Separation of Cobalt from Zinc Sulfate Solutions by Synthetic Cation Exchange Resins

Scholars' Mine

Masters Theses Student Theses and Dissertations

1950

Separation of cobalt from zinc sulfate solutions by synthetic cation exchange resins

Wen-Hsiang Chang

Follow this and additional works at: https://scholarsmine.mst.edu/masters_theses

Part of the Metallurgy Commons Department:

Recommended Citation Chang, Wen-Hsiang, "Separation of cobalt from zinc sulfate solutions by synthetic cation exchange resins" (1950). Masters Theses. 4971. https://scholarsmine.mst.edu/masters_theses/4971

This thesis is brought to you by Scholars' Mine, a service of the Missouri S&T Library and Learning Resources. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact [email protected]. SEPARATION OF COBALT FROM ZINC SULFATE SOLUTIONS

BY SYNTHETIC CATIO EXCHANGE. SINS

WEN-HSIANG CHANG

A Thesis submitted to the faculty of the

SCHOOL OF I lNES AND METALLURGY OF TF..E UNIVERSITY OF ISSOURI in partial fulfillment of the work required for the Degree of MASTER OF SCIENCE I METALLURGICAL E GlNEERING Rolla,: Missouri 1950 ----'"

Approved by ------lo,~rI------A. W. Schlechten, Chairman,. Department of etallurgical Engineering and Mineral Dressing. ii

ACKNOWLEDGMENTS

The author is greatly indebted to the Missouri School of Mines and Metallurgy for providing the opportunities and facil1ties necessary for carrying out this investigation. To Dr. A. W. Schlechten,. Chairman of the Department of

Metallurgical Engineering and Minera.l Dressing, the author ~s deeply grateful for his invaluable consulLation,. advice,. and correction of the theslB~ Sincere thanks are aJ.so due to MI'. E. J. Breton, Jr. for setting up the experimental apparatus and also for his helpful suggestions. 11i

TABLE OF CONTE TS

Page AcknO'\"lled ents •• •• ••••••••• i1 List of Illustrations. • ••• ••••••• iv List of Tables •••• •• •••.•••••• vi Introduction ••• .••.•••••• 1 Literature Review • ••••. •• 4 Theoretical Considerations •••••••• 15 EXperimental Apparatus, Equipment', and Materials 31 Experimental Procedure. •••••••• ••• 35 (1) Chemical Analyses ••••••• 35 (2) Operation •••••••••••••• 39 Experimental Results ••••••••••••••• 43 Discussion •••••• •• •• ••••••••• 80 Conclusions • ••••••••••••• 87 Summ.a.ry. •••.•••••••••.•• ••••• 89 Bibliography •••••••••••••••••••• 90 Vita .•...... 92 iv

LIST OF ILLUSTRATIONS Page

Photo. No.1 The Experimental Set-up • • • • • • ••• 31 Fig. 1 Amberlite IRC-50 ...... 15 Fig. 2 Amberlite IR-120 • ...... 16 Fig. 3 Ionic Radius and Free Energy of Ion Exchange • 19 Fig. 4 Ionic Radius and Extent of Exchange •••• • 21 Fig. 5 Adsorption of Cations by Organic Ion Exchangers 26 Fig. 6 Diagram of Experimental Bet-up •••••••• 32 Fig. 7 Exhaustion Cycle of IRC-50 at lower pH •• 48 Fig. 8 Exhaustion Cycle of IRC-50 at Higher pH . . . 50

Fig. 9 Exhaustion Cycle of IR-120 . .. • •• • • •• 52 Fig. 10 Elution with H2S0A.0n IRC-50 (Eluate Cone. vs Eluate Volume) ••• 54 Fig. 11 Elution with H2S04 on IRC-50 (% of Total Eluted vs Eluate Volume) · . ... 55 Fig. 12 Elution with 0.5M NH4Cl on IRC-50 (Eluate cone. va El ate Volume) ••• • • • • • 57 Fig. 13 Elution with 0.51< H4Cl on IRC-50 (% of lotal El ted va Eluate 01 me) ·. . . . 58 Fig. 14 Elution with 1M'NH4Cl on IRC-50 (Eluate Cone. VB Eluate olume) ••• ·... . 60 Fig. 15 Elution with 1M NH4Cl on IRC-50 (% of ~Otal Eluted vs Eluate Volume) ·. . .. 61 Fi • 16 Elution with O.5M NaCl on IRC-50 (Eluate Cone. vs Eluate Volume). . . · . .. . 63 Fig. 17 Elution '\-lit _ 0.5M NaCl on IRC-50 (% of Total Eluted va Eluate olume) · .. .. 64 Fig. 18 Elution with 0.5M NaCl on IRC-50 at Higher Flow Rate (Eluate Cone. va Elute Volume) • 66

Fig. 19 Elution with O.5!~ NaCl on IRC-50 at Higher Flow Rate (% of Total Eluted va Eluate Volume) 67 v

Fig. 20 Elution with 0.5M NaCl on IRC-50 at Higher pH (Eluate Cone. vsEluate Volume) ••••••• 69

Fig. 21 Elution with 0~5M NaC1 on IRO-50 at Higher pH (% of Total El ted vs E1 ate Volume) ••••• ~

Fig. 22 Elution Wl~h 1M Na01 on IRO-50 (Eluate Cone. vs Eluate Volume) . · · · · · • 72 Fig. 23 Elution with 1M NaC1 on IRC-50 (% of Total Eluted VB Eluate Volume). ···· V Fig. 24 Elution with 0.5M NH4Cl on IR-120 (Eluate cone. VB Eluate Volume) • • • · · • • 75 Fig. 25 Elution with 0.5M NH4C1 on IR-120 (% of Total Eluted vs Eluate Volume). •• · · 76 Fig. 26 Elution with 1M NH4Cl on IR-l20 (Eluate Cone. VB Eluate Volume) • · • , •• · 78 Fig. 27 Elution with 1M NH4Cl on IR-120 (% of Total Eluted VB Eluate Volume). • ·· • 79 vi

LIST OF TABLES Page,:

Tab~e I. The Quantitative.Relationship between Cobalt and Potassium Ferrocyanide". • •• 37 Table II. Exhaustion Cycle on IRC-50 at Lower. pH.. 47 Table III. Exhaustion Cycle on IRC-50 at Higher pH • 49 Table IV. Exhaustion Cycle on IR-120 ••••••• 51

Table V. Elution with H2S04 on IRC-50 ••••••• 53 Table VI. Elution with 0.5],1 J.lUI4C1 on IRC-50 •••• 56 Table VII. Elution with 1M NH4Clon IRC-50 • 59 Table VIII. Elution with 0.5 NaC! on IRC at Lower Flow Rate •••••••••••••• •• 62 Table IX. Elution with 0.5M NaCl on IRC-50 at Higher Flo\-! Rate ••••••••••••• 65 Tabla X. EIution with O.5M NaC1- on IRC-50 at Higl1er pH •••••••••••••••• 68 Table XI. Elution with 1M NaCl on IRC-50 ••• . .. 71 Table XII. Elution \iith O.5M NHL~Cl on IR-120 •••• 74

Table XIII. Elution with 1M ~rn4Cl on IR-120 • •• 77 1

INTRODUCTION

Ion exchange has a history of about one hundred years. However, its application as a metallurgical process was not realized until 1909 when Gans (1) suggested the possibility

(1) Sussman, S., Ion Exchange, Theory And Practice. Acade mic Press, 1949, p.254 of adsorbing gold from dilute solutions on manganese zeoli tes. Recently, as a consequence of the development of many new synthetic resins of relatively high capacity and stabi lity, considerable attention has been directed to using ion eXchange as a means of separation, recovery, and concentration of metals from industrial waste waters the treatment of which has otherwise not been practical or profitable. Metals such as copper, magnesium, gold, Silver, nickel~ the rare earths, etc. have been recovered by ion exchange either commercially or in laboratories. Ion exchange provides a very efficient and economical method for recovering metals from large volumes of extremely dilute solutions. It has been shown, for example, that in the recovery of copper by a cation exchanger the use of one pound of sulfuric acid performs the same duty as the evapora tion of about 4,200 pounds of water. (2) Another advantage

(2) Beaton, R. H., and Furnas, C. C., Concentration of Dilu te Solutions of Electrolytes by Base-exchan~e Materials. Ind. Eng. Chem., Vol.33, pp.1500-l5l3 (1941) 2 of ion exchange 11es in 1ts being a means of purification by preferentially removing a minor qu.ntity of impurities from a solution. Obviously such a process may be desirable in some metallurgical industries where purification has been only inefficiently achieved by other methods. On the other hand, ion exchange has limitations in 1ts application to metal recovery. The lack of selectivity under, ppactical operating conditions and the limited capacity of the exchangers explain why ion exchange, attractive as it appears, has not been extensively used in the metallurgical field. The purpose of the present research is primarily to in vestigate the possibility of the separation and recovery of cobalt from zinc sulfate solutions by means of cation ex change resins. It is thought that such an investigation, seemingly of academic interest, might throw some light on determining the possibility of processing some industrial solutions containing both cobalt and zinc. Cobalt, for ins tance, has been well known as one of the most detrimental impurities in zinc electrolytic solutions. In spite of va rious purification treatments, small amounts of cobalt are often present in the electrolytic solution, causing a reduc tion in current efficiency and the production of 1rregular cathodic zinc. The question was whether an efficient separa tion of the cobalt could be obtained by the use of cation ex change resins. As shown by the experiments presented later, the carboxylic acid res1n does bring about considerable sepa- 3

ration of cobalt from zinc when a proper elutriant is em ployed. However,. since zinc, rather than cobalt, is prefe rentially retained by the resin, this process does not offer a practical method for purifYing the zinc electrolytic solu tion. The experimental work in this research consists chiefly of two parts. The adsorbing characteristics of the synthetic cation exchange resins, Amberlita IRC-50 and Amberlita IR 120, with respect to cobalt and zinc, were first studied. Most of the work in this part was done using Amberlite IRe 50 because it showed much better selectivity than did Amber lite IR-120. Greeter emphasis was laid on·the second part, i.e., elu tion by cationic displacement. This was done because elution is of primary importance as far as separation is concerned. Experiments have been conducted to study the variOUB fac tors which affect the efficiency of separation and recovery of cobalt from the system. Here, too, Amberlite IRC-50 had

been used for most of the work. The author regrets that, be cause of the organic compounds interferring with the colori metric analysis of cobalt, elution by complex-ion formation was not resorted to. 4

LIT,ERATURE RIWIE\'T

The separation of the system Co-Zn by ion exchange has not been reported before. Therefore a general review of ma tal recovery by ion exchange will be given briefly below. ~: Gold was the first metal recovered by ion exchange. As early as 1909, Gans (1) suggested the use of manganese

(1) Sussman, S., op. cit., p.254

zeolites for precipitating gold from dilute solutions. About thirty years later Baur, (2) in an attempt to recover gold

(2) Ibid. p.255

from sea water, ,made an investigation of adsorbing gold on numerous adsorbents from solutions containing 5 milligrams of gold per CUbic meter of solution. His results showed a capacity of 0.001 milligram of gold per gram of synthetic so dium alumino-silicate cation exchanger. In view of the fact that the gold was present as a complex anion in the sodium chloroaurate solution, it was very likely that the gold was removed by ordinary adsorption rather than by ion exchange. Since gold, as well as the other precious metals, usu ally occurs as a complex anion in aqueous solutions, it can

be removed by anion exchange. This was done by Sussman et al. (3) who removed gold trom chloroauric acid solutions by 5

(3) Sussman, ,S., Nachod, F. C., and Wood, W., Metal Recovery by Anion Exchange. Ind. Eng. Chem., Vol.37, pp.6l8-624, (1945)

two synthetic anion exchangers in chloride form. Recovery of platinum from a solution of chloroplatinate by anion exchan

ge was also performed by the authors. ~ecause of the insolu- bility of ammonium chloroplatinate, elution with ammonium hydroxide was not successful. However, complete recovery of the metal was made possible oy ashing the anion exchanger and dissoving it in aqua regia. Copper: Copper has been commercially recovered from was te waters by ion exchange. The first application of the pro cess for copper was in 1934 when Austerweil and Jeanprost (4)

(4) Austerweil, G., and Jeanprost, C., Process for The Sepa ration, Preparation, And Purification of Salts! Salt so lutions And Other Solutions. U. S. Patent 1,97~,447 (Oct. 30, 1934) patented a process of removing cupric ions from a copper sul

fate 801utlo~ by a salt-regenerated greensand and of reco vering the metal by elution withe~dium acetate. Three years later, an acid-regenera~ed organic cation exchanger was used by Phillips and Pain. (5) The process developed by Tiger and

(5) Ets Phillips,and Pain, French Patent 808,997 (Feb. 19, 1937) and Goetz (6) was to pass the copper~containina solution

(6) Tiger, H. L., and Goetz, P. C., Canadian Patent 396,040 (April 22, 1941) through a bed of acid-regenerated cation exchanger Which, When exhausted, was eluted qy an excess of strong hydrochlo ric acid. The eluate was later distilled to concentrate the copper and to recover the excess HC1. An outstanding example of commercial practice of metal recovery by ion exchange is the treatment of the cuprammo nium waste liquors. The idea of recovering copper from such liquors by means of an inorganic zeolite regenerated with an ammonium salt was early attempted by Syrkin and Krynkina. (7)

(7) Syrkin, Z. N., and Krynkina, A. Y., Russian Patent 50,548 (Feb. 28, 1937)

The process, however, had not been started commercially until 1941 in Germany. The operations involved sand filtration of the IIblue waters" followed by passage through the cation exchange unit and the ammonia recovery system. Waste spin ning acid was used as the regenerant. After sand filtration, the regenerant effluent w'as heated a.nd trea.ted with 20 per cent sodium carbonate solution to precipitate basic copper sulfate which was concentrated on a thickener, filtered on a vacuum filter, and re-used in the rayon process to make up fresh cuprammonium spinning solution. 7

An extensive study o~ concentration and recovery of copper from dilute. solutions by means ot an acid-regenerated: cation exchanger was carried out by Beaton and Furnas (8) in

(8) Beaton, R. ~.; and Furnas, C. C., op. cit. pp.1500-l5l3

1941. They found a good correlation of their experimental equilibrium data with modified mass action law predictions for the sulfonated coal cation exchanger. Their results can be expressed .. by the equation:

