370
‘
ANALYTICAL CHEMISTRY
Several crystals of phenol and 5 t o 6 drops of acetic or propionic (B) anhydride are added and the mixture is warmed to obtain solution. Hydriodic acid is added t o the second reaction flask, b, so that the bulb portion is about one-half full, and the flask is put into place. Equal parts of 5% solutions of cadmium sulfate and sodium thiosulfate are added to the scrubbing compartment, d, t o fill it half way. Kater is run into tube f through ground joint k , and the top is closed quickly.so as to have the constrictions filled with water. The source of carbon dioxide is attached at ground joint h, and the gas is bubbled through for a couple of minutes until the excess water in tube f has been forced out. Two cubic centimeters of 3.85y0 alcoholic silver nitrate are placed in test tube g and the tube is put into place so that delivery tube f goes t o the very bottom. The ground joint at h is opened and tinfoil about the size of a dime is placed in reaction flask a, and then enough hydriodic acid t o fill the bulb part of the flask approximately half full. A bumping tube of the Pregl type may be inserted into a if desired but is not necessary, as the platinum boat reduces bumping. The source of carbon dioside is immediately attached. The flow of gas is iegulated so that the bubbles pass through the silver nitrate in test tube g a t the rate of about one bubble per second. Water is run through condenser c and a microburner is used to heat the acid in b to gentle boiling. (Caution. Fuming acid is formed upon pasbing gas through at room temperature, 7 . ) I n this condition, the apparatus is allom-ed to stand for 30 minutes. The mixture in a is then brought to boiling by means of a microflame and boilin’g of the contents of a and b is continued for one-half hour. -4fter this period, the water is drained from the condenser and boiling of the contents of both flasks is continued for at least an additional half hour. The stopper a t ground joint k is then opened, test tube g lon-ered, and the precipitate washed into the flask alternately by means of 1 to 200 nitric acid and alcohol. Then 1 to 200 nitric acid is added t o the contents of the test tube g until it is about four-fifths full, follom-ed by 5 drops of concentrated nitric acid, specific gravity 1.42. The contents of the tube are then brought just to the boiling point on a steam bath, immediately cooled, and filtered, using a Pregl filter tube and suction flask. The silver iodide precipitate is vashed with 1 to 200 nitric acid and then with cold concentrated nitric acid, specific gravity 1.42, by filling up the filter tube with the concentrated acid, allowing i t to soak through for several minutes, and then sucking off, This should be repeated several times. The precipitate is then R-ashed with 1 to 200 nitric acid followed by alcohol and dried in an oven at 120” C., after which the precipitate is weighed. Silver iodide (0.120 mg.) is added to the weight of the precipitate t o give the corrected value before calculation ( 3 , 5 , 9 , I I ,121.
