Non-aqueous solvents in anion-exchange separations

T&ma.

1961, Vol

8. PP

143 to 162

NON-AQUEOUS

Peqamon

Press Ltd

Prrnted

in Northern

Ireland

SOLVENTS IN ANION-EXCHANGE SEPARATIONS*

JAMES S. FRITZ and DONALD J. PIETRZYK Instttute for Atomic Research and Department of Chemistry Iowa State University, Ames, Iowa, U S A (Received 18 August 1960 Accepted 13 October 1960) Smnmary-Dtstrtbution coefficients are measured for the partttton of metal tons between amonexchange resm and orgamc solvent-water mtxtures contammg hydrochlortc actd The presence of an In many organic solvent causes metal tons to be taken up at lower hydrochlortc acid concentrattons cases, dtstribution coeffictents are sngmficantly higher than m water-hydrochloric acid systems If other condtttons are comparable, the order of dtstrtbutton coeffictents m alcohol-water-hydrochloric acid IS- isopropyl > ethyl > methyl alcohol Column separattons of metal ion mtxtures can be carried out by elutmg with alcohol-water-hydrochlortc acid mtxtures of different compostttons Successful separations of a number of mixtures are reported INTRODUCTION THE separation

of metals as chloro complexes from aqueous hydrochloric acid solutions usmg amon-exchange columns is now a well-estabhshed and valuable analytrcal procedure .l-* In tins method, metals that form amomc chloro complexes are taken up from strong hydrochloric acid solutions; other metals pass through the column. Then the metals on the column are eluted one or two at a time using aqueous solutions of hydrochloric acid that are progressrvely more dilute. It was observed that metals are taken up strongly and at lower hydrochlorrc acid concentratrons if an appreciable amount of a water-miscrble organic solvent is added to the aqueous hydrochloric acid. The purpose of the present work was to study the anion-exchange behavlour of metal chloro complexes in partly non-aqueous medra. The primary arm was to find conditions that would improve the convenience and broaden the scope of this type of analytrcal separation. Katzm and GebertD studred the absorptron of lithium chloride, hthmm mtrate, cobalt chlorrde and nickel mtrate on chloride- and nitrate-form amon-exchange resins from acetone solutton They noted that the whole salt is adsorbed and suggested that the salts might be taken up as complex anions. On the other hand, Davies and Owen10 postulated that the salts are adsorbed by the resin by a solvent extraction process. Tuck and Welchi found that the uptake of plutoruum’v by an anion-exchange resm from a nitric acid-drethyleneglycoldrbutyl ether solution 1s the result of anionic complex formation. In the anion-exchange separation of sulphate, sulphtte, throsulphate, sulphrde, selemte and tellunte, Iguchr %13 found that the addition of alcohol to the elutmg agent increased the adsorbabrlitres of all but the selemte and tellurite. Burstall et aZ.14 eluted gold (adsorbed as the cyamde) from an anion-exchange column using an acetone eluent contaimng 5% hydrochloric acid and 5% water. Korkrsch et a1.l5 separated uranium vr from certain other metals on a chloride-form anion-exchange * Contribution No 921. Work was performed in the Ames Laboratory of the U S Atomic Energy Commlsslon 143

