ENGINEERING RESEARCE INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR PROGRESS REPORT APPLICATION OF THE POLAROGRAPE TO ANALYSIS OF TITANIUM-BASE ALLOYS CIoAR~.,; 3'','', Project 2075 AIR RESEARCH AND DEVELOPMENT COMMAND, U. S. AIR FORCE CONTRACT NO. AF 18(600)-397, E.O. NO. R606-60O- R3Z December, 1952

;r -

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN:PRORESS REPORT APPLICATION OF THE POLAROGRAPH TO ANALY$TIS OF TITANIUM-BASE ALLOYS SUMMARY The present report covers the period from October 7 to December 7, 1952. During this period, ferric and chromic ions were studied polarographically in the presence of titanic ion. Potassium fluoride was employed as the supporting electrolyte in all these solutions, since titanium metal and titanium-base alloys can be dissolved easily in hydrofluoric acid. In addition, ferric and chromic ions seem to have well-defined waves in fluoride medium under the right pH condition. SUMMARY OF RESULTS A. Preliminary Investigation Heyrovskyl found the half-wave potential, E1/21 for Fe+3 in 0.1 M KRF2 solution to be about -0.6 v versus the saturated calomel electrode; in a similar medium, Zuman2 found E1/2 to be -0.57 v. We have found El1/2 for Fe+3 in a KF solution (ca. 0.3 M) at a pH of 5.0 to be -0.57 v. However, we also observed that E1/2 for Fe+3 in a 0.5 M KF solution adjusted to pH 6 with KO 0 is -1.35 v0 The latter value agrees with that of Stackelberg and Freyhold3, who reported that Fe+3 produced a reduction wave with E1/2 of -1o36 v which remained constant when the concentration of KF was increased from 0o04 to 0.8 M. The pH's of their solutions were not specified. On the other hand, West and Dean4, using a 1 M NaF supporting electrolyte in the range of 4 to 7 containing 0.004% gelatin, claimed to have found no reduction wave for Fe+3 prior to the discharge of sodium ion.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Kolthoff and Lingane5 suggest the differences in pH as an explanat;ion for these discrepancies. Apparently, E/2 is a critical function of pH; however, the possibility of decrease or of complete obliteration of the wave due to the strong ferric-fluoride complex is a more logical explanation, The ferric-ferrous-fluoride system has been investigated6; the stability constant for the ferric-fluoride complex was found to be 1012 while that of the ferrous-fluoride complex was less than 30. We also find that a concentration of less than about 1.5% (0.3 M) fluoride is insufficient to hold Fe+3 in solution when the pH is increased above 3 but not higher than 7. It is hardly possible to work with fluoride solution in glass apparatus below a pH of 5. B. Experimental Procedure In studying the polarograms of pure Fe+3 and Cr+3 ions, standard solutions of FeC13 and K2CrO7 were diluted with 0.5 M EKF solution to make approximately 0.5 mM solutions. Then 50% EKOH solution was added to adjust the pH to between 6 and 7. In the case of dichromate, a 1% solution of NE4HS03 was added to reduce the Cr+6 to Cr+3, since the chromate waves are not entirely satisfactory for polarographic analysis. In this medium Fe+3 was found tohave an 1/2 of -1.35 v and Cr+3 E1/2 of about -1.65. A mixture of Cr+3 and Fe+3 was then studied in 0.5 M EKF solution, where it was found that the two waves could be distinguished. In these solutions, the NH4HS03 does not affect the Fe+3, for the latter is tightly bound in its fluoride complex, C. Separation of Titanium as K2TiF6 The experiment of the previous section were then repeated with the introduction of titanic ion. Samples of pure TiO2 (0.4 to 0.8 g.) were dissolved in concentratod hydrofluoric acid and the same amounts of Fe+3 and Cr+6 were added. The Cr+6 was reduced to Cr+3 with NH4HSO3 and the excess bisulfite was removed by boiling the solution. The latter was concentrated to about 3 ml. In order to obtain suitable waves for Fe+3 and Cr+3, the majority of the Ti+4 must be removed, allowing us to use a higher sensitivity when recording a polarogram. The presence of a large excess of Ti+4 produces a large current which makes the precise measurement of the normal Fe+3 and Cr+3 currents practically impossible, The method used for removing titanium was based on that of Noyes and Bray7. The flow sheet diagram of Fig. 1 illustrates the process. The resulting polarograms indicated nearly quantitative removal of Ti+4, but the Cr+3 wave was also lacking. The Fe+3 wave was still observed, however. 2