ACKNOBLEDGMEST
The author is indebted to the following persons for their ryork in connection with this paper. The photograph of the setup (Figure 1 ) \vas made by Lester Hodas, n-hile the sket,ch of the apparatus (Figdre 2) \vas made by W.D. O’Connor, Jr. The apparatus used was fabricated by John Deakin and Emil Wiegand. The analpes s1ion.n in Table I \\-ere dolie by Esther Bass, Samuel Blackman, Claire Farkas, Janet Farnon-, Marian Faulkner, Gerhard IGsch, Bella Littnian, 3Iargaret Sullivan, and llarie Walker. -111 the above-mentioned are in t,he employ of this company or !yere at the time of their Iyork. LITERATURE CITED
(1) Clarke, E. P., “Semimicro Quantitative Organic Analysis,” p
70, Xew York, Academic Press, 1943. (2) Elek, A, “Microdetermination of Alkoxy1 Groups,” IXD. Eso. CHEY.,.~N.~L. ED.,11, 174 (1939). (3) Friedrich, 4.,“Die Praxis der quantitativen organischen Mikroanalme.” DD. 133-60. 151-9. Leinsie and Tlenna. F. Deuticke. 1933”; Z.*phUsiol. Chem., 163, i 4 1 (1927) ; Mikrochemie, 7; 185, 195 (1929). (4) Furter, XI. F., Helv. Chim. Beta, 21, 1144, 1151 (1938). ( 5 ) Grant, J., “Quantitative Organic Microanalysis,” 4th English ed.. pp. 146-52. Philadelphia, Blakiston Co., 1946. (6) Knoll, Iqlexander, private c&nmunication, work done at Columbia University. (7) Mellor, J. IT., “Comprehensive Treatise on Inorganic and Theoretical Chemistry,” second impression, 1-01. 2, p. 190, London, Longmans, Green and Co., 1927. (8) Xiederl, J. B., and Niederl, Y.,“JIicromethods of Quantitative Organic Analysis,” 2nd ed., p. 241, Sew York, John Wiley & Sons, 1942. (9) h’iederl, J. B., and Siederl, Y.,“Micromethods of Quantitative Organic Elementary Analysis,” pp. 187-92, Nev York, John Wiley 8: Sons, 1938. (10) Pregl, F., “Quantitative Organic Microanalysis,” tr. by Ernest Fyleman. pp. 150-8, London, 3. 8: h.Churchill, 1924. (11) Roth, H., ”Die quantitat,ive organische Mikroanalyse von Frits Pregl,” 4th ed., pp. 210-19, Berlin, Julius Springer, 1935. (12) Roth. H., “Quantitative Organic Microanalysis of Fritz Pregl.” 3rd English ed., tr. from 4th revised German ed. by E. B. Daw, pp. 171-9, Philadelphia, P. Blakiston’s Son & Co., 1937. RECEIVED July 12, 1947. Presented before t h e Division of Analytical and Micro Chemistry a t the 112th lfeeting of the . i J r E R x C a N CHEMICAL SOCIETY, New York. N. Y.
Colorimetric Microdetermination of Zirconium DAVID E. GREEN, Filtrol Corporation, Vernon, Calif. This paper comprises the development of an accurate colorimetric method for determination of zirconium in clays or silicate rocks, using the pink lake formed by the zirconium alizarin sulfonate complex. The method applies to a range of zirconium oxide content up to 0.255 mg. with an accuracy to 0.003 mg. of zirconium oxide.
A
X U R G E S C l existed for developing a n accurate determination of small amounts of zirconium in naturally occurring clays and their activated products. The gravimetric method by precipitation of zirconium as a phosphate (3) usually employed in silicate rock analysis, or as a selenite.(b) in the steel industry, fails t o detect zirconium oxide in amounts less than O.Ol%, using a t least a 1-gram sample. The steel industry developed a rough colorimetric determination of zirconium using p-dimet hylaminoazophenylarsonic acid ( 2 ) which permits slightly greater accuracy (to 0.005% zirconium). However, as stated by the authors, very apparent difficulties arise in the standardization of this method. Recent spectrographic techniques ( 4 ) also give comparable accuracies to 0.005rc zirconium. Consequently, esisting qualitative tests for zirconium as compiled by Feigl ( I ) and as given by Yoc and Overholser (IO) were examined for possible quantitative development. The zirconium
alizarin sulfonate lake was selected as a color complex least subject t o interference, and specific in its reaction for clay analysis. It proved t o be entirely satisfactory and very accurate, and fills a need in analytical chemistry for a determination of zirconium oxide t o an accuracy of 0.00065. EXPERIMENTAL
Development. The zirconium alizarin colors in this investigation were read with a Coleman Universal spectrophotometer, Model 11, and the p H values with a Beckman p H meter. The optical densities recorded by thm spectrophotometer are the logarithms to base 10 of the ratio of the intensity of the incident light t o that of the transmitted light. The test solutions were made up to 100 ml in volumetric flasks. A stock solution of 0.05% sodium alizarm sulfonate and a standard solution of zirconium oxychloride (1 ml. equivalent to 0.05 mg. of zirconium oxide), adjusted to a p H of 2.2 Kith hydrochloric acid were utilized.