(CUZ2 ) (R+) - 2[K (CuZ2)'" N}K (1) (Cu++) where Z denotes the zeolite, N represents the n~ber ot replaceable ions per unit weight of exchanger

N .--(HZ) of" 2( CuZ2) (2) and K, the equilibrium constant for the equation,. is

K =(CuZ2) (R+) (HZ) 2(Cu··)

In all the examples shown above" copper was removed by means of cation exchangers. Sussman (9) points out that anion

(9) SUBsman,:S., Ope cit., p.245 exchangers are also capable ot removing copper trom aqueous solutions in which it is present either as the simple cupric lon or as the cuprammonlum complex. Thus by passing a cOpper Bulfate solution through a bed of a salt of an anion exchan- 8

ge resin the copper can be removed presumably by v1~tue o~

the ~ollowing react~on:

I~ the copper is present as the cuprammonium complex, its removal can be effected by passing the solution through a bed~.o~ alkali-regenerated anion exchanger. The reaction may be represented by

(5) Silver: The recovery of silver by cation exchange wa.s first patented by Phillips and Pain. (10) However, the ten-

(lO)Ets Phillips, and Pain, French Patent 808,997 (Feb. 19, 1937) dency has since been to turn toward the use ot anion exchan gers. In the German practice ot recovering silver ~rom pho tographic-film manUfacturing and processing waste waters,(ll)

(11) ~~ers, F. J., Report PB40802, pp.2l-22, Office of Tech nical Services,.Washington (1946) the solution obtained by washing the film coating equipment was first treated to separate 'the AgBr, Agel, or AgI. The residual 2-10 p.p.m. of silver present was converted to a complex anion by adding sodium thiosulfate or sodium cyanide. The aolilt10n was then passed through a salt ot an anion ex changer which was la.ter eluted with sodium hydroxide. The exchanger was regenerated with an acid to convert it into a 9

salt form for re-US6. This three-step cyclic process can be represented by the following equations: RS04+ 2Ag(GN)2- R(Ag(GN)2)2 t 504 (6) R(Ag(GN)2)2-+ 20H-~R(OH)2-+2Ag(GN)2 (7)

R(OH)2+ H S0 ----+ RS0 + 2H 0 (8) 2 4 4 2 Ohromium: In attempting to recover the chromates con tained in electroplating waste "solutions, Grindley (12)

(12) Grindley, J., Treatment And Disposal of Waste Waters Containing Chromates. J. Soc. Chern. Ind., Vol. 64, pp.339-344 (1945)

treated the solutions with two anion exchangers, Zeo-Karb and Deminerolit B. He found that only a small portion ot the chromium adsorbed was recovered by regeneration in the nor mal manner. Better results were obtained by washing the sa

turated exchanger with N HGl and a 5 percent solution of am monium hydrOXide followed by distilled water. In a more extensive investigation, Sussman, Nachod, and Wood (13) found that a three-step cyclic process, like that

(13) Sussman, S., Nachod, F. G., and Wood, W., Ope cit. , p.6l9

used in the silvery recovery, was necessary to obtain an efficient recovery of the chromate. In addition, it was found that the chloride form resin was more effective than the sul fate form and the chromate was eluted more efficiently by an 10

alkali than by a neutral salt such as ammonium chloride. Ma5Aesium: Recovery of magnesium f'rom sea water has been performed by processes involving both cationic and anionic exchange. One of the British patents (14) proposed

(14) Ocean Salts Ltd, and Adams, B. A., British Patent 536,266 (~~y 8, 1941) to precipitate the magnesium as a hydroxide Which was dissol ved by acid and subsequently removed from the solution by a sodium cation exchanger. The exchanger was later eluted with a salt solution, yielding a solution of magneBi~ chloride contaminated with a little magnesium sulfate and sodium chlo ride. Another patent by the Company (15) was to precipitate

(15) Ocean Salts Ltd., and Adams, B. A., British Patent 541,450 (Nov. 27, 1941) the magnesium from sea water by lime. The precipitate was reacted in a slurry mixture with the hydrochloric acid-salt of an anion exchanger to form a solution of ma5neslum chlo ride. After being separated from the exchanger, the solution was passed through an acid-regenerated cation exchanger to remove the magnesium. The HOI formed was used to regenerate the anion exchanger. In America, more economical processes have been patented (16) to remove magnesium from sea water. Hunter, for example,

(16) Hunter, M. J., Process And Agents for The Recovery of 11

~~gnesium Ions from Brines. U. S. Patent 2,409,861 (Oot. 22, 1946) claimed the use of special carboxylated oation exchange re sin for the recovery of magnesium from brines and sea water. The magnesium was displaced from the exchanger by a solution of an acid having an ionization constant of at least 1.8 x 10-5 and an acid conoentration at least normal. Nickel And Cobalt: The separation of these two chemical ly similar metals was first attempted by Austerweil and Jean prost (17) who saturated a greensand cation exchanger with a

(17) Austerweil,G., and Jeanprost, C., Process for The Pre paration, Separation, And Purification of BaIts, Salt Solutions And Other Solutions. U. S. Patent 1,978,447 (Oot. 30, 1934)

10 percent solution of chemically pure cobalt nitrate. A so lution of technically pure co~alt nitrate containing nickel nitrate in the ratio of 1:12 of the cobalt nitrate was then passed through the exchanger. Nickel was retauled by the ex changer, leaving a purified cobalt nitrate solution.

Samuelson (18) noted that, when his lI exchanger Bit was

(18) Walton, H. F., Ion Exohange, Theory And Practice. Aca demic Press, .1949, p.17 saturated with CO(1~3)5Cl·· , the swelling was greatly dimini shed, :showing that the oomplex ion was taken up by the ex changer. He Also found quantitative separations for the 80- 12 dium and potassium salts of the complex cyanides of iron, chromium, cobalt, molybdenum, and tungsten.

~: !-!ost of the work of ion exchange involving zinc was chiefly concerned with the separation of the metal from binary systems. Griessback (19) reported that by starting

(19) Sussman,:S., Ope cit., p.258 with a solution containing 15 grams per liter zinc and 5 grams per liter copper, passage through a bed of cation ex change resin gave an effluent containing up to 17.6 grams per liter zinc and no copper for a substantial portion ot the run. Investigation on the same system was recently carried out by Breton. (20) B,y passing a solution containing su1ta-

(20) Breton, E. J. t Private communication, May, 1950 tea of copper and zinc (both of about 0.005N) through a co lumn of Amberlite IRC-50 in sodium form t , 90.3 percent of the zinc was recovered in 45 liters of the effluent, with a pu rity of 98.3 percent. Zinc and cadmium were partially separated on a sulfona ted coal cation exchanger by Kozak and Walton. (21)

(21) KozaK, R., and Walton, H. F., Separation of Metal Iona by Cation fiXchangera. J. Pby. Chem., V01.49, pp.471-472 (1945) 13

Rare Earths: The first clue for separating the rare earths by ion exohange wasiven by Russell (22) who demons-

(22) Russell, E., Swartout, J. A~f Hume, D. N., a.nd Kettelle, B. H., Project Work, May 1944 trated the selective elution of Zr and Cb from IR-l-adsorbed fission mixture with dilute oxalic acid. The technique was later employed by Tompkins et al. (23) in the Plutonium Pro-

(23) Tompkins, E. R., Khym, J. X., and Cohn, W. E., Ion Ex change as A Separation Method. I. The Separation of Fission-produced P~dioisotopes, Including Individual Rare Earths, by Complexing Elution from Amberlite Re sin. J. Am. Chem. Soc., Vol.69, pp.2769-2777 (1947) ject, using a 5 percent citric acid solution as the elutri ant. By controlling the pH of the citric acid solution with concentrated ammonium hydrOXide, they were able to separate the components from each other in the binary systems, Zr-Cb, Y-Ce, and Ba-Sr (aD{aline earths). As will be mentioned la ter, the separation is based upon the different tendency of the cations being complexed by the citrate ions. Spedding et ale (24) showed that a mixture solution

(24) Spedding, F. H., Voigt, A. F., Gladrow, E. M., and Sleight, N. R., The Separation of Rare Earths by Ion Exchange. I. Cerium And yttrium. J. Am. Chem. Soc., Vol.69, pp.2777-2781 (1947) containing Y 0 and C9 0 in the ratio of 1:25 could be elu- 2 3 2 3 14 ted to yield 60 percent of the yttrium in spectroscopically pure form before the cerium break-through. For the system Nd-Pr, (25) one passage of a 50 percent mixture solution

(25) Spedding, F. H., Voigt, A. F., Gladrow , E. M., Sle1ght, N. R., Powell, J. E., Wright, J. M., Butler, T. A., and F1gard, P., The Separation of Rare Earths by Ion Ex change. II. Nd-Pr. J. Am. Chern. Soc., Vol.69, pp~2786 2792 (1947) through the Amberllte IR-l column yielded about 22.5 percent of the Nd free of Pr, and 50 percent more than 98 percent pure. 15

THEORETICAL CONSIDERATIONS

General ConSiderations: Ion exchange is a process by which the cations are re distributed between a solid phase (the exchanger) and a li quid phase (the solution). The exChanger, composed of posi tively charged cations and negatively char ad anion roups, is virtually an ionic solid. The surface ions are bound less firmly than the internal ions of the same species. When pla ced in a polar solvent, these surface ions may become solva ted, resulting in a further lowering of the binding energy and a marked dissociation from the lattice. If a foreign electrolyte is added to the system, it is 10 ical to expect an exchange to take place between these surface ions and ions of the same charge of the foreign electrolyte. Amberlite IRe-50 derives its exchange activity from a carboxylic acid (-COOH) group and Arnberlite IR-120 from a sulfonic acid (-S03H) group. Their structures are shown in Fig. 1 and Fig. 2 respectively. The hydrogen ion in both re sins are more or less free to diffuse through the resin pha-

Fig. 1. Amberlite IRe-50 se as in a solution, and hence are replaceable. The sulfonic 16

,...,.vl1')D--

("r"' t.: ·~C·. 1.J\ -. l_ ." "J P. ..-' ~

Fig. 2. Amberlite IR-120 acid group, beip~ a stronger acid than the carboxylic acid group, shows a more complete dissociation than the latter. Consequently the hydrogen ions in Amberlite IR-120 are much more eas11y exchanged for another cation than those in Am- berlite IRC-50. The -COO and -803 anions themselves are free' to move with the network of the resin phase and to ro- tate and vibrate about the network, but their translational motion is limited by the immobility of the network aa a who le. These resins possess a gel type structure. Almost all of the ion exchange takes place in the interior of the gra nules of which the resins are composed. ~nen placed in an electrolyte- , therefore, there is a clear distinction be tween the ions associated with the resin and those associated with the solution. When two cations are adsorbed by a cation exchanger,.X-ray diffraction measurements have shown that a continuous range of solid solutions are formed. The fact that cation exchangers are neutral bodies re veals that ion excha,nge proceeds by eqUivalents. For every gram eqUivalent weight of cations adsorbed by the exchanger, 17

one gram equivalent weight of cations must be released~ This is necessary to maintain the electrical neutrality, since few anions are taken up by a cation exchanger. Fundamental Factors Governing The Separation of Cations: In general the separation of cations is baaed upon the difference in affinity of the exchanger for the cations. The affinity is a function of the following factors: 1. Valency of The Cations: Jenny (1) first discovered

(1) Walton, H. F., op cit., p.13

that the affinity of a cation for the cation exchanger in creases with increasing valence. On a group basis, the ad sorption affinity for cations will be: monovalent'" divalent < trivalent cations. We have seen that the cation exchanger is virtually a highly ionized solid in which the cations hold their positive charge, balanced by the negative charge of the anion groups of the exchanger. The two oppositely charged groups must be bound together chiefly by an elec tostatic force, presumably of Coulombal nature. Evidently, then,.a divalent cation which is doubly charged should be adsorbed more strongly and held more firmly than a singly charged cation, since, according to the Coulomb's law, the attractive force is proportional to the product of the charges. However, the above relationship between valency and ad sorption holds true only for low concentrations and ordinary 18 temperatures. According to Schachtschabel (2) and Wilkander,

(2) Kunin,: R., . Ion Exchange, Anal. Chem." Vol.21, p.87 (1949)

(3) at high concentrations the difference in the exchange

(3) Ibid. p.87 npotentialsll of ions of different valence diminish and in some cases the ion of lower valency may have higher exchange potential. Such a contradiction to the Coulomb's law, as sug gested by Jenny" (4) may be attributed to hydration of the

(4) Jenny, H., Ionic Exchange in Colloidal Aluminum Silica tes. J. Phy. Chem., Vol.36, p.2249 (1932) ions. 2. Size of The Cations: According to the Coulomb's law, the electrostatic attractive force between two opposite char ges are inversely proportional to the distance between the two charges. In the case of ion exchange, the attraction force obviously varies inversely with the effectice size of the cation. Furthermore, the synthetic exchange resins are composed ot a rigid cross-linked- net1Vork upon which the ion ic groups are affixed. It is to be expected that the small ions will be accommodated more easily than the large ions. Consequently, for cations of the same valence the ad sorption affinity will increase with decreasing effective 19 ionic size~ B,y effective ionic size is meant the hydrated ionic radius since the cations are 'more or less hydrated when placed in an aqueous solution, and it is this hydrated ionic radius that accounts for the difference in adsorption affinity. This reasoning was actually comfirmed by Boyd et al. (5) who found that for the reaction:

(5) Boyd, G. E., Schubert, J., and Adamson, A. W., The Exchange Adsorption of Ions from Aqueous Solutions by Organic Zeolites. J. Am. Chem. Soc., Vol.69"Pp.28l8 2829 (1947)

A~ + HX---~ H~ t" AX (1) the change ot free energies were related in a linear manner with the reciprocal a O, the closest distance of approachg1 ven in the Debye-Huckel activity equation: log (! =-AIU / (1 + Bao!U) (2) Their results are shown in Fig. 3.