144

JAMESS FRITZ and DONALD J PIETRZYK

column using a solution of hydrochlonc acid in mixed water and ethyl alcohol. The addrtton of IO to 25 % methyl akohol to an aqueous hydrochloric aad eluent has been found to improve the amon-exchange separatton of zmc and cadmium 16 Recently Kojimal’J* and Yoshmo and Kurrmurals studied the anion-exchange behavrour of several of the transition elements m nnxed solvents containing hydrochloric acrd Dtstrlbutron coefficients of metal ions were measured m various combmatrons of hydrochloric acid, water and orgamc solvent, and the results were compared with the distrrbution coeffictents found in aqueous hydrochlonc acid solutrons by Kraus and Nelson v In all cases the incorporatron of a non-aqueous solvent m the rmxture caused the dlstributron coefficients to be higher. An elution scheme for several metal ions was devised, and the method was applied to the determination of ah.nnmum and copper m certain zmc alloys EXPERIMENTAL Reagents Anzon-exchange resm Dowex 1 j 8 “analysed reagent” resm, either lOO- to 200-mesh or 200- to 400-mesh The finer mesh resm gives somewhat sharper separations and was used m most of the column separations Before use, backwash the resin to remove the very fine partdes, then wash a column of resin with 2 to 2 5M perchlortc acid to remove metalhc lmpurltles Wash with water, then convert the resm to the chloride-form by successive washmg w&h 2 to 2 5h4, lM, and 0 005M hydrochlorlc acrd Wash with water, then remove the water by suction filtration Wash the resin w&h absolute alcohol, then with acetone, and allow It to air-dry The air-dried resin has a water content of 5 to IO’/, by weight EDTA, (dzsodzum dzhydrogetz ethylenedzamznetetra-acetate dzhydrate) Prepare a 0 05M stock solution from reagent-grade salt Standardlse by titration of zmc 11from primary standard zmc metal or by tttratlon of cadmium” from the primary standard, HCdV 20 Naphthyl azoxme Indicator IS recommended for the standardlsatlon 20+21 Metal sa/ts Reagent grade metal chlorrdes were used when available Rare-earth, yttrmm and scandium chlortde soluttons were prepared by dlssolvmg the oxides m hydr~hloric acid Thorturn’” and uramumvl nitrates, which were converted to chlortdes by amon-exchange, were used, and vanadmmIv was used as the sulphate In a few experiments metal perchlorates were employed The solutions were 0 OSM to 0 1M Mixtures were prepared so that the amount of Organzc solvent-water-hydrochlorzc aczd mzxtures organtc solvent and water was expressed as percentage by volume and the hydrochloric acid concentration as molarlty. For example, 1 htre of 1M hydroc~orlc acid m 80% isopropyl alcohol IS prepared as follows MIX together 800 ml of rsopropyl alcohol, 83 ml of concentrated hydrochloric acid and 117 ml of water Disregard any changes m volume due to mixmg Procedures

Macro amounts of most metal ions were determmed by tltratlon with EDTA Condttlons are summarlsed m Table I Iron**x, chrommmI1l and uranmmv’ were determmed by standard redox methods Micro amounts of metal Ions were determmed by colorlmetnc methods coppeF with neouramumV1 with arsenazo,8* bnmuth”I cuprome,** zincI with zmcon,2* thorturn” with thorma with thlourea,32 and Iron*1 with 1 IO-phenanthrohne s8 Measurement

of dzstrzbutzon coefJiczents

Weigh accurately approximately 1 g of arr-dried anion-exchange resin mto a 12%ml glassstoppered Erlenmeyer flask. Pipette mto the flask 4 ml of metal salt solution and 50 ml of the approprlate orgamc solvent-water-hydroc~orlc acid mixture Stopper the Aask and shake for 12 to 18 hr (SO% or less of orgamc solvent) or 22 to 32 IX (more than 80% organic solvent) at 24” + lo. Pipette an ahquot from the supernatant hquld, evaporate the acid and organic solvent, and determme the metal Ion content by an appropriate analytlcal method Determme accurately the water content of the resm and calculate the dlstrlbutlon coefficients on a dry weight basis

Amon-exchange

separations

145

TABLE I-TITRATIONOFMETALIONSWITHEDTA

I

Ions tttrated

PH

CdB+, Co2+, Cu*+, Nta+ ’ Zn2+ Mna+ voz+

5 5-6 5 10 5 5-6 5

Rare earth3+, Ys+ Th4+ Sc3’ BIJ+ Zr4+

5 5-6 1 7-3 3045 1 3-2 1 3-2

Naphthyl azoxme Ertochrome Black T Naphthyl azoxme (backtttrated wrth Zn2+) Arsenazo Arsenazo Methylthymol blue Thtourea Thtourea (back tttrated wtth Bi3+) Thymolphthalexone Ertochrome Black T