V O L U M E 20, NO. 4, A P R I L 1 9 4 8 Table I.
371
Solution Densities with Interfering Substances
(To 100 nil., 1.00 mg. of sodium alizarin sulfonate, color developed in 1 hour, 5200 A.) Interfering Substances Added t o Test Solution, Mg. .11ClrGH*O FeCla TiClr as Ti02 T h ( I i 0 ~ ) c . l H z Oas ThOr
474.0 50.7 1.5 0.2
Test Condition, 5 >Il. of 1 N HC1 Added t o Give pH 1.5 Blank Al+++ F e + + + Ti + + + Tht+'0.004 0.023 0.118 0.015 0 . 0 0 4 0.11 mg. of ZrO2added 0.065 0.080 0.112 0.068 0.066 Test Condition, 10 M1. of 1 N HC1 Added to Give p H 1.1 0 . 0 0 4 0.004 0.079 0.004 0.065 0.066 0.088 0.066 0 . 1 1 nig. of ZrOz added
0.004 0.063
Test Condition, p H 1.1, F e + + t Reduced in Silver Reductor Fe++ 0.004 0.004 0.089 0.004 0.004 0 . 1 1 nig. of ZrOs added 0.066 0.065 0.108 0.065 0 . 0 6 5
0.004 0.066
conium complex, these two transmittance curves gave the poiGt of minimum transmittance for the zirconium pink lake a t 5200 A. For convenience, the transmittance curve for 1.00 mg. of sodium alizarin sulfonate and no zirconium oxide at a pH of 1.5 is plotted in Figure 1 along with the curves containing variable amounts of alizarin and 0.275 mg. of zirconium oxide. I n subsequent determinations, a 1.00-mg. addition of alizarin was selected not only to eliminate interfeience in reading densities of the developed lake from the yelloxv of the alizarin but also to peimit full development of the lake by retaining thrx indicated slight excess of alizarin reagent. This excess of alizari! is shonn by the slight dip in the transmittance cyrve at 4180 A. Zirconium lake densities are then read at 5200 A.
Since the pH of the hydrolysis of a zirconium salt to the hydroxide is 2.8 ( 9 ) ,an initial survey was made in t,he acidic range from 1.5 to 4.0 pH t,o determine the effect of pH on the color of a O . O O l ~ csodium alizarin sulfonate solution. I t was found that change iu solution pH over this pH range had only a negligible effect on the density of the yelloy alizarin color a t the point of niinimuin transmittance of 4180 -4.The pH of the solutions \vas adjusted with hydrochloric acid. Wit,li the pI-1 constant a t 1.5 and the zirconium oxide content at 0.275 mg., transmittance curves were run with varying amounts ol sodium alizarin sulfonate. Since the alizarin a t the two lowest pcrrentagcs \vas entirely consumed in the formation of the zir-
Figure 2.
Colorimetric Zirconium Standard Curve Density
Figure 1. Transmittance Curves of Zirconium-Alizarin Complex Solutions containing 0.275 mg. of zirconium oxide with varying amounts of sodium alizarin Sulfonate
US.
