\\ere analyzed and the results are reported in Table 11. These results show that all the hydrogen added to the sample was extracted. This experiment also indicated that no significant quantity of hydrogen was absorbed by the metallic mirror. This point was completely demonstrated by introducing a known quantity of hydrogen into the furnace tube n-hen the bath was a t 670" C. and a tin mirror was present. The hydrogen was pumped out and again measured. The difference between the quantity introduced and that collected 11as 0.32c, which is the same magnitude as the error in measuring the quantity of hydrogen. The mass spectrographic analysis of the extracted gas shon ed the reliability of the assumption that only hydrogen is evolved from the sample. The possibility of error due to evolution of gases other than hydrogen is therefore
excluded for calcium samples of high purity, Samples of calcium which contained significant amounts of CaCOa and Ca(OH)? would probably evolve some COZ or CO and water when heated under vacuum and, in this case, it would be necessary to analyze the evolved gas to determine the quantity of hydrogen. Operation of the tin bath a t lower temperatures was investigated with the hope of decreasing vaporization from the bath. A temperature of 550" C. gave accurate results for low hydrogen samples but samples containing more than 0.1 wt. yo hydrogen gave low values. On heating a tin bath from which the evolution of hydrogen had ceased at 550" C., additional hydrogen was evolved beginning a t 640" C. A temperature of 670" C. seems to be the lowest temperature a t which rapid and complete evolution of hydrogen is
achieved. Higher temperatures resulted in more rapid vaporization of metal from the bath and deterioration of the furnace tube. LITERATURE CITED
(1) Bastien, P., Fonderie No. 92, 3570-98 (1953). (2) Bergstresser, K. S., Waterbury, G.
R., U. S. Atomic Energy Comm. Rept. LAMS 1698 (1954). (3) Frazer, J. W., Schoenfelder, C. W.,
Ibid., UCRL 4944 (1957). (4) Mallett, M. W., Gerda, A. F., Griffith, C. B., ANAL.CHEM.25, 116-19 (1953). (5) Yreaton, R. -4~, Vacuum Vol. 11, 115 (1902).
RECEIVED for review September 21, 1961. Accepted January 15,1962. Contribution No. 1060. Work performed in the Amea Laboratory of the U. S. Atomic Energy Commission.
Forms of Aluminum Determined by an 8-QuinoIinoIate Extractio n Method SIR: In low-acidity solutions, aluminum ions undergo polymerization to form polynuclear hydroxoaquo complexes (2, 6, 6) with aluminum atoms probably linked to one another with diol bridges as in
It seems likely that a series of polynuclear ions with different degrees of polymerization, ranging from monomer to colloidal hydrated alumina, exist in solution in equilibrium (6). I n spite of its great importance, few attempts have been made to determine any particular form of aluminum among this series of complexes, except Tanaka ( 7 ) who used an ion exchange technique to separste ionic aluminum. Recently, however, a study of the chloroform extraction of aluminum 8-quinolinolate showed that higher molecular weight aluminum ions are not extracted as the 8-quinolinolate (1). Thus, an extraction method has been described for the determination of lower molecular weight aluminum ions in the presence of higher ones, but no clear-cut distinction has been made between the forms of extractable and unextractable aluminum.