,,-..., 000 CD r-I o 600 -{ Fig. 3. Ionic radius and G1 400 o free energy of ion exchan ~ ge. (after Boyd et al.) o 200 1 o

0.1 l/aO (A.U.) 20

The next point of importanoe sU8sested by the above au thors is the correlation o~ adsorption a~finity with the pa rameter a O• I~ a O is taken as a relative measure of the hy drated ionic radius, the relat1ve adsorption affinity of oat ions of the same valence is obtained directly from theacti v1ty coeff1c1ent-concentration plots. Thus for a series of salts w1th common anions, the higher their curves lie in the plot, the larger 1s a O and consequently the weaker is the adsorption of the cation component by the cation exchanger. In such a manner the above authors predicted the sequence of adsorption among the alkali cations to be:

f Ost> Rb > K+> NH4-+ > Na+; for the alkal1ne earth cations:

-:sa'"'> Sr +; ~ C·"a)o Mg..t.; while for the divalent transition metals: _ """ ~'. ok ...... ~ ~ > Cu > N1 ~ Co > Fe • As the authors pointed out, their predictions, based alone on the parameter, are oversimplified. Under different cond1tions other factors may come into play modifYing the order of adsorpt10n. For example, it was shown by Gr1essbaoh (6) that copper was more strongly adsorbed than z1nc by cat~

(6) Sussman, S., OPt cit., p.258. ion exchangers. The same result was recently comfirmed by Breton. (7)

(7) Breton"E. J., Private communication. May, 1950. 21

The relationship between the exchange capacity and hy drated ionic radius and valence of the exchanging ion was studied by Nachod and Wood. (8) As shown in Fig. 4, the lar-

(8) Nachod, F. C., and Wood, W., The Reaction Velocity of Ion Exchange. J. Am. Chern. Soc., Vol.67, pp.629-631 (1945) ger the hydrated ionic radius, the lower the exchange capa- . 0-(.. ~9 H '0 .s 8 Ie;

0.7 0.8 0.9 1.0 Exchanee, meq. / g.

city. All values of the ions of the first group fallon a

Btrai~lt line and the same holds true for the second group. The slope is the same for both groups indicating the ionic size really determines the relative accessibility of an ex change group. The shift of the tVlO lines indicates the Itva_

ll lence effect • 3. Atomic Weia.:ht of The E>c.changi115 Cation: The results of Jenny (9) in stUdying the aluminosilicate exchangers 22

(9) Jenny, H., Ionic Exchange in Colloidal Aluminum Silica tes. J. Phy. Chem." Vol.36, ,p.2249 (1932)

showed that for alkali metal and alkaline earth metal cations the strength of binding by the exchanger increases as the atomic weight of the cation increases. In the alkali me tals the order of adsorption is:

Cs > Rb > K > Ne. > Li; and in the alkaline earth metals the order is:

Be. > Sr > Oa :> Mg. As seen from Fig. 3, the same orders were found by Boyd et al., with a sulfonic acid resin. The existance of such a re lationship may be explained in terms of the difference in hydration of the cations. It is known that the atomic wei ghts of the elements in each group of the periodic table in creases from top to bottom while the tendency of hydration decreases in that direction as a consequence of the increase of the sh1elding effect of electronic orbits. In addition to the effects of valency, ionic size, and atomic weight, ion exchange 1s also influenced by the chemi cal characteristics of the resin as well as of the cations involved. As pointed out by Bauman, (10) the type of ionic

(10) Bauman; W. C., Ion Exchange, Theory And Practice. Aca demic Press, 1949, p.45.

group w1l1 determine the ion exchange properties of the re- 23

sultant exchanger. For instance, hydrogen ions will be more strongly adsorbed by a weak carboxylic acid exchanger than by a relatively stronger sulfonic acid exchanger. Similarly, the chemical affinities of one exchanger for cations A and B may be Dr the reverse order from those of another exchanger so that a resultant reverse order of adsorption may be ob served.

The actual separation of cations is brought about by elution, i.e., removal of the adsorbed cations from the ex changer by washing the latter with a solution containing a nother cation, either simple or complex. Elution may be ef fected in three ways: 1. Elution by Cationic Displacement: If an exchanger has a higher affinity for cation A than for Cation B~~ then when the eXchanger,. already saturated with A and B, 1s brou ght into contact with another cation C, B will be displaced by C more easily than will be A. In such a way A and B can be separated. In choosing a proper elutriant, it is appa rent that the factors of valency, ionic size, and the chemi cal characteristics of the exchanger and cations involved should be taken 1nto consideration. For example, the cat ions of lower valency can be first removed with a dilute acid and those of higher valency by a more concentrated acid so lution. The usual elutriants used for this pu~ose are sul furic acid, hydrochloric acid, sodium hydroxide, ammonium hyaroxide, ammonium chloride" sodium chloride, etc. 2. Elution by Complex-ion Formation: For cations which 24 have a low "separation factor" this method is used which con- - sists of passing a solution containing an organic compound, such as citric acid or oxalic acid, through the exchanger originally loaded with the cations to be separated~ The se paration is based upon the different tendency of the organic anion to complex the cations involved or upon the different stability of the complex ions thus formed. Here the concen tration and acidity of the elutriants are most important. An outstanding example is the separation of the alkaline earths from the rare earths by eluting with a 5 percent citrate so lution. Both the alkaline earths and ral'e earths form complex ions with citric acid. Yet the pH at which citric acid appreciably complexes cations of different valenoes varies by at least one pH unit. Consequently they are separated from each other by adjusting the pH of the citric acid solution with concentrated ammonia. In elution by complex-ion formation two factors are pre vailing,. i.e., the affinity of the exchanger for the cations and the stability ~f the complex ions formed by the cations and the eluting organic anion. If they work in favor of each other the separation will be very simple. Thus in the above case, the alkaline earths can be easily separated from the rare earths because those cations which are mostly strongly held by the resin also form the weakest complexes. (11)

(11) Sohubert, J., Ion Exchange, Theory And Practice. Aca demic Press, 1949. p.182. 25

3. Elution by Combination of Cationic ~1splacement And Complex-ion Format1on: In case one of the adsorbed cations forms a complex ion with a certain compound but the others do not, separation can be effected by removing that cation by complex-ion formation and the others by cationic displace- mente

Ion Exchange Equilibria: The nature of ion exchange equilibria has been studied from two different points of view, 1.e., the adsorption 1so- therm and the mass action law. The relationship between the amount of base exchange and the equilibr1um concentration of added electrolyte ap pears to be similar to an ordinary adsorption isotherm. For a given amount of ion exchanger and a fixed total volume of solution a curve like those shown 1n F1g. 5 1s obtained.(12)

(12) Boyd,G. E., Schubert, J., and Adamson, A. W., op c1t., p.2819

For a s1mple case of the simultaneous competit1ve adsorption of two singly charged cations, A+ and B,~ Boyd et ale der1ver, by analogy With the LangmUir equation for ad sorption from a binary gaseous mixture, the follOWing equation for the adsorption of one the ions, A•• from a dilute,; electrolyte solution: 26

0.02 0.04 0.06 0.08 Equilibrium cone. moles;1itel'

Fig. 5. Adsorption of cations at 250 by organic ion exchan gers. (Boyd et at.) where (X/m)A is the amount of A+ ion adsorbed per unit wei ght of adsorbent; CA and CB are the respective eqUilibrium concentrations (activities) of ions A·and B+ in solution; and K, bl and b2 are constants. Since the number of unsaturated anionic exchanging groups is always small, the quantity unity in the denomena tor of E~uat1on (3) can be neglected relative to the quan

C b C ), tity of (bl A + 2 B X) _ kblCA (iii A- h.. C + b C -.1. A 2 B 27

bl (CA k~ -) CB - (4) ltbl CA (C ) b2 E Equation (4) can be rearranged into:

CAICB _ b2 C 1 A (5) (X/m'A - blk + CB it so that a (CA/cB)/(x/m)A vs C.t/CB plot should give a. strai- ght line.

By ins ec·tion of Equation (4) it is clear that the amount of base exchange, (X/m)A' 1s dependent upon the ratio of the equilibrium concentrations of the exchanging iona, and hence that the extent of adsorption will be independent of t dilution. For CB» CA' the adsorption of A becomes directly proportional to CA; for cA»cB,.the adsorption of At becomes indenpendent of concentration. By similar reasoning, the authors show that the ex change adsorption of two cations of unequal charge, A+m and Btn, will be given by: kb1 (CA+"" ) m (~)Am = bl(CA.. ...,)n~b2(CB ..P'\)il (6)

For the adsorption of a singly charged cation Mi from a mixture of singly charged cations in solution the expression is:

Evidently Equation (7) shows that the greater the ionic strength of the mixed electrolyte the less the exchange ad- 28

sorption of any single cationic species held at constant concentration. There are other emperical equations for ion exchange equilibria. That of Jenny (13) is given as:

(13) Walton, H. F., Ion Exchange, Theory And Practice. Aca demic Press,: 1949. p.7.

y =k ( c )p (8) a - c where y is the exchange per gram of solid; a, the initial equivalent concentration of the added salt; and c, the equi librium concentration of the added ions which remains after the exchange has taken place. One particular point of Equa tion (8) is that it does not involve the concentration of the ion originally in the exchanger.

The most general equation, derived by analogy with the Freundlich adsorption isotherm, is that of Rothmund and Kornfeld: (14)

(14) Ibid. p.5

where mAX and mSX are moles of cations A and B per gram of exchanger, A+ and BT are molar concentrations in solution, and k and p are constants with p less than unity. Experimental results have been obtained, in agreement 29 with the above equation. However, it can not be concluded that ion exchange is an adsorption phenomenon in any special sense since equations of the same forms can be derived from the mass action law. The second branch of studying ion exchange equilibria is characterized by assuming that the ion exchange is ana logous to an ordinary metathesis reaction and that, at equi librium,.the mass action law applies to the heterogeneous system: ion-exchanger + aqueous solution. For the exchange reaction (10) where A and B are two monovalent cations and R represents the insoluble structually bound anionic part of the exchan ger" the thermodynamic equilibrium constant, Ka "is defined by

(11) where agand aB+are the respective activities in aqueous so B~, lution of the ions A+ and and aAR and aBR. are the activities in the solid state.

a~ and a B+ are each related to the product of the con centration of the respective ions in solution by their ionic activity coefficients. The appropriate activity coefficient in to be used is that for the ionAthe mixed electrolyte in equilibrium. \~ere the ionic strength of the equilibrium solu tion is low, the mean activity coefficient of a salt in a mixture can be assumed to be the same as that of the pure 30 salt at the same ionic strength. Further, at high dilutions, the mean activity coefficient ratio for two electrolytes of the same valency may be taken as unity. In evaluating the activity coefficients in the solid state, advantage is taken of the fact that the components of the solid phases form completely miscible solid solutions with one another. In general, the activities of the compo nents of the binary solid solutions should be taken as com plex functions of their respective mole fractions. For sim- plicity, let us assume that the ions in the solid are mono valent and have apprOXimately the same size so that ideal solid solutions are formed. In this case, then, aAR and aBR may be set equal to the mole fractions XAR and XBR with E quation (11) changed into

Ka = (aBot )XAR / (aA" )XBR

- (aB" )(nAR) / (aA+ )(nBR) (12) where nAR and nER are the number of moles of A and B in the solid respectively. For the exchange of two cations of unequal charge, AP and Eq, where q;> p, the exchange equation can be 'l'rritten q APT (p/q)BR =ARt (p/q)3 (13) The expression for Ka , assuming ideal solid solutions, is then

K (&Bq)P/q(JC.AR) / (a ) (X ) a = AP BR E (p/q - 1) _ (&Bq)q(nAR+nBR) (nAB) (aAP)(~)Pq 31

EXPERIMENTAL APPARATUS,. EQUIPMENTS, AND MATERIALS

Appamtus .:A1'ld Eguipments: The experimental set-up, as shown by Photograph No.1, is very simple. There are two identical units one of which is illustrated in Fig. 6. The influent, effluent, and back washing water tanks are simplY'glass jars with a total vo-

Photograph No. 1 Experimental set-up.

lume of about 10 liters. In the elution cycle, the intluent tank is used to hold the elutriant and the effluent tank to hold the eluate. The column is made of smooth, colorless py rex tUbing having an inside diameter of 2.44 centimeters and a length of 100 centimeters. Both ends or the column are fit ted with a rubber stopcock and are connected to the tanks by 32

Influent clolashing

tank ve.tertank

D c

to ~ ~

III J I< G> J.. ~. 'ip 0 .,-i III ;J) J..

sand

Effluent B

Dac shin ter outlet

Fig. 6. Experinmte.l t-up 33 glass and rubber tUbings as shown. In Fig. 6, A, B, and C are metallic tUbe clamps. A and B are used to contro~.the flow rate respectively during the exhaustion (or elution) cyole and backwashing. C.:is opened. only during backwashing to let the backwashing water drain through the, outlet tube. Vent D, placed parallel to the upper end of the column, has the function of maintaining a liquid level well above the resin bed in case the influent tank is exhausted. The bottom of the pyrex glass oolumn is covered with a thin layer of glass wool above which is placed a layer of sand about two inches thick. This sand bed supports the re sin which is fed in upon it. The resin bed is usually of such a depth as to leave on top a free space about 50 to 100 percent of the volume of the bed. This free space is ne cessary to provide enough' room for bac~lashing. The other equipment used are as follows: 1. A stirrer actuated by compressed air used in regene rating the resins in a beaker. 2. A portable pH indicator assembly manufactured by the Leeds & Northrup Company. 3. A pair of 50-ml Nessler tubes and an electric lamp used to determine cobalt the colorimetric method.

Materials: The cation exchange resins selected for 'the research are Amber11te IRC-50 and Amberlite IR-120, both manufactu red by the Rohm & Haas Company. They are comparatively new 34 synthetic resins and are claimed to possess 'high chemical and physical stabili~y and high exchange capaoity. Amberlite IRC-50, in the form or white opaque beads, derives its exchange activity from the weak carboxylio acid group. It, therefore" has a strong affinity for the hydrogen ion and is supplied in the hydrogen form. BY leaching with a 4 percent NaOH solution, it can readily be converted into the sodium form. Amberlite IR-120 is produced in the form of dark brown bead-like particles. Deriving its aotivity from a nuclear sulfonic acid group, it has far less affinity for the hydro gen ion. It is supplied in the sodium form and can be trans rerred to the hydrogen form by regenerating with either sul furic acid or hydrochloric acid. The following are some of the characteristics of both resins as reported by the Compa~ Amberlite IRC-50 Amberlite IR-120 Density: 0.64-0.77 gm.per 0.77 gm. per mI. mI. (wet, H-form) as shipped Screen grading: 16-50 mesh 16-50 mesh (Wet) Effective size: 0.35-0.50 rom. 0.4-0.6 mm. Voids: 35-40 percent. 45-50 percent. 35

EXPERI1-1ENTAL PROCEDURE

Analyses: 1. Analysis for Cobalt: Cobalt is determined by a colo rimetric method using the Vogel's reaction. (1) Five rol. of

(1) Snell.. F. D., nd Snell, C. T., Colorimetric Methods of. Analysis. Vol.I,.Van ostrand Co., 1936, p.328

60 percent 1\!H4CNS is added to a 20-ml sample in a. 50-ml Neg... sler tube. (In case-the sample is too concentrated in cobalt, a smaller sample is used and diluted to 20 mI. with water.)

