represents. Transfer the chloroform extract to a round-bottomed flask, distill until about 25 ml. remain, and then transfer this remainder to a 125-m1. Erlenmeyer flask with a little chloroform and distill almobt to dryness; in this may most of the solvent is recovered. Remove the last traces of solvent by warming slightly under diminished pressure and evaporating twice more after adding two small portions of carbon tetrachloride. Dissolve the residue by warming under a reflux condenser with exactly 25 ml. of carbon tetrachloride, cool, and seed if necessary to induce crystallization. If a t this point crystallization does not take place readily or occurs only in small amount because of low rotenone content, add an accurately weighed quantit of pure rotenone to bring the amount up to at least 1 gram, andrkeep the flask at 0” C. overnight. Pour the contents of the flask quickly through a small frittedor sintered-glass funnel of medium porosity, allowing the flask to drain as completely as possible. (These glass funnels and also crucibles have been found very useful by the author and can be cleaned easily with acetone and used repeatedly.) Continue the suction for about 5 minutes after the solvent has drained off. Place the flask in which the solvate was crystallized, and which still contains some crystals adhering to the sides, under the funnel containing the filtered solvate. Dissolve the solvate by pouring small portions of acetone through the funnel, about 25 ml. usually being sufficient. Evaporate the acetone solution of the solvate to dryness on the steam bath, and then place it under gentle suction for a short time to remove all traces of acetone. Treat the residue from the acetone with 25 ml. of ethyl alcohol saturated with rotenone a t room temperature, and boil the contents, still in the flask originally used for crystallization of the solvate, gently under reflux until completely dissolved. If necessary, add more alcohol to effect complete solution. Allow the alcohol solution to come to room temperature. Usually crystallization begins readily, but it may be induced, if necessary, by adding a crystal of pure rotenone; it then usually proceeds very rapidly. Shake the flask vigorously for 2 full minutes, and allow to stand at room temperature for 30 minutes. Filter the rotenone crystals through a weighed sintered-glass or Gooch crucible, using about 10 to 15 ml. of alcohol saturated Kith rotenone at room temperature to transfer and wash them. After aspiration for 3 to 5 minutes dry to constant weight at 105” C., xhich takes about 30 minutes. Weigh the final product, which consists of purified rotenone, add 0.07 gram to correct for the solubility in 25 ml. of carbon tetrachloride, and subtract any amount of pure rotenone that may have been previously added. Multiply the net weight by 100 and