5 0 0 0

10 10

Ca2+ Mg2f

I Reference

Indtcator

21 22 23 24 24 22 25 26 27 22

Separation of mrxtures Prepare a sample mtxture by ptpettmg mto a beaker known amounts of standard metal salt soluttons Carefully evaporate the sample mixture nearly to dryness Add 5 ml of 0*3&f hydrochlortc acid m 97 o/0ethyl alcohol When the restdue 1s dtssolved, transfer the sample to an anion-exchange column usmg 0 3M hydrochlortc actd m 95 ‘A ethyl alcohol as a rmse solutton Use an ton-exchange column havmg an mstde diameter of either 12 mm or 22 mm, the hetght of Dowex 1 x 8 resm bed for vartous separations is mdtcated m Tables II, III, IV Elute the various sample constttuents from the column wtth the types and quantmes of eluent indicated m Tables II, III, IV with a flow rate of 0 25 to 0 33 ml/mm Evaporate the effluent fractions collected, and determine the amount of metal salt present m each effluent fractton by an approprtate tttrtmetrlc procedure MEASUREMENT

OF DISTRIBUTION

COEFFICIENTS

Measurement of distribution coefficients of various metal ions over a wide range of conditions is a good way to avoid choosmg elutmg conditions for column separations by a strictly trial and error method. The batch distribution coefficient, D, is defined: D=

meqmv. metal on resin/g of dry resin mequiv. metal m solution/ml of solution

Although this distribution coefficient is measured on a batch basis, it can be used to predict elution behaviour for metals eluted from an ion-exchange column. To separate two substances, conditions should be selected such that the distribution coefficient of one is low (preferably 1 or less) so that elution from the column will be rapid. The distribution coefficient of the other substance under the same conditions should be as large as possible so that this substance will be tightly held by the column. In aqueous hydrochloric acid solution, the distribution coefficient of most metal ions in contact with anion-exchange resin increases as the concentration of hydrochloric acid is increased. After a certain concentration of hydrochloric acid is attained, the distribution coefficient remains about the same or decreases somewhat with further mcreases m hydrochloric acid concentration. In partly non-aqueous solutions of hydrochloric acid, two factors cause an increase m the value of D; an increase m hydrochloric acid concentration, or an increase in the proportion of non-aqueous solvent in the water-non-aqueous solvent mixture. In solvent mixtures containing an appreciable amount of non-aqueous solvent, a D value sufficiently high to assure quantitative uptake of a substance on an ion-exchange column occurs at a much

146

JAMESS FRITZ and DONALD J FIETRZYK

lower hydrochlo~c acid concentration than m completely aqueous systems. Also, the maximum distnbution coefficrent obtamable in partly non-aqueous systems containing hydrochloric acid is frequently much higher than 1s possible m aqueous hydrochloric acid solutrons. The drstribution of manganeserr between chloride-form amon-exchange resin and hydrochloric acid solutron 1s an outstanding illustration of the latter. In aqueous hydrochloric aad, the dist~bution coefficient is always less than 10; m alcohol-hydrochloric acid systems containmg only a little water a distribution coefficient of several thousand can be reahsed. I I I 1000 _9op9 f ~_______--*---------95% - f -

I

I

I

I

I

I

I

I I I

FIG. I-Dlstnbutlon coefficients of coppe+ chlonde m various concentrations of isopropyl alcohol us hydrochloric acid concentrations

In Fig. 1 the drstnbution coefficrent for copper11 is plotted as a fun&on of hydrochIonc acid con~ntration at several isopropyl alcohol-water solvent composrtions. It will be noted that the presence of an increasing proportion of isopropyl alcohol has a pronounced effect on the distrrbution coefficrent. In Figs. 2-5, distribution coefficients of other metals are plotted as a functron of hydrochloric acid for solvents containmg different percentages of isopropyl alcohol. The effect of different orgamc solvents on the ion-exchange behavrour of various metal ions was studied. Fig. 6 shows that the various organic solvents studied have about the same effect on the drstribution coefficients for copper when the solvent contains an appreciable proportion of water. However, when the solvent contains less than 10 or 15 % water, the drfference in D in changing from one orgamc solvent to