Upon addition of an equal volume (29 IDl.) of acetone and af ter shaking briskly, a beautiful blue color is developed due to the-following reaction:

00++ ~NS- + 2NH4+ (NH) () + • 4 2ab CNS 4 In another Nessler tube' 5 ml. of the NH40NB:: and some wa ter are added. The amount of water should be such that the total volume of the water and of the standard cobalt sulfate solution to be added equals 20 IDl. After Balning some expe

riencej this can be determined easily by inspecting the color of the sample and figuring the'amount of standard cobalt sulfate solution required. Twenty five ml. of acetone is then poured into the: tube: and standard co It sulfate solution is added,. a few drops at a- time:.:. The t"ro Nessler tubes are shaken well and brought together for comparison. The. standard cobalt sulfate solu tion is continuously added until the color of the two tubes 35 match. For example, assume that a 5-ml sample is used and ta kes 10 mI. of 0.002N cobalt sulfate. solution to give the same color as the sample. The concentration of cobalt in th~ sample is calculated as follows: Let X =-Normality of cobalt in sample then 10 x 0.0020 - 5X and X =0.0040N

2. Analysis for Zinc: Zinc can be determined by titrat ing '\'1ith potassium ferrocyanide ••<\ccording to Low, (2)however,

(2) Low,.A. H." '",1enig,. A. J., and Achoder, W. P., Technical Methods of Ore Analysis.John Wiley & sons"1927, p.275. the presence of cobalt interfers with the titration. The usual procedure then is to precipitate zinc as ZnS by pass ing H2S tllrough the sample which has been acidified with sulfuric ac1d to a pH of 2 to 3, redissolve the ZnS With Hel· and sUbsequently titrate the solution with potassium ferro cyanide. In the present work this method is not desirable because the time required would be too long to permit hand ling more than 10 samples daily. In order to get around this difficulty, a series of tests were performed to find if there was a quantitative relationship between cobalt and potassium ferrocyan1de. The results of these teats are shown in Table I. It seems that such a relationship does exist,. with or without the presence of zinc. The results show an average.er- 37

ror of about 3 percent which is well justified by the rapi- dity of this method and which is accurate enou for the present work.

Ta Ie I: Determination of the quantitative relationship be" tween cobalt and potassium ferrocyanide. Volume of samples: 250 mI. Blank test: 1.5 nl. of 0.1131 4Fe(Cll)6 for 250 mI. of '\Vater.

No. mI. 0.153N rol. O.lN mI. 0.113N K4Fe (CN)6 %error ZnS°4 C0804 used calculated I' 0 10 16.0 2 0 5 8.4 8.7 3.5 3 10 0 15.0 15.0 0.0 4 I I 4.4 4.3 1.5 5 3 2 8.6 8.6 0.0 6 5 1 9.8 10.0 1.5 7 1 5 9.9 10.1 2.0 8 5 10 24.4 22.9 8.3 9 10 5 23.0 22.7 1.3 10 10 10 31.5 29.9 5.4 11 5 15 32.0 30.1 6.3. 12 20 5 35.0 36.8 4.9 13 30 5 50.0 50.8 1.6 14 1 20 33.0 31.7 4.1

One peculiar point that'~can be observed from Table I is that the equivalent weight of cobalt with respect to po tassium ferrocyanide is not 58.94/2, but 58.94/2 times a factor of about 1.5. Eased on the data in Table I the. analysis of zinc is then carried out as follows: A. Determine the concentration of cobalt of the sample by the colorimetric method. B. Take a suitable sample and dilute it to 250 mI. C. Add about 5 grams of NH4Cl and 5 mI. of cone. HCI. 38

D. Heat" the solution to about 70 0 C. E. Titrate the solution with potassium ferrocyanide. If only zinc is present"s white precipitate will appear due to the reaction:

2K4Fe(CN)6 -t 3Zn++ -lC2Zn3(Fe(C )6)21" 6 t If only cobalt is present, a yellow precipitate will be pro duced. With both zinc and cobalt present, the precipitate is' green. The nature of these reactions is not known as yet~. F. The end point is determined by bring a drop of the solution in contact with a drop of 10 percent uranium nitra tesolut~on and is indicated by the appearance of a brown tinge, owing to the formation of uranyl ferrocyanide ° G. Calculate the concentration ot zinc from the volume of potassium ferrocyanide used atter deducting the amount of potassium ferrocyanide reacted with cobalt. 3. Calculation: A. Quantitative factor between co alt and K4Fe(CN)6~ Take No. 1 of Table I as an example:

Vol. of 0.113N K4Fe(CN)6 required to ree.ct with 10 mI. of O.IN C0804 - 16 - 1.5 =14.5 ml. Concentration of cobalt with respect to K4Fe{CN)6 = 14.5 x 0.113 / 10 =0.164N Theretore the quantitative factor =0.164 / 0.1 =~ It should be noted that the concentration of the C0504 solution was determined by weighing a certain amount of

CoS04 °7H20, dissolving it in water, and diluting the solution 39 to one liter. Since the cobalt sulfate crystals easily chan ge their amount of~water of crystallisation, the concentra tion of cobalt thus date ined is only approximate. Conse- q ently the quantitative factor should be determined for each newly prepared cobalt sulfate solution. According to the author's. experience, the true q antitative factor betwe en cobalt and potassium ferrocyanide s ems to be 1.5. B. Determination of Zinc: Take No. 22 of Table II as an example: Concentration of cobalt: 0.0055N Vol. of 0.113N K4Fe(CN)6 ken by a 250-ml sample: 32.5 - 1.5 =31.0 mI. Let X =concentration of zinc in normal then 250X t 250 x 0.0055 x 1.64 = 31 x 0.113 and X= 0.0050N

Operation: 1. Exhaustion Cycle: In order to prepare the resina·, in a form more easily displaced by cobalt and zinc, the Am berlite IRC-50 is converted into the sodium form by agita ting it in a beaker with concentrated NaOH until the mix ture shows a constant pH of 7. The Amberlite IR-120~: sup plied in the sodium form, 1s regenerated into the hydrogen form with a 10 percent sulfuric acid solution up to a level of 30 meq. of H2S04 per mI. of wet resin. The regenerated resin 1s now fed into the column, the bottom of which is already covered first with a thin layer of glass wool and then with a layer of sand about 2 inches thick. The column is then ready for a IO-minute bac ash. The

purpose of bacl~iashing, is to remove the fine resin particles, impurities, and entrapped air in the bed, and to give the

bed a hydrostatic classification. The ai~" if not removed, will cause channelling during the operation. Baclrnashing 1s done by closing the metallic clamp A in Fig. 6 (p. 32), and opening Band C to let the water from the backwashing water tank flo1tl upward through the column,: pass C and then dra.in through the outlet for ten minutes. The flow rate" control led by means of B, is so adjusted as to give a 50-75 percent,. expansion of the resin bed. After baclruashing, Band C are closed and the upper end of the column is connected with the influent tank where a

sulfate solution of cobalt and zinc, both of about 0.005N in

concentration, is stored. Clamp A 1s then opened to let~the

influent solution pass through the colQ~, upward through the vent D, and finally dO\''lnward into the effluent tank. Clamp A is adjusted to give the desired flow rate, which is expressed in milliliter per square centimeter (of the cross sectional area of the column) per minute (ml./cm2/min.). It is observed that for flow rates over 0.5 ml./cm2/min., some leakage will take place along the circumference. As the operation proceeds, the flow rate will gradually decrease as a result of the decreased hydrostatic head in the influent tank. This is corrected by rea.djusting the clamp A. 41

TJ:ie adsorption of cobalt by the resin bed (Amberlite IRe-50) imparts to the latter a pink color which becomes dar-er and darker as it moves down the column, a fact indi cating the displacement of cobalt by zinc. This is an advan tage in that the breakthrough point for cobalt is clearly observed and can be easily determined. As soon as the breakthrough point approaches, samples are tal en frequently and analyzed for cobalt and zinc. After the maximum effluent concentration is passed, s mples are taken less frequently. When the column is exhausted, as evi denced by the effluent concentrations bein almost identical with (usually slightly lower than) those of the influent so lution, the operation is stopped. In order to regenerate the exhausted resin for re-use,

a 10 percent sulfuric acid solution, containing as many e~

quivalents of H2S04 as those of cobalt plus zinc adsorbed on the column, is passed through the Amberlite IRO-50 bed at a rate of 2 ml./cm2/min. The resin is then taken out by re movin the "bottom rubber stopcoc c and is leached in a bealcer With concentrated NaOH to convert it the sodium form as des cribed above. For Amberlite IR-120, a 10 percent aCl solu tion is employed for regeneration and the resin is later a81 tated with concentrated sulfuric acid to transfer it into the hydrogen form. 2. Elution Cycle: The elution cycle consists of two parts: A. Exhaustion: The procedure is identical with that 42 desoribed under (1), exoept. that the operation is stopped as soon as the breakthrough point is reached. The-resin bed, loaded with oobalt and zinc, is bac ~ashed again. for ten minutes to reclassify it and to remove the residual influent solution. The column is then ready for elution. B. Elution: The eluting solution, elutriant, is stored in the rinsed influent tank and passed through the oolumn at a prescribed rate. Samples are ta ~en from the-'be- ginning. This is continued until all the oobalt and zinc has been eluted out as shown by t e analyses. In case the eluting rate is very slow, the operation is discontinued after_ ten liters of elu.ate have been collected. The resin bed is then re enerated as in the exhaustion cycle. EXPERIAE ~L RESULTS

I. Exhaustion Cycles:

Three runs were made to study the adso~Jtion selectivi ty and capacity of the resins for cobalt and zinc under the arbitrarily chosen operating conditions.

Run No.1: This was a preliminary run in which Amberli te IRC-50 was used at an influent pH of 3.5. The operating conditions and results are shown in Table II and Fig. 7. Sam ples 10 to 17 were analyzed for zinc ,precipitating the zinc as ZnS" redissolvin it with HCl,.and titrating with potassium ferrocyanide. The subsequent samples were analysed by the co-titration method. This second method was used throughout the later experiments.

Run No.2: Amberlite IRC-50 was used at an influent pH of 5.0. The results are given in Tabla III and Fi8. 8. The recovery and purity of cobalt and zinc are as follows: Co fed in: 95.245 x 0.0053 x 58.94/2 = 14.9 gms. Co in effluent: 77.6(units of area) x 5(liters) x 0.001(norma1) x 58.94/2 =11.4,gma. Recovery of Co: 11.4.x·-100/14.. 9 =--76.5% Zn fed in: 95.245 x·o.o048 x 65.38/2 =14.92 gms. Zn in effluent: 24 x,5 x~O.OOl x.65.38!2 =-3.92 gms. Recovery of Zn: 3.92 x 100/14.92-= 26.2~ Purity of Co: 11.4 x 100/(11.4+ 3.92) =74.5% Purity of Zn: 100 - 74.5 =25.5% Run No. ); Amberlite IR-120 was used at an influent pH of 4.2. The results are shown in Table IV and Fig. 9. The_ same calculation gives the recovery and purity as follows: Cobalt Zine Recovery, percent: 32.0 32.8 Purity, percent: 46.0 54.0

II. Elution Cycles: Elution was exclusively performed by cationic displace ment. Fifteen runs \'lere made" twelve on Amberlite~ me-50 and three on Amberlite IR-120. The. preference for the carboxylic. resin is justified by its better selectiVity between cobalt and zinc tha.n obtained '\'1i th the sulfonic resin. Sulfuric acid,. ammonium chloride,; sodium chloride, ammo nium hydroxide and sodium hydroxide were employed as the el:0. triants. Only one run was made using sulfuric acid since it showed too strong an eluting power to effect appreciable se paration. Three runs "'ere made, one using 1aOH and the other two using NH40H. The bases caused precipitation of hydroxi des of cobalt and zinc. Ammonium chloride was employed in four runs, two on each resin. Sodium chloride, , the cheapest of all these reagents, was used for the rest of the expep1 ments, all on Amberlite IRO-50. A. Runs Using Amberlite IRe-50:

Run No.4: Elution with 0.5 percent H2S04. Results are shown in Table V and Figs. 10 and 11. In Fig. 10, the eluate concentration in normality 1s plotted against the eluate wolume in liters, while in Fig. 11 is plotted the per cent of total eluted against the eluate volume. The percen tage of total eluted is obtained by acummulating the number of equivalents of cobalt (or zinc') in all the samples and dividng the acummulated number of equivalents by the total number of equivalents of cobalt (or zinc) ori inally adsorb ed in the column.

Run No.5: Elution with 0.51-1 NH4Cl. The results are shown in Table V1 and Figs. 12 and 13.

Run No.6: Elution with 1M NH4Cl. This experiement was made to study the effect the elutriant concentration. The results are shown in Table VII and Figs. 14 and 15.

Run No.7: Elution with O.SM NaCl at pH 5 and flow rate 0.43 ml/cm2/min. The results are shown in Table VIII and Fig. 16 and 17.

Run No.8: Elution with 0.5M NaCl at pH 5 and flow.

rate 2 ml/cm2/min. The purpose of this run wa.s to study the

effect of flow rate. The results are given in Table I[ and plotted in Figs. 18 and 19.

Run No.9: Elution with 0.5M NaCI at pH 7 and flow rate 2 ml/cm2/mln. This run was made to study the effect of

elutriant pH. The results are shown in Table X and Figs. 20 and 21.