Jahr and Brechlin ( 2 ) prepared a series of basic aluminum nitrate solutions and determined the mean degrees of polymerization of aluminum ions in these solutions from salt cryoscopic measurements (3, 4), in which eutectic point depressions of the KNOa-HzO system were measured as in the usual cryoscopy. The salt cryoscopic method has an outstanding feature in that ions common to those comprising the eutectic system do not produce any depression of the eutectic point. It is thus possible with the KN03-H20 system to determine the degree of polymerization of aluminum ions with considerable precision even in the presence of a large amount of potassium and nitrate ions. The present investigation was designed to show, with the help of the salt cryoscopy, that only aluminum ions with the degree of polymerization a t or very close to unity are determined by extraction. EXPERIMENTAL
Reagents. Basic Aluminum Xitrate Solutions. A series of basic aluminum nitrate solutions with basicities (molar ratio of O H to Al) ranging from 0.5 to 2.0 were prepared according to Jahr and Brechlin ( 2 ) . Aluminum nitrate solution was first prepared from the purest grade of aluminum nitrate nonahydrate. The basicity of this solution was adjusted
exactly to zero by adding a required volume of nitric acid. Basic aluminum nitrate solutions were prepared from this solution under vigorous stirring a t about 50" C., by adding calculated volumes of 4'37 potassium hydroxide solution through a capillary. 8-Quinolinol Solution. Dissolve 2.0 grams of 8-quinolinol in 5 ml. of glacial acetic acid and dilute to 200 ml. with distilled water. All chemicals used were of purest grade and supplied, unless otherwise stated, by Kanto Chemicals Co., Japan. Extraction of Aluminum 8-Quinolinolate. Pour 10 ml. of distilled water and 2-ml. portions of 8quinolinol and sodium acetate (1M) in a separating funnel. Add to this exactly 10 ml. of chloroform, being careful not to disturb the aqueous layer. Add a n appropriate volume of sample, containing less than 0.08 mg. of extractable aluminum. Immediately thereafter shake the funnel vigorously for 10 seconds ( I ) , and allow the mixture to separate into two layers. Run the chloroform layer into a test tube containing a small amount of anhydrous sodium sulfate crystals (Koso Chemicals Co.) to remove droplets of water in the chloroform layer. The amount of aluminum extracted is determined from the absorbance of the chloroform layer a t 390 to 420 mF, depending on the amount extracted. Any delay between the addition of sample and the separation of chloroform layer will cause considerable depolymerization of higher aluminum complexes to take place. VOL. 34, NO. 4, APRIL 1962
581
Cryoscopic Measurements. All cryostopic measurements were made in a way similar t o t h a t reported b y Jahr et al. (2, 3).
the DP of unextractable aluminum a t basicity 0.5, when extractable aluminum was assumed to be monomer. Such a negative value may be due t o the presence of aluminum in such a form which is rapidly depolymerized during the course of extraction. At this basicity] however, the amount of unextractable aluminum is very small so that the calculated D P of the latter is not sufficiently accurate. When the D P of extractable aluminum is assumed t o be 2, the calculated DP of unextractable aluminum levels off a t certain basicities below 2, the assumed D P of extractable aluminum. This is unreasonable. Thus there is no doubt that the extracted aluminum is of a very Ion degree of polymerization. The exact value of the mean D P of unestractable aluminum should not be greater than the DP value calculated on the assumption that extractable aluminum is monomer. It is conceivable, therefore, that aluminum ion 11ith degrees of polymerization greater than the smallest of the figures in the second column of Table I1 are not extracted. Such a consideration reveals that aluminum ions larger that hexamer can no longer be determined by extraction. I n contrast, the mean DP of estractable aluminum was also calculated by assuming values of 2 and 3 for the D P of unextractable aluminum (Table 11). Examination of this table clearly shows that a t low basicities] the calculated values are very close to unity if it is assumed that the mean DP of unextractable aluminum is greater than 3. At higher basicities] the calculated D P is somewhat higher. Since, however, the mean D P of the unextractable aluminum should not be smaller than the DP value averaged for all aluminum species, it will be easily seen that the mean DP of unextractable aluminum should be greater than 4 a t basicity 2.0. With this figure for the DP of unestractable aluminum, 2.48 is obtained for the mean DP of extract-
RESULTS AND DISCUSSION
Eutectic point depression and the concentration of extractable aluminum were determined a t each basicity a t concentrations of total aluminum 0.1 to 1.0 mole per liter. Apparent molar depressions were obtained from the eutectic point depressions by extrapolation as in the usual way. The results were accurate to within less than 0.02" C. The ratio of extractable to unextractable aluminum was found to be independent of the concentration of aluminum at a given basicity. The reproducibility of the determination of extractable aluminum was about =t2%. The results are summarized in Table I. The mean degrees of polymerization, DP, were calculated from the apparent molar depressions on the assumption that all aluminum is in the monomeric state a t zero basicity. These figures are fairly consistent with tho,qe reported by Jahr and Brechlin ( 2 ) . From the mean D P and the amount of extractable aluminum, calculations m-ere made for the mean D P of unextractable aluminum on the assumption that extractable aluminum is nionomer/or dimer. The results are shown in Table 11. -4negative value ivas obtained for Table 1. Apparent Molar Depression, Mean DP, and Proportion of Extractable Aluminum
Apparent Molar DepresBasion, sicity O C. 0 0.1 0.5 1.0 1.5 2.0
1.70 1.70 1.33 1.18 0.91 0.48
Table II.