Amon-exchange

FIG 2-Dtstnbution

separations

147

coefficients of metal chIondes (IS.hydrochloric acid concentration m 74 % %sopropyl alcohol.

another is more pronounced. The distribution coefficients of copper as a functron of hydrochloric acid concentration for several solvents are plotted m Fig. 7. Drstribution coefhcrents for several metal ions are plotted in isopropyl alcohol, ethyl alcohol and methyl alcohol in Figs. 8-10. In Figs. 11 and 12, distnbutron coefficients for various metals are plotted as a function of ethyl alcohol concentration for solutions that are 0.344 in hydrochloric acid. In these figures the concentration of hydrochloric acid is kept low for two reasons. When the proportion of non-aqueous solvent m the mixture 1s high, only a low concentration of hydrochloric acid 1s required for metals that form chloro complexes to have high distribution coefficients. Also, the use of much higher concentrations of

148

JAMWS. FRITZand DONALD J PIETRZYK

I

3ooc

I

I

I

I

I_HYDR&HURIC FIG 3-Dlstrlbutlon

coefficients

ACID

C&NlRATlDN

blDLAR:

of metal chlorides us hydrochloric 65 % Isopropyl alcohol

acid concentration

m

hydrochloric acid would make it lmposslble to mamtam the desired concentration of non-aqueous solvent and still use concentrated aqueous hydrochloric acid m preparing the solvent mixture. Figs. 8-10 show that the dlstrlbutlon coefficients of the metal ions studied are higher m ethyl alcohol than m methyl alcohol, and are still higher m isopropyl alcohol. By mixing the alcohols m varying proportions and keeping the hydrochloric acid concentration constant, a gradual transition of D values from one pure alcohol to another is possible Figs. 13 and 14 show the dlstrlbutlon coefficients for several metal chlorides m methyl alcohol-isopropyl alcohol and methyl alcohol-ethyl alcohol, respectively SEPARATIONS

The batch dlstnbutlon coefficient, D, can be used to calculate the volume of elutmg agent required to elute a metal Ion from a given ion-exchange column. The relatlonships used m such a calculation are as follows: D_DXgofvresm U* = V(D, + 1) where D is the batch dlstrlbutlon coefficient, D, is a volume distribution coefficient, V is the mterstltlal volume (ml) of the resin bed, and U* is the volume (ml) of eluent required to elute a metal ion to a point where the effluent from the column has the

Amon-exchange separations

FIG 4-Dlstnbutlon

149

coefficients of metal chlorides us. hydrochloric acid concentration m 55 % Isopropyl alcohol. High adsorptlon group

maximum concentration in the metal ion eluted. The volume of eluent required to elute all of a metal ion depends on the band width of the metal ion eluted. A good discussion of the selection of column dimensions and operating conditions, when the distribution coefficients of the substances to be separated are known, is given by Comish.= In the present work, batch distribution coefficients served as a valuable guide in selecting conditions for column separations. However, the volume of solution requned to elute various ions from a column was always determined by collecting actual fractions from a column and analysing the fractions. Under the conditions used, this

JAMESS. FRITZ and DONALDJ PIETRZYK

2

I

3

HYDROCHLORIC ACID CONCENTRATION FIG

5-Dlstnbutlon

I IT m IJZ

4 MLAR)

coefficients of metal chlorides us. hydrochloric acid concentration m 55 % Isopropyl alcohol. Low adsorptlon group.

ETHYL ALCOHOL WPIWPYL ALCOHOL ACETONE DIOXANE

_I

1; lO_

P-

z6 I: 06

I 20 40 PER GENT ORWW

FIG

6-Dlstnbutlon

60

I60

!zQLvENT

coefficients of copper” chloride m IA4 hydrochloric acid us varymg percentage of orgamc solvent

was found to be much more reliable than simply calculatmg the reqmred volume of eluent from the distnbutlon coefficients. To separate two metal ions on an amon-exchange column, a solvent composltlon and hydrochloric acid concentration IS chosen such that one metal will be rapldly

I

I

ETHYL ALCOHOL

1

P

I

I

I

ACID

I

I

I

I

I

CONCENTRATION

I

I

of copper= chloride m 74 % orgamc acid concentration.