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Since the Cr+3 wave appears at a very negative potential (-1.65 v) and K+ ion discharges somewhat before -2 v, it was desirable to separate these waves further. The use of Li2CO3 instead of K2C03 to neutralize the hydrofluoric acid seemed to be a possible answer, since Li+ ion has a more negative El/2 value than K+ ion. However, LiF is so extremely insoluble that Li+ ion could not be used. In another variation, 99.99% titanium metal, secured from the Foote Mineral Company was used and solutions were prepared by dissolving pure titanium, iron, and chromium metals in hydrofluoric acid and oxidizing the metals to Ti+4, Fe+3, and Cr+6 respectively with persulfate and a trace of Ag+t The excess persulfate was removed by boiling the solution, and then the solution was treated as in Fig. 1. D. Polarographic Studies of Cr+3 and Fe+3 After Ti+4 Removal The filtrates from the previous separations were studied polarographically and many curious but, unfortunately, undesirable phenomena were observed. The first major trouble was the irregular drop-time observed for the dropping mercury electrode. As the applied potential is increased to more negative values, the drop-time increases until the mercury flow ultimately becomes a stream, and the current drops nearly to zero. This effect was somewhat overcome by the addition of gelatin, but the effect varies with each solution, so a further control of variables is necessary. This irregular drop-time has been observed with four different capillaries. Most of these solutions possessed a hydrogen wave which completely obliterated the Fe+3 and Cr+3 waves. The reason for the nonreproducible phenomena sometimes observed after removal of Ti+4 may be attributable to a rather poor procedure of controlling the variables. Since the sample is dissolved in concentrated hydrofluoric acid (27 M) and then concentrated to a small volume, the amount of fluoride ion present varies from sample to sample, The amount of K2C03 necessary therefore varies with the samples. Also the final pH adjustment with KOH requires different amounts of KOH for each sample, which means that final solutions contain variable amounts of both fluoride and potassium ion, A better control over these variables is essential and will be maintained in the future. FUTURE WORK A systematic independent study of chromic and ferric ions with respect to pH, concentration of fluoride ion, and other environmental and _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _._ _ _ _ _.5.

Solid Sample TiQ2 or Ti Add HEF Solution (Add Solution K2Cr2O7 l TiF6= (colorless) FeC13)a TiF6, FeF6, Cr207= (orange) Add NH4HS03 Precipitate K2TiF6 (white) (l)Boil to Solution destroy HS035 (2)Add K2C0 iF, FeF6, r+3 (3)Filter TiF6, FeF6= Cr+3, *..............., 3 (5 )Filter xs NH4HS03 (light green) Filtrate 1 FeF6= Cr+3 (light green) Add KOH to pH 6 - 7 Solution Solution FeF6, Cr+3 (1)Add gelatin | (light green) F)Dilute to FeF6E, Cr+3 (light green) pH =6 - 7lOOml. Polarographic Test Solution a The iron and chromium would be present in an actual sample. Figure lo Preparation of Low-Titanium-Content Polarographic Solution. 4

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN experimental variables is necessary in order to obtain a standard optimum set of conditions for further work. We have not, of course, abandoned the possibility of there being a more suitable supporting electrolyte than fluoride ion. We still propose to study this system in other media such as tartarate, citrate, etc.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN BIBLIOGRAPHY 1. Heyrovsky, J., Polarographie, Springer-Verlag, Wein, Germany, 1944, p. 106. 2. Zuman, P., Collection Czechoslov. Chem. Commun., 15, 1116 (1950). 3. Von Stackelberg, M., and Von Freyhold, H., Z. Electrochem., 46, 120 (1940) 4. West, P. W., and Dean, J. F., Ind. Eng. Chem., Anal Ed., 17, 686 (1945) 5. Koltoff, I. M.,, and Lingane, J. J., Polarography, Vol. II, 2nd ed., Interscience Publishers, New York, 1952, P. 479. 6. Kolthoff, I. M., and Auerbach, C., J. Am. Chem. Soc., 17, 1452 (1952). 7. Noyes, A. A., and Bray, W. C., Qualitative Analysis for the Rare Elements, Macmillan, New York, 1943, p. 105. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~~~6 - _ _ _ _ _ _ _

UNIVERSITY OF MICHIGAN 3 9015 02826 8608