mg. of zirconium oxide
Standard Curve. Test solutions (100 ml.) were made up with additions of zirconium from the standard zirconium oxychloride solution. The zirconiu? lake color densities were read a t intervals of time a t 5200 A . Color development is complete in 1 hour and is stable.for at least 4 hours. However, upon standing overnight, the lake precipitated and settled out. The zirconium alizarin sulfonate lake follows Beer's lair. Zirconium oxide contents over 0.275 mg. show insufficient alizarin to permit the full development of the color complex. The standard curve of color density versus milligrams of zirconium oxide is plotted in Figure 2. Interfering Substances. The following substances (6) in solution are considered to interfere with the formztion of the zirconium lake and the measurement of its density: fluorine sulfates, phosphates, silicon, molybdenum, antimony, tungsten, and organic hydroxy acids. These substances for meither com-
ANALYTICAL CHEMISTRY
372
plex radicals or precipitates with zirconium ( 1 ) . However, they were not investigated because they were not present in the clay material, were eliminated by the outlined analytical method, or occurred in such small quantities that the procedure a$ outlined eliminated their interference. , Interference from aluminum, ferric iron, titanium, and thorium, which cause coloration with sodium alizarin sulfonate, was investigated. Acid conditions in the test solutions eliminated interference from thorium, and the lowering of the p H t o 1.1 by the addition of 10 ml. of 1 N hydrochloric acid from titanium and aluminum. Iron interference was overcome by reduction from the ferric to the ferrous state by passing the solution through a silver reductor (8) of 12-ml. capacity. I n these tests, quantitiex of aluminum, iron, and titanium were taken to approximate the analysis of the clays using a 0.5-gram sample-namely, 20, 5 , and 0.37,, respectively, calculated to the oxide. The results art’ recordcd in Table I
Procedure. A O.5-grani *ample of clay is weighed into a 20gram nickel crucible, approximately 4 grams of sodium hydroxide pellets are added, and the fusion is made. The completed fusion is then digested on a hot plate with distilled water. The insoluble portion of the digestion is filtered out with a 9-cm. No. 40 Whatman filter paper and thoroughly washed with distilled water. Since the insoluble will contain the sodium zirconate, the filtrate is rejected. The insoluble is digested with about 12 ml. of hot concentrated hydrochloric acid and washed with hot distilled water into a 100-ml. beaker. The total amount of solution should not exceed 75 ml. The filter paper with residue is retained for further analysis. The solution is cooled to room temperature, 2 drops of phenolphthalein indicator are added, and the solution is neutralized cautiously, not permitting the temperature to rise, with a 50% sodium hydroxide solution. Five milliliters of 1 N hydrochloric acid are added, and the mixture is passed through a silver reductor of 12-ml. capacity. The reductor is rinsed out with 5 ml of 1 -4- hydrochloric acid and 5 ml. of distilled water. The reduced solution is received in a 100-ml. volumetric flask, 2 ml. of 0.05% sodium sulfonate are added, and the solution is brought up to 100-ml. volume with distilled water. The final solution is mixed well by shaking. The pH should be 1.1. I t is allowed to stand 20 hours or overnight fgr color development. The density of the solution is read a t 5200 A. The filter paper from the hot hydrochloric acid digestion of the sodium zirconate is burned a t 925” C., and the residue is fused with a few pellets of sodium hydroxide. This second fusion is digested and filtered in the same manner as the initial fusion. The insoluble residue is similarly leached with hydrochloric acid and washed. The resulting leachate is also adjusted to a p H of 1.1. The small amount of iron present permits the omission of the reduction step. The number of milligrams of zirconium oxide contained in the two solutions are taken from the previously plotted standardization curve of density versus milligrams of zirconium oxide and added for the final zirconium oxide content of the sample. The method is accurate to 0.000640 zirconium oxide with the 0.5-gram sample used.
Table 11. Triple Fusions Cndried, 0.5-Gram Sample, Cheto, biz., Clay ZrOl, Lknaitz. 0.09; 0.008 0 003
1st f U S l 0 l l
2nd fusion 3rd fusion Triple fusiori toral
0.010
0.000 0174
0.125
% ZrOi found % ZrOz certified ab-.
0.0250
0.041
Table 111. Density Readings Yndried 0.5-Grain Pampie, %hero, Ariz., Clay ZrOx, Single fusion 0.055 mg. of %vOr
added
1)enaitz.
mg.