HYDROCHLORIC

coefficients

I

ETHYLENE GLYCOL

1p DIOXANE

m ACETONE

II lSOPROPYL ALCOHOL

I

I

FIG. 7-Dlstnbutlon

I

100 -

300

solvent

WOIAR)

1

I

2

I

I

us hydrochlortc

1

I

FIG 8-Dlstnbutlon Isopropyl alcohol

coefficients of some metal Ions m 96 % DS hydrochloric acid concantratlon.

HvDRocmancmcoNGENTRATK)N~)

-- --___

152

JAMFSSFRITZ

and DONALDJ

PIETRZYK

eluted from the column (D should be low, preferably 1 or below) while the other metal ion has a lngh D and is tightly held by the column. When several metal ions are to be separated from each other, the scheme is to choose different conditions for successtve eluttons so that only one metal ion is eluted by each elutmg agent. In selecting the composmon of elutmg agent, we have three variables to work with the non-aqueous solvent, concentration of non-aqueous solvent m the non-aqueouswater mixture, and concentration of hydrochloric acid. In alcohol-water mixtures containing a htgh proportion of alcohol, only a low concentratron of hydrochloric acid is required for metal ions to be strongly taken up by an anion-exchange column. As shown m Figs. 7-10 the distribution coefficients of most metals change little with moderate changes m hydrochloric acid concentration. However, inspection of Figs. 11 and 12 reveals that changes m the alcohol content of the solvent affect the D values of metal ions to varying degrees. Thus one elutton scheme is based on the use of a series of eluents with a constant hydrochloric acid concentration (0*3&Q, but with decreasing proportions of alcohol. Ethyl alcohol was chosen because the D values of the “non-adsorbable” metal ions such as nickel11 and calciumI are too high m isopropyl alcohol (see Fig. 8) Methyl alcohol was not chosen because D values for “adsorbable” metal ions are sigmficantly lower than m ethyl alcohol (Figs. 9 and 10) The ethyl alcohol elutlon scheme is as follows: Sample is added to column m 95 % ethyl alcohol-0.3M HCl NP IS eluted with 82 % to 95 % ethyl alcohol-0*3M HCl. Mnrr is eluted with 72% to 82% ethyl alcohol-0*3M HCl. Co11 is eluted with 72 % ethyl alcohol-0*3M HCI CurI and Fern are eluted with 40% ethyl alcohol-O 3M HCI. Znrr is eluted with 0.005M aqueous HCl. Examples of quantitative column separations using this scheme are given m Table II. In some instances the composmon of the sample was such that the elutron of one of the sample constituents could be accomplished using a lower concentratton of ethyl alcohol than stated m the preceding scheme Several separations are not possible using this elutlon scheme. UranmmvI, copperrt and nonI can be separated only as a group. NtckePI, calcmmn, vanadmmrv and chromium111 can also be separated only as a group. ManganeseI can be separated from this latter group, but the separatton is lengthy and only moderately successful. Separation of manganeseI from cobalt IS not possible At higher ethyl alcohol concentrations, tallmg of the elutton bands 1s a problem. Tallmg of nickel11 when eluted with 0*3M hydrochloric acid m 95 % ethyl alcohol IS acid-82% ethyl particularly bad. When nickel is eluted with 0 3M hydrochloric alcohol, the tailing is lessened but IS still somethmg of a problem. An attempt was made to ehmmate tailing by making the 0*3M alcoholic solution of hydrochloric actd up to O-1&4 m perchlonc acid This speeded up all eluttons and diminished tailing of the nickel somewhat, but the over-all improvement was not sufficient to recommend the use of perchlorlc acid in the eluents. Another elution scheme uses isopropyl alcohol instead of ethyl alcohol (see Figs. 2-5). In this method the isopropyl alcohol concentration is held constant at 55 %, and the concentration of hydrochloric acid is varied. The elutron scheme is as follows Sample is added to the column in 0*3M HCl-95 % ethyl alcohol