1.00 1.00 1.28 1.44 1.87 3.54
100 100 95.8 65.6 43.7 20.8
Calculated Values for DP of Unextractable and Extractable Aluminum D P of Extractable A1 D P of Unextractable Al
Basicity 0.5 I .O 1.5 2.0
Extractable Mean illuminum, DP %
Assumption Extractable A1Monomer Dimer -0.24 8.82 5.68 9.70
1.73 1.83 2.46 4.98
Assumption Unextractable A1 Dimer Trimer Octamer
.
1.26 1.25 1.il -1.82
1.25 1.13 1.26 11.0 ~~~
Table 111.
~-
Ion Exchange Behavior of Aluminum Ions in Basic Aluminum Salt Solutions
Sample Basic aluminum chloride, basicity 2 . 0 Basic aluminum nitrate, basicity 2 . 1 9 582
~
1.25 1.24 0.94 1.12
ANALYTICAL CHEMISTRY
Al, Mg./Liter Extractable Unextractable Before ion After ion Before ion After ion exchange exchange exchange exchange 5.6
0
16.4
4.5
0.5
0
4.9
2.5
able aluminum. Therefore, there is no doubt that even a t basicity 2.0, the exact value of the mean D P of extractable aluminum lies below 2.48. Moreover, if the unextractable aluminum is assumed to be octamer, as has been concluded by Matijevit et al. ( 5 ) , the calculated D P of extractable aluminum becomes very close to unity a t all basicities, as will be seen in the last column of Table 11. Therefore] it seems very likely that practically monomer alone is determined by the extraction provided that the extraction period is sufficiently short to minimize depolymerization of polynuclear complexes. The aluminum referred to as ionic by Tanaka (7') can further be classified into groups, since in certain cases, much of the unextractable aluminum is removed by ion exchange (Table 111). Solutions of basic aluminum d t s were analyzed for eatractable and unextractable aluminum bcfore and after passage through a column of ion exchange resin (Amberlite IR 120, hydrogen form). The ion exchange treatment releases hydrogen ion n hich may cause the equilibrium betnecii polymeric and monomeric forms to be displaced. In a preliminary investigation, however. it was found that the depolymerization of the polymeric forms is rather slow a t low acidities. Therefore it is deemed that no significant change in the form of aluminum has taken place during the course of ion exchange. Application of the extraction method reveals that many colorimetric methods currently available for the determination of aluminum involve changes in the form of aluminum during the course of determination, thus giving rise to a lack of reproducibility. Details of such a n investigation rvill be published later. LITERATURE CITED
(1) Goto, K., Ochi, H., Okura, T., Bull. Chern. SOC.Japan 31, 783 (1958). (2) Jahr, K. F., Brechlin, il., 2 anorg. u. allgem. Chem. 270,257 (1952). (3) Jahr, K . F., Brechlin, A., Blanke, M., Kubens, R., Zbid., 270, 240 (1952). (4) Jahr, K. F., Kubens, R., 2. Elektrochern. 56, 65 (1952). (5) MatijeviC, E., Mathai, B.G., Ottewill, R. H., Kerker, M.,J . Phys. C'hem. 6 5 , 826 (1961). (6) Tanabe, H., Yakugabuzasshi ( J . Pharm. SOC.Japan) 74,253, 866 (1954); 75,954 (1955); 77,33 (1957). (7) Tanaka, M., Bull. Cheni. SOC.Japan 27, 98 (1954).
TAKESHI OKURA' KATSCW GOTO TAKAO YOT~YANAGI Faculty of Engineering Hokkaido University Sapporo, Japan Deceased. RECEIVEDfor review March 7 , 1961. Accepted December 21, 1961. Taken from a thesis submitted by Takao Yotuvanagi in partial fulfillment of the requirements for the Bachelor's degree in engineering at Hokkaido University, 1961.