Run No. 10: Elution with 1M NaCI. The effect of elu- 46

triant concentration was studied. The results are shown in Table XI and Figs. 22.and 2).

Run No. 11: Elution with O.lM NH40H. The cobalt was precipitated in the column as cotl8.ltous hydroxide as evi denced by the appearance of the blue color. Zinc hydroxide must have also been formed since the eluate showed neither of the metallic ions.

Run No. 12: Elution with 5 pee cent NaOH. The same results as in Run No. 11 were obtained.

B. Runs Using Amberlite IR-120:

Run No. 13: Elution with O.5M NH4Cl. The results are shown in Table XII and Figs. 24 and 25.

Run No. 14: Elution with 1M NH4C1. This, tod~ was for the purpose of studying the effect of elutr1ant concen tration. The results are shown in Table XIII and Figs. 26 and 27.

Run No. 15: Elution with O.1M ~m40H. Ne1ther cobalt nor zinc were detected in the eluate. They must have been precipitated as hydrOXides in the column as was the case for Amberlite IRO-50. 47

Table II. Exhaustion Cycle at Imler pH Resin: Amberlite mC-50 in Na-form Regeneration level: pH 7 Bed before: run: 4.64 em2 x 56 em Bed after Run: 4.64 em2 x 41 em Flow rate: 5.6ml/em2/min. Influent: Co, O.OO48N; Zn, 0.OO59N pH: 3.5 Co zn mI. tot. ml., 0.113:N Spl effluent N N ~ vol. ,liter spl K4Fe (CN)6 1 21.85 0.00016 2 22.55 0.00033 3 23.25 0.00072 4 24.45 0.00140 5 25.15 0.00290 6 25.85 0.00430 7 26.55 0.00490 8 27.75 0.00920 9 28.96 10 30.16 0.01100 150 0 0 11 31.16 0.00960 1·50 0.8 0.00060* 12 32.56 0.00950 150 0.9 0.00070* J..3 34.76 0.00785 150 3.1 0.00230 14 37.06 0.00850 150 2.5 0.00170* 15 39.41 0.00734 200 3.8 0.00290* 16 41.76 0.00683 200 4.4 0.00340· 17 45.16 0.00616 200 5.3 0.00400* 18 47.51 0.00650 250 33.5 0.00384 19 49.66 0.00580 250 33.0 0.00472 20 52.21 0.00550 250 33.0 0.00529 21 57.06 0.00630 250 33.5 0.00450 22 _ 59.41 0.00550 250 32.5 0.00500 23 61.76 0.00510 250 32.5 0.00570 24 64.26 0.00530 250 32.5 0.00530 25 66.61 0.00510 250 32.0 0.00540 26 68.96 0.00560 250 33.0 0.00560 27 71.81 0.00520 250 31.5 0.00530 28 75.66 0.00520 250 32.0 0.00550 29 78.01 0.00520 250 32.0 0.00550

* Determined by H25 method. ~ ~ f t-I- t- -I-I-+-l-~I 1-,- _~ 1- r'-1_ - ,'.::.. 1- 1 , I f- I- h- J', I .. j :.. • 1- i I• - :I / r •I + •• !- , • ..~ -. I, I .' I , r ft- . I , I --

~ ~ ~ -, ;;. t! I, I", .•t' r 11' -:- !. rt:: .~. • .Hi at r\' II-- . ~ "J 1- I ~ ~ Tn!4'1~j:IR~t=J&t-'+--+--+-+-H-t-t1I--t-+-+t::r+-+-t-H1 "'0 .':;. 1:'" ':0: : , t ~ ., ~. , li:, .. . 'r d ', l "!/--T'It-+-t~t-T-t-+-r-t-t-HH-t-+-toot-t-4-'!-'-t · •• t ~~ , •,,•. ~.... "r • ~ t t t • •. • .• , co • '''''! ,,~r - ., ~, ,-"., . i t ~ --+--+-r-r r-T-!-I-nLH-rl-++++++++-H-+++++++++t--t""++-IH-+t-H ~: -, . -e. ::: .,. -j- ~ J i f-H·\.~-+-+ - i-,.-jH--t-++--t+++-+-I-H-i-+f.-4++-H'IH-I-H'-i ~, •- .--•.-H- ., .-,. -, t- c-! f: I !, H-+-+-+I\.'.-I-+-+-++++++-+-l-++++++++++++++d-+-l-+-+-i · .. ' ...+-~ ,-. •• :..: : ~ :--:'"'-rt-r-It-+--jf-H~l-I'-++++-+-+-t-Hf-H-++-t-++-+-+-HH-I-+-+'d++-+-t-I ·. .. . .- ...... 1 ...- -+- " • I! ..- f~ ·.. -L:...... •.. ..'.t- .. ~ • .-'-·H'~-t--H-++++-H--H-++++++..-+++-l!'t++-++i • ., ..... i_~ •• :•• ~ • '!-r :'.- ~ ·., -t:' I' r" ~·I-I-!-t"oo1'.,-h-H-H-f-HH-l....j-T,....l.-f++I'\oiI"'"l-tM-I-'-''-H

- ~..... T •• •• j'f I- I- ' '-:-~!-r' -,....'. I I• 1- ~ ,·t i' ttt!! 1 I' r H.-L.-' :. - "=:! f:1' ,.. t • 1 . r· I I- J: f! I 1-l-1-!-"-+-t1-H-t-t-f _! L ,--s-t.' ~ -• It.:, ~:-:.. t-1-.' 1_ too ~I- ...... +-!-H-·I-+-...... I-i-.- 1""t-~-H-t-+-+-1'"'ioo...h+++-;-+-+-J r-i-,...,.....-1f-+-,r-\-i -r t • ;. I .- r-rH-IH-J'-; ;:: ~;:: I~ ....f-; t 1- - ., '" n,~-+-+-t-t--"-1-+-1--"-1-+H1-l-l t. ;, ~, ,.~ . ," H-I- -I ,;I '-+-t-+-~++-tl-+-t-+r-r-I I' H-H-H-I I ,. I I' PI ~ H-+-J aj: ~+++-Q*-I-~+"-+-~+-l-I-I -mzlo"k:~'j]'~E1m~rrtL-1 ~ r" It-HrH-H-H++-+-+-H I• '"? . I

•• I I ~ • j I • • _t ~- I. ~ t- .-, •I , : t-f - J . . , ..., 1 ,-~- 1-1- .I 1_ - • ,.. , .. I- I- t •• I • f .- , H-l-I-If-H-I I t '..LL '1.- t t- j , 1 H t· t· ! I' Table III. Exhaustion cycle Resin: Amberlite IRC-50 in Na-form Regeneration level: pH 7 Bed before run: 4.64 cm2 x 56 cm Bed after run: 4.64 cm2 x 31.7 cm Flow rate: 5.4 ml/cm2/min. Influent: Co: 0.0048 Zn: 0.0053N pH: 5

Effluent E~fluent cone. , norma.1 liters Co zn 30.750 0.00002 0 31.010 0.00003 0 31.730 0.00004 0 32.750 0.00011 0 33.950 0.00023 0 36.410 0.00091 0 41.010 0.00440 0 43.610 0.00470 0 47.210 0.00610 0 54.660 0.00900 0 55.660 0.00970 0 58.560 0.00970 0 59.360 0.00900 0.00032 60.160 0.00930 0.00041 60.960 0.00870 0.00050 61.885 0.00870 0.00050 63.585 0.00850 0.00076 65.885 0.00810 0.00140 70.185 0.00750 0.00230 74.485 0.00670 0.00320 79.785 0.00610 0.00450 84.085 0.00570 0.00480 86.885 0.00530 0.00540 88.185 0.00600 0.00430 89.485 0.00570 0.00450 90.785 0.00570 0.00470 92.095 0.00580 0.00470 95.245 0.00570 0.00470 I, I I I I'

~ 0. ~- .p • >:: Q I ~ • , . .-4 ~ •• ~ I- s.. I ~ .r{ -"'i- ..ct I • I 51

Table IV. Exhaustion cycle Resin: Amberlite IR-120 in H-form Regeneration level: 30 meq.H2S0~ml. resin Bed before run: 4.64 em2 x 30 em Bed after run: 4.64 em2 x 27 em Flow rate: 3.45 ml/cm2/min.

Influent: CO: 0.00493N Zn: 0.00509N pH: 4.2 ml. Effluent Etfluent cone •• normal BPl volume. liter Co Zn 265 25.580 0.00032 0.00033 725 27.078 0.00176 0.00267 265 27.970 0.00352 0.00410 275 29.818 . 0.00451 0.00462 910 30.411 0.00462 0.00477 500 31.116 0.00462 0.00477 620 31.676 0.00462 0.00477 500 32.236 0.00462 0.00477 960 32.966 0.00462 0.00477 ii' ; , t . j f 1 ") f' , 52

I I I •I I •• I I t- • • I· .-. . I 1 •I 2lJ

I, .I 0.006 ,....----~----_r_----T_---__,

f1.g. 9. l;¢C\.~sti., n Oycle'

------, . Zn

I .I Co

---+------

I ,I,

---:1 R~Bin: ArIDerlite iR-l~ (B-fo~) ....I ned hefol''9 run: 4~64Cmf :x 300m 3ad after run: 4.64cr. x 270m l P10'ff' rate: 3.45 ml/cl'12/min. ~nf~.rttt p:- J 4~2 .001 to: O,OQ1G~ eq~iYa_ t~: U.005~D9 f~liv. --I

.. ~9 31 Z3 fluent volw~, liters i I-. , I ~ 1 I I _~J -L . -. , 53

Table V. Elution with sulfuric acid Resin: Amber1ite IRC-50 Bed: 4.64 cm2 x 36.7 em Total Co in column: 0.15810 equiv. Total Zn in column: 0.16200 equiv. Elutriant: 0.5% H S0 2 4 Flow rate: 2.16 m1/cm2/min. %of tot. mI. m1.tot. Normality No. of eqUivalents ~e=l=u~t~ed~_ ep1 eluate Co Zn Co Zn Co Zn 155 155 0.00007 0 0.00001 0 0.01 0 145 300 0.00533 0 0.00077 0 0.48 0 100 400 0.02500 0.00610 0.00250 0.00061 2.14 0.37 100 500 0.03900 0.01430 0.00390 0.00143 4.66 1.26 100 600 0.04500 0.02490 0.00450 0.00249 7.61 2.80 100 700 0.05400 0.03010 0.00540 0.00301 11.12 4.65 100 800 0.07020 0.03730 0.00702 0.00373 15.70 7.85 100 900 0.07400 0.04580 0.00740 0.00458 20.55 .80 100 1000 0.08533 0.05220 0.00853 0.00522 26.00 10.55 100 1100 0.09333 0.06540 0.00933 0.00654 32.20 14.53 100 1200 0.10470 0.07480 0.01047 0.00748 39.00 19.30 100 1300 0.11000 0.08270 0.01100 0.00827 46.20 24.25 100 1400 0.12000 0.09100 0.01200 0.00910 54.00 29.90 100 1500 0.14000 0.09700 0.01400 0.00970 63.10 36.40 100 1600 0.14500 0.10480 0.01450 0.01048 72.60 43.50 100 1700 0.09500 0.11530 0.00950 0.01153 78.80 51.60 100 1800 0.07400 0.13240 0.00740 0.01324 83.60 59.36 100 1900 0.05300 0.12400 0.00530 0.01240 87.10 69.90 100 2000 0.04600 0.10550 0.00460 0.01055 89.42 71.30 100 2100 0.04300 0.08730 0.00430 0.00873 91.50 76.50 100 2200 0.03960 0.08520 0.00396 0.00 52 94.30 81.50 100 2300 0.02800 0.08520 0.00280 0.00852 95.70 85.70 100 2400 0.02520 0.08000 0.00252 0.00800 97.65 90.60 100 2500 0.02120 0.06290 0.00212 0.00629 98.10 93.50 100 2600 0.01720 0.04640 0.00172 0.00464 98.70 95.00 100 2700 0.01240 0.02530 0.00124 0.00253 99.30 96.10 100 2800 0.00680 0.01870 0.00068 0.00187 99.70 97.00 100 2900 0.00440 0.01240 0.00044 0.00124 99.90 97.20 100 3000 0.00060 0.00380 0.00006 0.00038 99.95 97.45 100 3100 0.00034 0.00150 0.00003 0.00015 99.97 97.50 100 3200 0.00030 0.00150 0.00003 0.00015 100.00 97.55 100 3300 0.00000 0.00080 0.00000 0.00008 100.00 97.60 rrlT'll~I' ., .,. .I '/ . •I 54 I --rl·····. ii' I I I I I I j t I' !: r I .15Ot------+----...... ------4----~--.·

Fig.lo.1 II tion with RfS04 I

',)1.:; ~ .1 - ~[

I• I I

-+----+----~.~ -'I----i .)(() r

1 I -I __ __-+-_~.-.._+--...... ----..L--__---L----..L.-~w- ...... J ~-r I ...... 14:: I 21:( 2:CC

'I:h~r.t~ -, 1urr.e J 7111.1 . \ . ! ~~9in: ;~.~l ita !~C-~ , . ~ 1,' ) ~,,.. • 3. .,/k. :..l...':r.c . •. -:';~-+----J,4U----- ~qtal Co:~ olur..~: o. c~u1v • .:.'dto.l Zn ~n col"lann: C. 620 eo liv. :::: futr_~.llt: C .5~ :J'-'~.:l ~c r.cid ? O\'T ;-- te: 21.:....;I• rl en',')Vnin. I,

t ", ., , II,

I , I I ; • -~- -+--...-+,...... --~+_...... ,..._.---t------i r-~ I' "; II ' I:. I I• " . •I,,, l I ••

I•• •• • If;

. ,,,: I I :!' i' f1£;- 11 ~ E utlbil' ~11.t. .. ll2S04; :: ~J.UG.te vo ume va %: ~~t$~q: ---J-- -1001---+'- _.' .- -. -~---t------t-r=:;~--+---+-----t I , Ii'

~ I' I :: ~ i

em 'x :;6.7 e

$m11om2/m n.