Amon-exchange TABLE II-SEPARATION MIXTURES ON DOWEX RESIN AMOUNT

METALS

AND

153

ANALYSIS OF SYNTHETIC METAL ION

1 x 8,200-TO

APPEAR

OF WASH

separations

400~MESH, CHLORIDE-FORM

IN THE ORDER

OF ELUTION

SOLUTION INCLUDED

WITH

WITH

THE

THE FIRST ELUTED

METAL

Taken,

Found,

mg

mg

NI~~-CIP (3 x 1 1 cm) NI 22 ml 95 % EtOH-0 3M HCI 10 ml 74% EtOH-0 3M HCI Cu 15 ml 40 % EtOH-0 3M HCl

7 17 7 07

7 17 7 01

NllI-Cull-Znll (3 x 1 1 cm) NI 2 ml 95 % EtOH-0 3M HCl 9 ml 74 % EtOH-0 3M HCl Cu 15 ml 40 % EtOH-0 3M HCl Zn 40 ml 0 005M aqueous HCl

7 17 7 07 7 15

7 17 7 02 7 16

MnlI-CuI1 (3 Y 1 1 cm) Mn 2 ml 95 % EtOH-0 3M HCl 11 ml 74 % EtOH-0 3M HCl Cu 15 ml 40% EtOH-0 3M HCl

7 16 7 07

7 13 7 02

NP-Mnlr-Cull (3 Y 1 1 cm) Nl 2 ml 95 % EtOH-0 3M HCI 41 ml 95 % EtOH-0 3M HCl Mn 10 ml 72% EtOH-0 3M HCl Cu 25 ml 0 005M aqueous HCl

7 17 7 16 7 07

711 7 12 7 02

NP-CO~~ (6 Nl 2 ml 13 ml Co 15 ml

x 1 1 cm) 95 % EtOH-0 3M HCl 82 % EtOH-0 3M HCl 40 % EtOH-0 3M HCI

7 17 764

7 23 7 62

NP1-CoU-Cull-Znll (6 Y 1 1 cm) NI 2 ml 95 % EtOH-0 3M HCl 12 ml 82 % EtOH-0 3M HCI Co 17 ml 72 % EtOH-0 3M HCl Cu 15 ml 40 % EtOH-0 3M HCI Zn 40 ml 0 OOSM aqueous HCl

7 17 764 7 07 7 15

7 20 7 62 704 7 20

Coll-Felll (3 x 1 1 cm) Co 2 ml 95 % EtOH-0 3M HCl 10 ml 72 % EtOH-0 3M HCl Fe 23 ml 40% EtOH-0 3M HCl

764 690

7 62 6 92

NP-Coll-Felll-Znll (6 x 1 1 cm) NI 2 ml 95 % EtOH-0 3M HCl 13 ml 82 % EtOH-0 3M HCl Co 15 ml 70% EtOH-0 3M HCl Fe 20 ml 40 % EtOH-0 3M HCl Zn 40 ml 0 005M aqueous HCl

7 17 7 60 690 6 91

721 7 63 6 90 6 82

NI1l-Felll-BP1 (3 x 1 1 cm) NI 2 ml 95 % EtOH-0 3M HCl 6 ml 72 % EtOH-0 3M HCl Fe 12 ml 40% EtOH-0 3M HCl BI 25 ml 1M aqueous H,SO,

7 17 6 90 690

7 17 6 88 6 82

Metal mixture (column dlmenslons) and elutmg agent

-

JAMESS FRITZ

154 5OooL

I

I

1

I

and

DONALD J. PIETRZYK

l

’

’

’

’

’

1

’

I,

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I-

%!m

I

I

I

I 02

WDROCHLORlC FIG g-Dlstrlbutlon

4

”

”

”

Ati

CONCENTRATION (MOLAR)

coefficients of some metal ions m 96 % ethyl alcohol hydrochloric acid concentration

us.