0.098
0.163
0.128
0 219
Dried a t 140’ C., 2 Hours, 0.5-Gram Sample, Bureau of Standards Clay 98 ZrOz, Density mg. 0.072 0.121 0 103 0.176
It is must important in thr procedurr t o ueutralize th6 hydroohloric acid, used in the digestion of tke sodium zirconate, a t room temperature, t o form zirconium hydroxide, Zr(OH)(; other\rise, a t higher temperatures thP acid-insoluble xirconyl hydroxide, %rO(OH)Q( 7 ) ,will be formed. Color Development. Zirconium lake color development is rapid but limited to 4 hours for the standard and test. interfering substance solutions. However, the solutions made up with the clay samples show a very slov development’ in color and remain stable for at least 30 hours before precipitation takes place, The high concentration of sodium chloride in these solutions possibly influences this behavior. \Then a standard zirconiuni solution is added to a clay solution sample in which the color has already been developod, the development of additional color is also d o n . Density readings taken after addition of this standard zirconium solution were exactly equivalent to the amnunt of zirconium oxide added, a8 shown in Table 111. CONCLUSION
This nesvly developed colorimetric method for the quantitative determination of zirconium, by using the pink lake formed by the zirconium alizarin sulfonate complex, permits the detection of as little as 0.003 mg. of zirconium oxide in clays. It is believed that this investigation has laid the foundation for the accurat,e determinat,ion of small amounts of zirconium in rock analysis. With possible modifications, it undoubtedly can he extended to advantage in other fields of chemistry.
DISCUSSION
Dissolution of Sample. I n clay material, zirconium usually occurs as a silicate, the mineral zircon, which is broken down and put in solution with difficulty. Sodium carbonate fusions of a clay which contained about 0.035% zirconium oxide gave only about 60% of the zirconium oxide found by using a sodium hydroxide fusion. This substantiates the facts as given by Venable (7). . Triple sodium hydroxide fusions were made on Natiorial Bureau of Standards plastic clay 98 and a raw Cheto, Ariz., clay. The third fusion of the residue from the hydrochloric acid digestion of the sodium zirconate shows (Table 11) no additional zirconium oxide. Therefore, two fusions must be made on every sample to ensure complete breakdown of the zirconium compounds present. Plastic clay 98 was found t o contain 0.02507c zirconium oxide. The Bureau of Standards certificate of analysi. shows an average of 0.041% zirconium oxide or variation in separate analyses between 0.032 and 0.0570.
mg.
0.164
Dried a t 140° C., 2 Hours, 0.5-Gram Sample, Bureau of Standards Clay 98 ZrO:, Density mg. 0.068 0.114 0.009 0.011 0,003 0.000
LITERATURE CITED
b’eigl, F.,“Qualitative Analysis by Spot Test,” p. 124-9, New York, Nordemann Publishing Go., 1939. Hayes, W. G., and Jones, E. W., I N D .ENG.CBEM.,ANAL.ED., 13,603 (1941).
Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” pp. 442-51, New York, John Wiley &- Sons, 1929. Pierce, W. C., Torres, 0. R., and Marshall, W. W., IND.ENG. CIIEM.,AKAL.ED.,12,44 (1940).
Simpson, S. G., with Schumb, W. C., Ibid.,’5, 40 (1933). Smith, L., and West, E’. W., Ibid., 13, 271 (1941). Venable, F. P., “Zirconium and Its Compounds,” A.C.S. Monograph, pp. 34-41, New York, Chemical Catalog Co., 1922.
Walden, G. H., Jr., Haminett, L. P., arid Edmonds, S. M., J. Am. Chem. SOC.,56,350 (1934). Willard, H. H., and Diehl, H., “Advanced Quantitative Analysis,” p. 45, New York, D. Van Nostrand Go., 1943. Y o e , J. H., and Orerholser, L. G., ISD. ENG.CHEM.,AXIL. ED.,
.
15, 73 (1943).
RECEIVED April 26, 1947.