350

. , • a.: ...... --+-_.,---. 56

Table VI. Elution with 0.5M NH4C1 Resin: Amberlite IRC-50 Bed: 4.64 cm2 x 24 em Total Co in column: 0.06870 equiv. Total Zn in column: 0.06540 equiv. E1utriant: 0.5M HH4C1 at pH 4 Flow rate: 0.5 ml/cm2/min. %of tot. mI. ml.tot. Normality No.of equivalent eluted BPI eluate Co Zn Co Zn Co Zn 105 105 0.06610 0 0.00694 0 10.1 0 25 130 0.24600 0 0.00615 0 19.1 0 25 155 0.25800 0 0.00645 0 28.4 0 30 185 0.21700 0 0.00651 0 37.9 a 27 212 0.16300 0.00210 0.00440 0.00006 44.4 0';1 25 237 0.13600 0.00500 0.00340 0.00013 49.3 0.3 50 287 0.11900 0.00920 0.00595 0.00046 58.0 1.0 50 337 0.08960 0.01120 0.00448 0.00056 64.5 1.8 59 396 0.06720 0.01300 0.00397 0.00077 70.2 3.0 50 446 0.04250 0.01100 0.00213 0.00055 73.1 3.9 54 500 0.03660 0.00920 0.00198 0.00050 76.2 4.6: 50 550 0.03240 0.00770 0.00162 0.00039 78.4 5.2.2 50 600 0.02730 0.00760 0.00137 0.00038 80.5 5.8 50 650 0.02310 0.00743 0.00116 0.00037 82.1 6.4 100 750 0.02060 0.00728 0.00206 0.00073 85.2 6.9 100 850 0.01550 0.00710 0.00155 0.00071 87.5 7.7 100 950 0.01120 0.00695 0.00112 0.00070 89.0 9·1 100 1050 0.00930 0.00680 0.00093 0.00068 90.5 10.1 100 1150 0.00930 0.00661 0.0009~j 0.00061 91.8 11.1 100 1250 0.00740 0.00647 0.00074 0.00065 92.9 12.1 100 1350 0.00580 0.00632 0.00058 0.00063 93.6 13.0 100 1450 0.00480 0.00620 0.00048 0.00062 94.4 14:.0 100 1550 0.00360 0.00600 0.00036 0.00060 95.0 14.9 100 1650 0.00360 0.00588 0.00036 0.00059 95.5 15.9 100 1750 0.00270 0.00574 0.00027 0.00057 95.9 16.7 100 1850 0.00280 0.00580 0.00028 0.00058 96.2 17.6 100 1950 0.00200 0.00550 0.00020 0.00055 96.5 18.4 100 2050 0.00160 0.00540 0.00016 0.00054 96.7 19.3 100 2150 0.00150 0.00530 0.00015 0.00053 97.0 20.5 225 2375 0.00090 0.00520 0.00021 0.00117 97.2 21.9 175 2550 0.00069 0.00500 0.00012 0.00088 97.5 23.2 210 2760 0.00046 0.00485 0.00010 0.00102 97.6 24.8 215 2975 0.00038 0.00470 0.00008 0.00101 97.6 26.3 213 3188 0.00022·· 0.00446 . 0.00003 0.00094 97.7 27.8 222 3410 0.00012 0.00423 0.00003 0.00094 97.8 29.2 250 3660 0.00008 0.00410 0.00002 0.00103 98.0 30.8 180 3840 0.00000 0.00382 0.00000 0.00069 98.0 31.8 I l I"~ I I I I I I ,- TI '11!! 1 t.14--l-~-~-I 'I .I r I I I- I I I I I, I ! IJ- 1- - , ri' j j I ! ::! :: l ___~_-.LJLW-+-,-!-,-L l-":'-':+-I-+-+-++++-Ho,-+I-+,T T,++-,t+i4,-J.-t,"';"t-r-.i-r!,~,I~.-,-.-.-,---J. t I I ,;I I I II i 11- f I :.,., ':,'. . .• j II r '. f. ,. :: I I I j i H- ;! I,•, •• I 0.26 I I - - r: -T-- v.e. 0.2_

~ .-~- ~v - • j .. !. .. - -. "lit t{ ...

~-- . e in: ~~~ lte IE045< :: ~. 0.18 • t • t T'- ,- . -- . ., ; 4. lJ:.qrrrJ'f:~ 0 : .- .. :. 4

I- ..

~- - . .a~ -"...... - H I-H-+++t-I-Hf-H'++-~''''+-+-+-Hi tTl -t+-H-H -0 " Co • . o " +- I • •II, 9.10 I -.--- . . '.: .;-: lL f-.-·HH-+++-l+t+I++++f-I-H--t++-f+f+i-+-I-:....jI -' . -t- •• r.....-:-~·H..:l-4-i+++t-+-I-++-H-..l.-f-+-H:'-~~~..,....';"-~I I I .~' .- - t- · ro. ,,0 . I I . - .-~ '.. .. I --.. .--- • J~- r,t',ij.:t-=:tttt~I-.'~~i.:..,t-:.j · _.'- . f • ~ + • .. fl' 1-t · - .- . • p 1-t-l-4-+-...;-j..II-H-i-+-H-~ - ~.~ - . 'g.O I} 1 t I, ~ --- ....-;--:.....-+-1. - '\ ." · '-I' ...... ~ I •• : =- : i ...... i' t':; T- .- 1- I I .. .•. • i I ~ - -;-.-

R=-=' I• 1-· .., •• I• • 1- . ,. . '1-1Ii+~-H- •. ,.. ;_. . •. t- o· , • • •. 1-

1- •• +- I ~. h-. ~H -t-H-I-t+; .. t f·. • ";: :-"r; II',-,,! i'-l-J.-I--I-+. ~ -. t· • I .••.••• ~k1-1- ~ . ~- +4-1-,4-!---.J.4-1 ;~ H4~~ •• I ':.::' ".: I t- t-. , .. r f • i -j - I' •• 1 1 I ·-.-+-Ir -+-+-I-I-4-4-1 't"'f lll •• ! ,

1 II ,I", . , , ,, I I 1 I I'. l r ' , 'I I I ! I II rI

It. t

, -,---1--+0""'flo~t- I,., V

~-- vv,." I --

...-

I I I , z.n -- -._- .- :/ .. o - / I 1 2 3 4 5 --

------+-----+------+----'---:-,I +-----:---I------t-----i I ,. , ,. i. ' I . . ," t. . " I , I .. ! II ~- , . ._------t-.-----+---_--o-.-.---f------j t----l ,I I t I•I •I• t II II

I t " ! I ~ ....L...J-.~ "-"-,'--"'--'--'i~ ~ L --'- --1.- 1· ...... j' I ...... __ 59

Table VII. Elution with 1M NH~Cl* Resin: Amber11te IRC-50

Bed: 4.64 CE2 x 27 cm Total Co in column: 0.0805 equiv. Total Zn in column: 0.0829 equiv. Elutriant: 1M NH4C1 at pH 4.5 Flow rate: 0.43 rnl/cm2/min. Cobalt. mI. rol. tot. number of %of tot. epl eluate normality equiva.lents eluted 85 85 o 0 0 53 118 0.00084 0.00003 0.03 25 143 0.00301 0.00008 0.13 50 193 0.00535 0.00027 1.46 50 243 0.00715 0.00036 1.90 100 342 0.02600 0.00260 4.14 50 393 0.03110 0.00156 6.06 100 493 0.02270 0.00227 8.90 50 543 0.01810 0.00091 10.00 50 593 0.02640 0.00132 11.65 100 693 0.02360 0.00236 14.60 50 743 0.01680 0.00084 15.60 100 843 0.01930 0.00193 18.00 100 943 0.01600 0.00160 20.00 100 1043 0.01740 0.00174 22.20 100 1143 0.01640 0.00164 24.40 200 1343 0.01530 0.00306 28.00 200 1543 0.01430 0.00286 31.60 ~O 1843 0.01260 0.00378 36.20 360 2203 0.01150 0.00414 41.30 550 2753 0.00945 0.00520 46.80 965 3718 0.00756 0.00730 56.80 548 4266 0.00700 0.00384 61.70 490 4756 0.00685 0.00336 65.80 574 5330 0.00577 0.00331 69.90 670 6000 0.00494 0.00331 74.20 562 6562 0.00414 0.00233 76.90 675 7237 0.00350 0.00236 79.80 450 7687 0.00273 0.00126 81.40 830 8517 0.00260 0.00216 84.20 344 8861 0.00231 0.00080 85.20 860 9521 0.00227 0.00195 87.60 595 10116 0.00218 0.00130 89.20 745 10861 0.00180 0.00134 90.80 710 11571 0.00155 0.00110 92.20 * Only a trace of Zn wa.s detected by precipitation with H2S• , ' •I• ,I 1 . " . . 60 I

F:'--: ._14. _snutiqn i:~.t;: l.~ rl~Cl _

( 31uc to CCUC. 3 01 1=- t.c: vlc,lt~c) I I ~03~1~:------Alrmc

-Jod: 4.";4 C::12 v ~~( r . .ch IElutr~~nt:: 1.:: ItI4Cl t ~ 4.5 ~/min. IFla•1 r.:. to:1 •

I • '.. .. ~ I l---+-rli~ I I I .. I '/ C' .. '1 <; 2 I•4 • ~ 1.~a.1I. I1tarn l- I i

•• 'T ~ t. 1 • . . I . . , .

IIII ..-. ....s:l lD 'd ~ !

t- •

t') ~ 62

Table VIII. Elution with 0.5],1 Na.Cl~r Resin: Amberlite IRC-50 Bed: 4.64 cm2 x 27 em Total Co in column: 0.08573 equ1v. Total Zn in column: 0.08625 equiv. Elutriant: 0.5M NaC1 at pH 5.0 Flow rate: 0.43 ml/cm2/min. mI. ml. tot. Cobalt': BpI 01 ate normality number of %of tot. equivalents eluted 55 55 0 0 0 28 83 0.00028 0.00002 0.03 20 103 0.00264 0.00005 0.09 50 153 0.00797 0.00040 0.55 100 253 0.01345 0.00135 2.10 54 307 0.01175 0.00064 2.80 51 358 0.01365 0.00070 3.70 120 478 0.01470 0.00176 5.70 100 578 0.01810 0.00181 7.85 100 678 0.01720 0.00172 9.85 50 728 0.01720 0.00172 10.90 100 828 0.01470 0.00147 12.50 100 928 0.01600 0.00160 14.40 108 1036 0.01510 0.00163 16.30 100 1136 0.01430 0.00143 18.00 100 1236:- 0.014-50 0.00145 19.70 100 1336 0.01280 0.00128 21.20 200 1536 0.01240 0.00248 24.10 375 1911 0.01410 0.00529 30.20 315 2226 0.01220 0.00384 34.70 244 2470 0.01050 0.00256' 37.70 270 2740 0.00982 0.00265 40.80 275 3015 0.00903 0.00248 43.70 325 3340 0.00798 0.00260 46.70 450 3790 0.00715 0.00317 50.50 437 4227 0.00630 0.00275 53.60 660 4887 0.00588 0.00388 58.20 865 5752 0.00494 0.00427 63.10 840 6592 0.004-10 0.00244- 67.10 500 7092 0.00336 0.00168 6 .00 660 7752 0.00315 0.00208 71.50 800 8552 0.00301 0.00241 74.40 620 9172 0.00252 0.00156 76.10 718 9890 0.00245 0.00176 7 .20 650 10540 0.00217 0.0014-1 80.00 , f I I, ,I I •,I I . , 63 I I .I .II I •I•

0.020 I I Fi_.l6- Elut~ n of c0bf'l t r~nd za..rc I " j !< E;lUllto cohee VB eluhte volU!lle I I - -, -! ' ,-- .- ---4--+---i t· ~~o~in: Ambprlito IRCr50 F{ ..". pOd: 4.64 mE x 27 c~ ~ I 8 i '" . \'. Elutr'innt:, O.51~ HaCl at pH ? L_ 0- .-tr. :r1 I • FIOl"' ritt10' 4Jj mIle 2/ml.lr." I I e I I +' I '\ ~ Q) I , 8 I I 0 r~'------.s BS ~ I' g ~._~ • ·-r--- I

z 4 a :na-.:-' ..

.-.- -I------+------+~.:...-----.--.r_-:-.--- , t· . .. I .::.:! I• t I, •• I ••• I HI: .. :;. :. ! Ii! i I•

---L -L... .L-:: ...... ';:::::.;.....;_:_;_:_iI"....L..-.1..L·_.--.L.,;_~~~I.-..J....J_"~ ; . ..: .:\ L._' .. _... w. . ~-,-. ,I t , 1---;- I I' t ' I I r I . --- - t .. - .- -

~- 1OO.....-----+----4----+--...... -+--....-~ .j 90 IF'-S·17. E ution of ooalt and zine I I- -- -+-~HI-- ~f tot 1 elu~ed_ " eluate ;_0_,~_':lf1!:_e_}_-+-_--I-~...,--t

----t'I--.l:tA+-

-1-- - 65

Table IX. Elution with 0.5M NaCl at higher flow ~te. Resins: Amberlite IRC-50 Bed: 4.64 cm2 x 19.5 cm after reclassification. Total Co in column: 0.07349 equiv. Total Zn in column: 0.07510 equiv. Elutriant: 0.5M NaCl at pH 5 Flow rate: 2 ml/cm2/min.

Cobalt mI. ml.,tot. number of %of tot. BPl eluate normality equivalents eluted 50 50 0.00202 0.00010 0.13 50 100 0.06468 0.00323 4.51 100 200 0.06720 0.00672 13.35 100 300 0.03675 0.00368 18.70 275 575 0.02360 0.00650 27-.40 553 1128 0.01302 0.00720 37.20 710 1838 0.00903 0.00641 45.90 987 2825 0.00504 0.00497 52.60 650 3474 0.00389 0.00253 56.00 760 4235 0.00329 0.00250 59.45 875 5110 0.00263 0.00250 62.85 900 6010 0.00236 0.00212 65.60 720 6730 0.00215 0.00155 67.90 910 7640 0.00189 0.00172 70.30 870 8510 0.00168 0.00146 72.20 960 9470 0.00139 0.00133 73.90 940 10410 0 .. 00124 0.00117 75.50

0)(. No zinc ",as detected by precipitation with hydrogen sulfide

67 1 I

::.~~_~ r------;..-.------+------.,

ric::. ~-90 Elution \\'ilil o.m: lIa.cl nt 111$hol" flow rata.

(rerc611 ~ elu ted vs elua.to ml.lIDO) , ,

~ .s ~ I , . .:{ .,.. I%l --+ ---j ..... /~ .I ~ .s /" ~osin: Al"ilorlU,e I~C-~a fo4 0 :ed c 4.64 ICl} :x 19.5 om 4-' ;~.1 s:1 ,ID Elutriant.: ().61~ ~l t pE 5 1°l., I%l fl.. 2 Flow I'llte ~ 2 rril/c.ln / n.

f"' ------~---~------_--1 ... a I ., j ... :c i : ::U\.lQ. te vO~'.::Ie,n ters

I_ I I· I

I I I -. '1 ~-;- ... - .. 1 I I I ... I I I ~-_._- ._.-1 68

Table X. Elution with 0.5M NaCl at higher pH')! Resin: Amberlite IRC-50 -,ed: 4.64 cm2 x 17 cm Total Co in column: 0.03587 equiv. Total Zn in column: 0.03733 equiv. Elutriant: 0.5M NaCl at pH 7 Flow rate: 2 ml/cm2/min.

Cobalt ml. rol. J tot. number of %or tot. apl aluate normality eqUivalents eluted 50 50 0.00067 0.00003 0.08 50 100 0.00294 0.00015 0.50 100 200 0.00294 0.00029 1.31 100 300 0.00280 0.00028 2.08 100 400 0.00269 0.00026 2.82 253 653 0.00233 0.00059 4.46 650 1303 0.00227 0.00148 8.60 690 1993 0.00189 0.00130 12.20 730 2723 0.00201 0.00147 16.30 780 3503 0.00214 0.00167 21.00 850 4353 0.00214 0.00182 26.00 1180 5533 0.00214 0.00253 33.10 900 6433 0.00214 0.00193 38.40 850 7283 0.00210 0.00179 43.50 820 8103 0.00210 0.00172 48.25 800 8903 0.00197 0.00158 52.60 980 9883 0.00176 0.00172 57.50

i: No zinc wa.a detected by preci pitation with hydro~en sulfide 69 I t I l ,I~'r"

I 0

, ., I

r •I

F il.~" 20. "$.\1 tion th L5:,! !,il,C 1 at !:igher pI1 I C~ ..~ 1ffil-u·JQ&-i'--- o .Hoe5 {Elu.a1M eluate

Co I l- ·..oo~e , -. I

ita !

oX 17 em I -~~I

.,,; :I I :, .. ~-- I

•I I 0 -r------"j------t---"'H't"r-+~~--+----~-..L- I I 0 2 10 I I. c ~;~ 0 "I. 0 I 1: 't t ... ~ 0': h o .".. jj: ::.. • - r--,... ..-' ." \ ..L...~~...... L...__.1 ---..1------l... _ .••• _ 0 ,_ • r

I Fig.21. Elutioh \fitJ O.ft·: lIaCl e. t 3igher pI! (~eluted 'Vs elun+-.e vol\.lrle~ 1 ------j---+---

!lesin: Amberli !~-50

"d v ,I-' ;j .-I CD F10l'T rate: 2 !"'l/c~/nin • ~ lS ~ 6G f- 0 ~ I .~ I ICD I~

-to -----

I /i ;.7' ------T- J ./ I I . '

4 6 1 8

-j I I . r I

------. --.------L_: L--_ 71

Table XI.E1ution with H/I NaCl Resin: Amber1ite IRC-50 Bed: 4.64 cm2 x 19 em Total Co in column: 0.05610 equiv. Total Zn in column: 0.05778 equlv. E1utrlant: 1M NaCl at pH 5 Flow rate: 0.43 ml/cm2/min.

%of tot. IDl. IDl. tot. orma1ity No. of eguivalents eluted spl eluate Co Zh Co Zn Co zn 45 45 o 0 0 0 o 0 20 65 0.00256 a 0.00005 a 0.09 0 50 115 0.02265 0.00052 0.00113 0.00003 2.15 0.06 50 165 0.02353 0.00054 0.00118 0.00003 4.60 0.12 60 225 0.02940 0.00057 0.00176 0.00003 7.35 0.18 58 283 0.02100 0.00060 0.00124 0.00004 9.55 0.24 46 329 0.02900 0.00065 0.00133 0.00003 11.90 0.30 101 430 0.03233 0.00069 0.00327 0.00007 17.75 0.40 90 520 0.03400 0.00073 0.00306 0.00007 23.30 0.52 110 630 0.03150 0.00078 0.00347 0.00009 29.40 0.68 350 980 0.02710 0.00082 0.00949 0.00029 46.35 1.18 485 1465 0.01763 0.00085 0.00856 0.00041 E1.55 1.89 616 2081 0.01092 0.00090 0.00673 0.00056 73.56 2.86 685 2766 0.00672 0.00094 0.00460 0.00065 81.77 4.00 832 3598 0.00427 0.00099 0.00355 0.00082 88.00 5.40 590 4188 0.00280 0.00101 0.00165 0.00060 91.00 6.45 620 4804 0.00220 0.00105 0.00136 0.00065 93.40 7.56 295 5103 0.00164 0.00108 0.00048 0.00032 94.30 8.10 326 5429 0.00151 0.00110 0.00049 0.00036 95.10 8.75 229 5658 0.00120 0.00112 0.00038 0.00026 95.75 9.20 280 5938 0.00105 0.00115 0.00029 0.00032 96.25 9.75 635 f$573 0.00080 0.00119 0.00051 0.00076 97.30 11.10 560 7133 0.00067 0.00123 0.00038 0.00069 98.00 12.30 645 7778 0.00054 0.00127 0.00035 0.00082 98.60 13.70 705 8483 0.00041 0.00131 0.00029 0.00092 99.00 15.30 895 9378 0.00025 0.00136 0.00022 0.00122 99.50 17.30 750 10128 0.00014 0.00141 0.00011 0.00106 99.70 19.20 882 11010 0.00006 0.00148 0.00005 0.00131 99.80 21.50 560 11570 0.00000 0.00156 0.00000 0.00087 99.80 23.00 ..~-- ~ •I T~~ I .. I I',

')G

~lU4101

•I _~onci~_~~_ 011.lD.~d I~~ , ,

I i I

x 19 OJ',1 , I; t 1------+- - _. - ~- I ~t p riant: lH Ua.Cl :5: .. I r I • 1 ') r.tl/am2j

-r ------+------+---+------1 I ... I a I I , t , t I

. ,

I--~-~-~..-----

r o I 3 : ,11 !. ato VOlllil-~ 'litars . I

• , , I ; i { I , • I, I I , , I ,I j • I I :,1 !It,I; ,II' I I ! ~-~ ,-- r '

r' J

4 _ , " ---

-- • t"---,.-.' "f" '"t iT" ~ :- 1 .. r t r . ~ ., ,-+-.-. ', \J., .. , ,," .,, r'") .. - ,." .- :, ".: . ";!:::: ~Tt= I"'- .: ":, I ,"'" ,1 "-;- (> ~' .. i : ~~ 'I :;":,. 1 r r "] -r-~.,' , • • ~-', 1 I '" ,l l'~ •,~ j I I .'. , . I I I ;--1-, •:; ;-1:- I :::: :ii >--- ' ~ ~,~ ' .... :" ~i -~''-r ': '" iT" 1=1.~.'j:: 3i ' :'4 : .: .:;: 'J-ft- 1- :-{lli-TT1 tf\_ -.... " .. • ••• I I r I t- t- I. [-+++-1 I nf" '~I ,• .,., .. ,...I II ',,I', I., "" '"

-'-

L.-__ 74

Table XII. Elution with 0.5M NH4Cl

Resin: Amberlite IR-120 Bed: 4.64 crn2 x 21.5 cm after reclassification Total Co in column: 0.08183 equiv. Total Zn in colu~n: 0.08567 equiv. Elutriant: .?51JI NH4C1 at pH 4.7 Flow rate: 2 rnl/cm2/min.

%of tot. rol. ml.tot. Normality No. of equivalents eluted spl eluate Co Zn Co Zn Co Zn 50 50 0.00074 0.00131 0.00004 0.00007 0.05 0.08 100 150 0.11256 0.14310 0.01126 0.01431 13. 0 16.80 100 250 0.18060 0.23610 0.01806 0.02361 35.90 44.40 120 370 0.13650 0.14157 0.01638 0.01699 56.00 64.20 100 470 0.06300 0.07710 0.00630 0.00771 63.60 73.00 460 930 0.03024 0.02874 0.01391 0.01322 80170 88.60 875 lLJ05 0.01176 0.00836 0.01029 0.00732 93.20 97.10 890 2695 0.0040.... 0.00218 0.00364 0.00194 97.80 99.40 350 3045 0.00205 0.00086 0.00072 0.00032 98.50 99.70 390 3435 0.00131 0.00040 0.00051 0.00012 99.20 99·90 985 4420 0.00050 0.00003 0.00048 0.00003 99.80 100.00 650 5!l70 0.00011 0.00000 0.00007 0.00000 100.00 100.00 _ .. ~ 1-·" • ., T 75

~~8------""';-----~----+------' , 0. ~ ,'1",. 24. ,01",:" ",j \,! tf, o. 51:! l'HlfC1

...~_, .., ...... c r "C ( ,l I '/1-- _'...

• '. ' ~ I. .2 - '"11 r C• . ""CL. 'r.lJ'.· Cl... - '-_ • .,J .... 8 r.~·tc:" :"'CC =.~.r'. r.: ~::-~C:.t~O) J 0.16 o .....p 1 f! :'J.01'; 1 [', to: '),- l'..-/C.,:_/... .., n. .0.> r:1 Q) o g 0.12 Co) l --'L-I.-----1

I r-- .--~ Co

. ,

t- .... 1 -

---+------. - -. ---f------+++----+------1 . .- .J ~. ______.-l---l. --.J:..-.__. _._'_ ::: r • t .I r I 76

-, . - _.. ,.,.- ( .:-=:: _. __ /:.J_

;',(:(~. ~ -~- ~e I 1.,..,\ _ •. VG 0,1 ;L;, :.,f] ol\.:,j ) , ·c· _0- .~ I I !.r . ret.s ;;1 ,.-l Q)

r-{ Jj >- .B r.-. 0 1;; II) 0 J-. Q) -~ -:y..> --~

I ___ I

t. I '"T" I t,;. ';(J''':"'~' 77

Table XIII. Elution with 1M NH4C1

Resin: Amberlite IR-120 Bed: 4.64 cm2 x 21 cm after reclassification Total Co in column: 0.08700 equiv. Total Zn in column: 0.08875 equiv. Elutriant: 1M NH4C1 at pH 5 Flow rate: 2 ml/cm2/min. %of total rol. ml.tot. Normality No. of equivalents eluted -snl eluate Cd Zn Co Zn Co Zn 50 50 0.00210 0.00289 0.00011 0.00015 0.12 0.17 100 150 0.24360 0.32330 0.02436 0.03233 28.20 36.60 100 250 0.26250 0.34298 0.02625 0.03430 58.00 75.20 100 350 0.13125 0.13140 0.01313 0.01314 73.40 90.00 100 450 0.07035 0.04755 0.00704 0.00476 81.50 95.50 260 710 0.03306 0.01040 0.00860 0.00270 91.50 98.50 345 1055 0.01390 0.00234 0.00480 0.00080 97.00 99.50 690 1745 0.00282 0.00062 0.00195 0.00042 99.10 99.90 510 2255 0.00052 0.00000 0.00027 0.00000 99.40 9.90 410 2665 0.00010 0.00000 0.00004 0.00000 99.50 99.90 365 3030 0.00016 0.00000 0.00006 0.00000 99.60 99.90 580 3610 0.00000 0.00000 0.00000 0.00000 99.60 99.90 ,-+ ••

~-,~I ~ r-.: ,f-.-- -I 79

::'00 _.....---=:==L----.--.----. Co

·0

' - ...' " ( ,> 0. 'JL"-",... C.,,1

1" IT ,.." ,J. -" , :ipJ.- " v

4_'), ···l/c-:·/-_ .... _(..,,.:..

u o

- .. -'\ '" ~."" _:. I• '1"""\ - -oJ __..,-'J 80

DISCUSSION

1. Selectivity of The Res ins: Figures 7 and 8 show that the breakthrough point of co balt appeared much earlier than that of zinc, indicating that the adsorption affinity of the carboxylic acid resin, Amberlite IRC-50,. for zinc is stronger than for cobalt'. Th:is is in agreement with the order of adsorption predicted by Boyd et"al." using activity coerficient~concentrationplots as the criterion (p.20). Cobalt and zinc are both divalent and do not differ greatly in their atomic weights (The atomic weight of cobalt is 58.94 a.nd that'of zinc is 65.38). Their hydrated ionic sizes will therefore be the decisive factor in determining the adsorption affinity. Such a selectivity, h.bwever"wa.s not shown by Amberlite IR-120 which contains a nuclear sulfonic acid group as the exchan e center. Fi re 9 shows that both cobalt and zinc ap peared in the effluent at approximately the same point. It was thought that the new batch of the resin which had not been pre-conditioned might be responsible for the result. The experiment was repeated using the same batch after it had been acid-regenerated•.The same result was obtained. The difference between the results shown by the resins can hardly be accounted for in terms of the effects of va lence; . atomic '\-leight, and hydrated ionic radius. It seems to suggest that the chemical properties of the cations and the active groups in the resins are of more influence. Thus 81

it could be imagined that in the case of Amberlite IRC-50, the carboxylic acid-zinc complex is more stable than the

carboxylic acid-cobalt complex~ In the case of Amberlite IR 120, the stability of the sulfonia acid-zinc complex is ap proximately equal to that of the sulfonic acid-cobalt com plex. No matter what the cause may be, it is definitely esta blished that considerable separation of cobalt and zinc can be achieved even in the exhaustion cycle by the use of Am berlite IRC-50. The calculation or recovery and purity or co balt and zinc is shown on p.43. 2. EXchange Capacity of The Resins: No particular attention was iven to determining quanti tatively the capacity of the resins. However, Figs. 7 and 8 indicate that the capacity of Amberlite IRC-50 is relatively high and increases with increasing influent pH. The latter is chal'aoteristic of all weak acid type resins. Even in the sodium form, a considerable number of hydrogen ions are a·s sooiated with the acid group of the resins. As the pH of the influent solution is raised, these residual hydrogen ions oombine y;ith the OH- ions in the solution, leaVing more a vailable positions for cobalt and zinc. No experiement was conducted to test the effect of in fluent pH upon the exchange capacity of Amberllte IR-120. It is reported that the effect will not be appreciable since sulfonic acid is relatively stronger than carboxylic acid. By comparing Table III and Table Tv, it is seen that 82

Amberlite m-120 actually showed a much lower capacity than Amberlite IRe-50, a fact contradictory to the manufacture's claims. This could be due to the loss of moisture content of the sulfonic acid resin which had been stored for ten months before use. 3. Elution with Sulfuric Acid: Figure 10 shows that cobalt and inc were eluted simul taneously by a 5 per cent sulfuric acid solution. This re sult is to be expected in view of the tremendous affinity be tween the weak carboxylic acid resin and hydrogen ions. Here is an example in which the adsorption and desorption are so lely governed by the chemical properties of the exchanger and the cations involved. 4.EUut1on ",1th Ammonium Chloride: As seen from Table VI and Figs. 12 and 13, the 0.5M

NH4Cl solution showed a fairly high eluti~~ strength. In this instance, the valence effect, which works against the elu tion, is apparently outweighed by the ma~s action effect of the excess amount of ammonium ions. In the first liter of eluate, 90 per cent of the cobalt was eluted together with only 10 per cent of the zinc, re sulting in an appreciable separation. The significanoe oan be better appreciated from the foll"ow1ng calculation:

A. Purity of Co~lt: Total cobalt in column: 0.0687 x 58.94/2 =2.03 gms. 83

Total zinc in column: 0.0654 x.65.38/2 = 2.14 gms. Purity of cobalt in column: 2.03 x 100 / (2.03+ 2.14) = 48.7% Total cobalt in the first liter of eluate: 0.9 x 0.0687 x 58.94/2 =1.83 gros. Total zinc in the first liter of eluate: 0.1 x: 0.0654 x 65.38/2 =0.22 groa. Purity of cobalt in the first liter of eluate: 1.83 x 100 / (1.83+0.22) - 89.5%

u. Concentration of Co It: Total volume of influent solution passed duri the eXhausting cycle: 13.35 liters Concentration ratio for cobalt: 13.35 x 0.9 : 1 =12 : 1 For Amberlite IR-120, zinc and cobalt ,,,e::re eluted si multaneously by the 0.5M ammonium chloride solution. This is to be anticipated since Run No. 3 showed that the affinity of the resin for both zinc and cobalt was about the same. By doubling the concentration of the ammonium chloride, the eluting strength of the solution was found to be much roduced in the case of Amberlite IRC-50. Ninety per cent of the cobalt was not eluted until ten liters of eluate had been collected. Meanwhile, only a trace of zinc was detected thrOUghout the elution, by precipitation with H2S. This red ction in the eluting strength 1s another phe nomenon difficult to explain. It is true that the hi&ler the 84 concentration of rm4Cl, the less the percenta. e of NH4+ ions entering into exchan e. comparin t.he followin.g activity coefficients of NH4Cl, (1) however,. the lM solution of ammo-

(1) Latimer, W. M., The OXidation States of The Elements, Prentice-Hall Inc., 1938, p.325 nium chloride still contains a larger amount of ammonium ions and therefore should possess a higher elut etre th than the 0.5M solution.

Concentration, molar ActiVity coefficient (250 ) 0.5 0.62 1.0 0.57 This experiment was repeated and the s me result was obtained. For the sulfonic acid resin a logical result in accor dance ~Tith the activity coefficients was produced as shown in Table XIII and Figs. 26 and 27. The 1M NH4C1 solution ex hib~ted a higher e1 tin strength than the 0.5 solution. 5. Elution with Sodium Chloride: A. Eluting Strength of 1:aC1: As shown in Ta Ie VIII and Figs. 16 and 17, the 0.5M aCl solution has a much lower eluting power than the ;0.5 4C1 sol tion. Only 81 per cent of the total cobalt was eluted after eleven liters of the e1 triant had been passed. By passing H2S gas through a lar ge sample, a faint turbidity appeared indicating the presen ce of merely a trace of zinc. 85

The most important factor that accounts for the diffe rence in eluting strength between sodium chloride and amms nium chloride is perhaps the effective sizes of the cations.

According to Creighton, (2) it is aeen that since the lI appa-

(2) Creighton, H. J., Princinplea and Applications of Elec trochemistry, Vol. I. John illey & Sona, 1943, p.l40. rent ra.dius" of sodium ions is 7.90 A" while that of ammonium .. ions is 5.37 A, the former will not be adsorbed as easily as will be the latter. In other words, ammonium ions, owing to their smaller effective size"have a higher displacing power than sodium ions.

B. Effect of Flow Rate: For obvious reasons, the slower the flow rate, the longer the contact time between the elu triant and the resin,. and hence the more the exchange. By comparing Table VIII and Table IX,. it is noted that at a flow rate of 0.43 ml/cm2/rnin., 80 per cent of the cobalt was elu ted at the eluate volume of 10.5 liters while by increasing the rate to 2.0 ml/cm2/min., only 75 per cent was eluted at the same eluate volume. This is only a qualitative illustra tion since the rtm at the hj. er rate was performed after re classification. The bed after reclassification was loosened up to expose more exchanging surface. The effect of the hi gher flow rate in reducing the eluting rate '>las thus dimlni- shed. 86

C. Effect of The Elutriant pH: While the previous elu ate concentration-eluate volume plots all show a rapid drop soon after the peak of the curve is passed, the shape of the curve in Fig. 20 is different. The right-side portion prac tically levels off, indica.ting a very slow eluting rate. This phenomenon is also characteristic of the weak acid resin. As mentioned before, Amberlite IRC-50 has a very high affinity for hydrogen ions. For any acidic solution, part of its elu ting pOI'ler is contributed by the hydrogen 10ns present. Con sequently, when the hydrogen-ion concentration is decreased by raising the pH, the eluting strength of the elutriant is reduced, giving a lower eluting rate.

The dip in the curve of Fi • 20 has no sl nificance 0 ther than to ShOYT that there wa.s some leakage during the opera.tion. D. Effect of Elutriant Concentration: For Amberlite I C-50, the 1M NaCl solution showed a higher eluting power than the 0.5M solution, as expected. Since 23 per cent of the zinc was eluted together with all the cobalt after 11.57 liters of the elutrlant had been used, the gain in increased eluting rate is offset bJ the 10SB in separation. 87

CONCLUSIONS

The experimental resUlts show that coba.ltous ionc can be separated from zinc sulfate solutions by means of the carboxylic acid cation exchange resin,. Amberlite IRC-50. The separation is based upon a higher affinity of the resin for zinc ions than for cobaltoue ione, and can be achieved in two cyclee. In theexhauetion cycle, partial separat10n ie obtained with the breakthrough point of cobalt appearing in the effluent ahead of that of zinc. In the elution cycle, either partial or complete separation is feasible, depending on the type of elutriant as well as on the operating condi tions such as the concentration and pH of the elutriant. The sulfonic a.cid cation exchange resin, Amberlite IR 120, does not' exhibit appreciable selectivity between cobalt and zinc and, consequently, fails to give the desired sepa~ ration. A tentative explanation for thls lec of selectivity is that the sulfonic acid resin may form equally stl9.ble com plexes with both of the cations. The extent of the s9paratlon made by using Amber11te IRC-50 depends on the following factors: 1. Type of Elutriant: Ammonlum chloride and sodium chlo ride are shown to be suitable elutriants. The el ting stren gth of ammonium chloride 1s higher than that.of sodium chlo ride. Neither 1nor snic acids nor bases can be used for elu tion. The former possess too great a displacing power to permit appreciable separation while the latter cause preci- 88 pitation of the hydroxides of cobalt and zinc. 2. Concentration of The Elutriant: The effect of elu triant concentration upon the separation seems to depend on the nature of the elutriant. Thus, while an 1M solution of NaC1 showed an increased elutin rate with a poorer separa tion over the O.5M solution, the reverse was true for NH4C1 in the case of Amberlite me-50. 3. pH of The Elutriant: A pH value of 5 appears to be optimum. At higher pH's, separation is improved, however, at the expense of the el tin rate. On the other hand, a more acidic solution would rapidly desorb both zinc and cobalt. Although cobalt can be separa.ted from zinc sulfate so lutions as described above, the carboxylic acid resin, Am berlite IRC-50 does not offer a practical means of purifying zinc electrolytic solutions since, zinc, instead of. cobalt, is preferentially adsorbed by the resin. 89

SUMMARY

The possibility o~ separating cobalt from zinc sulfate solutions by means of synthetic cation exchange resins was studied in this research. Sulfate solutions of cobalt and zinc, both of about 0.005N in concentration, were treated on Amberlite IRC-50 and Amberlite IR-120. It was fo nd that the carboxylic acid resin (IRC-50) s owed a considerable higher adsorption affinity for zinc ions than for cobaltous ions, thus permitting a partial separation even in the exhaustion cycle. The sulfonic acid resin (IR-120) did not exhibit ap preciable selectivity between these two cations. Elution by cationic displacement was then studied. Am berlite IRe-50 was used in most of the experiments. It was shown that ammonium chloride and sodium chloride-were capa ble to effect either partial or complete separation of co balt from zinc when both had been adsorbed by the carboxylic acid resin. Inorganic acids and bases failed to accomplish the purpose. The affects of elutriant concentration, elutriant pH, and flO'\'1 rate were separately st died by using sodium chlo ride. The latter was selected because it is the cheapest of the rea ants employed. The res Its showed that complete se paration could e 0 tained y sing a 0.5 aCI solution at a pH of 5 and a flow rate of 2 ,ml/cm2/min. Greater elut~ns rate could e achieved y increasin t e elutriant concen tration and lowering, the elutriant pH, bnt the separation would be sacrificed. 90

BIB!.. IOGRAPHY

1. AusteI'wei11, G., and Jeanproat,. C., Process foI'_ The Pre paration, Separation, And Purification of Salta, Salt so lutions And Other Solutions. U. S. Patent 1,978,447 (Oct. 30, 1934) 2. Bauman, W. C., Ion Exchange, ~heory And Practice, Aca demic Press; 1949, p.45 3. Beaton, R. H., and Furnas, C. C., Concentration of Dilute Solutions of !lectrolytes by se-exchange Materials. Ind. En . Chem.,~ Vol.33, pp.1500-15l3 (1941) 4. Boyd, G. E.,. Sch bert, J., and Adamson,. A. W., The Ex change Adsorption of Iona from Aqueo a Solutions by or anic Zeolites. J. Am. Chem. Soc., Vol.69, pp.2818-2829 1947) 5. Ets Phillips"and Pain, French Patent 808,997 (Feb. 19, 1937) 6. Grind1ey,~J., Treatment And Disposal of Taste Waters Con taining Chro.mates. J. Soc. Chem. Ind.,.Vol.64, pp.339 344 (194;) 7. Hodman, C. D."Hankbook of Chemistry And Physics. Chemi cal Rubber Pu lishin Co., (1949) 8. Hunter, M. J.,. Process And Agents for The ecovery of Magnesium Ions.from Brines. U. S. Patent 2,409,861 (Oct. 22, 1946) 9. Jenny;;H., Ionic-Exchange in Colloidal Aluminum Silica tes. J. Phy. Chern., Vol.36, p.2249 (1932) 10. Kozak, R., and Walton, H. F., Separation of Metal Ions by Cation Exehan ers. J. Phy. Chem.,.Vol.49, pp.471-472 (1945) 11. Kunin,. R. Ion Exchange. Anal. Chern., Vol.2l, p.8? (1949)

12. Latimer, I. M., The ~idation States of The Elements. Prentice-Ha.ll Inc., 1938~ p.325 13. Low, A. H.,.- reinig,. A. J., and Achoder, W. P., Technica.l Methods of Ore Analysis. John Wiley ~ Sons. 1927, p.275 14. Myers, F. J., Report PB40802, pp.21-22, Office of Tech nical Service, Was ington (l946) 15. Nachod, F. C., and Wood, W., The Reaction Velocity of Ion 91

Exchan e. J. Am. Chern. Soc., Vol.67, pp.629-631 (1945) 16. Russell, E., Swartout, J. A.p,Huma, D. N., and Kettal1e, B. H., Project. ork (May,.1944) 17. Schubert, J., Ion Exchange, ~neory And Practice. Academ ic Press, 1949, p.1B2 18. Scott, W. W., The Standard Chemical Analysis, 5th ad., D. Van Nostrand Co., 1938, p.1060 19. Snell,.F. D., and Snell, C. T., Colorimetric_Methods of AnalY6i6,~ Vol. I., D. Van Nostrand Co., 1936, p.328 20. Spaddlng, F. H., Voigt, A. F., Gladrow, E. ., and Sleight, N. R., The Separation of Rare Earths by Ion ex chan e. I. Cerium And yttrium, • Am. Chern. Soc., Vol.69 pp.2777-278l (1947) 21. Spedding, F. H., Voigt, A. F., Gladrow, E. M., Slei t, N. R.,.Powell, J. E.,.Wri .t, • M., Butler, T. A., and Fi erd, P .', The Separation of Rare Earths by Ion Exchan 6e. II. Nd-Pr., J. Am. Chern. Soc., Vol.69, pp.2786-2792 (1947) 22. Sussman, .S., Ion Exchange, Theory And Practice, Academic Press, 1949"pp.235-258 23. Sussman, S., Nachod, F. C., and Wood,W., Metal Recovery by Anion Exchange, Ind. En.g. Chem., Vol.37, pp.61B-624 (1945) 24. Tompkins, E. R., }{hym, J. X., and Oohn,.W. E., Ion Ex change As A Separation ethod. I. The Separation of Fi ssion-produced Radioisotopes, Including Individual Rare Earths, by Comp1exing Elution from Amberlite Resin. J. Am. Chern. Soc., Vol.69, pp.2769-2777 (1947)

2~. Walton, H. F., Ion Exchan e, Theory And Practice, Aca demic Press, 1949, pp.3-26 92

VITA

Wen-Hsiang Chang was born on ~~y 17, 1922 at Hunan, China. He ",as gradua.ted in 1946 with a B. S. degree in Metallurgical Engineering from the National Hunan Uni versity of China. He came to United states in Februa~J, 1949 and transferred to this School in September, 1949 from Colorado School of Mines.