I

I

”

” 01

”

’

”

I

I

I

02

HYDROCHLORIC ACID CONCETRATION (MOLAR) FIG IO-Dlstnbutlon

coefficients of some metal Ions m 96 % methyl alcohol hydrochlonc acid concentration.

us.

FIG.

111

I

I

50

I/:1

I I

70

IIk4I I

I 00

I

I

I

I

I 90

chlorides m 0*3M hydrochloric High adsorption group

PER CENT ETHYL ALCOHOL

00

11-Dlstnbutlon coefficients of metal percentage of ethyl alcohol

40

II I

I

I

acid cs

I

PER CENT

ETHYL

ALCOHOL

1

I

I

coefficients of metal chlorides m 0 3M hydrochloric Low adsorptIon group. acid vs percentage of ethyl alcohol

FIG. 12-Dlstrlbutlon

..A

.

JAMESS FRITZ and DONALD J PIETRZYK

156

NG, MnII, CaII and DyIII are eluted with 3M HCl-55 ‘A isopropyl alcohol. Co11 1s eluted with 1*3M HCl-55 % Isopropyl alcohol. CuII and FelI1 are eluted with O*lM HCI-55 % isopropyl alcohol. ZnII is eluted with 04IO5M aqueous HCl, or BP 1s eluted with 1M aqueous sulphurlc acid. TABLE III -SEPARATION AND MIXTURES RESIN

ON

DOWEX

ANALYSISOF SYNTHETICMETAL ION

1 X 8, 200-

TO

4I%MESH,

CHLORIDE-FORM

METALSAPPEARINTHEORDEROFELUTIONWITHTHEAMOUNT

OF WASHSOLUTIONINCLUDED

WITHTHEFIRSTELUTEDMETAL

Metal mixtures (column dlmenslons) and elutmg agent NGFe”‘-BinI Nl 4 50 Fe 45 BI 90

Taken, v

Found, w

ml ml ml ml

(6 x 2 2 cm) 95 % EtOH-0 3M HCl 96 % MeOH- 2M HCI 0 1M HCl-55 % 2-PrOH 1M aqueous H,SO,

8 96 9 13 8 66

8 93 9 13 8 50

(6 x 4 ml 60 ml 45 ml

2 2 cm) 95 % EtOH-0 3M HCl 1 3M HCl-55 % 2-PrOH 0 1M HCl-55 % 2-PrOH

9 51 9 13

9 58 9 12

NGMnrr-Cur1 (6 x 2 2 cm) Nl 8 ml 95 % EtOH-0 3M HCl 47 ml 96 % MeOH- 2M HCI Mn 80 ml 3M HCI-55 % 2-PrOH cu 50 ml 0 1M HCl-55 % 2-PrOH

8 96 8 96 8 94

8 94 900 8 89

N~lr-Mnlr-CoIr-CuI1-Zn’T (6 Y 2 2 cm) Nl 8 ml 95 % EtOH-0 3M HCl 47 ml 96 % MeOH- 2M HCl Mn 80 ml 3M HCI-55 % 2-PrOH co 70 ml 1 3M HCl-55 % 2-PrOH cu 50 ml 0 1M HCl-55 % 2-PrOH Zn 60 ml 0 OOSM aqueous HCl

8 8 9 8 8

96 96 51 94 95

8 8 9 8 9

99 88 50 88 06

NI”-CO~~-CU~~-BI~II(6 x 2 2 cm) Nl 8 ml 95 % EtOH-0 3M HCl 47 ml 96 p/, MeOH- 2M HCI co 65 ml 1*3M HCl-55 % 2-PrOH cu 40 ml 0 1M HCl-55 % 2-PrOH BI 85 ml 1M aqueous HaSO

8 9 8 8

96 51 95 66

9 9 8 8

06 51 97 54

CoII-Fem co Fe

Dyl’l-Mnll-CuII-FeI1l-B1”I (6 Y 2 2 cm) 8 ml 95 % EtOH-0 3M HCI DY 47 ml 96 % MeOH- 2M HCl Mn 80 ml 3M HCl-55 % 2-PrOH Cu-Fe* 50 ml 0 1M HCl-55 % 2-PrOH cu Fe BI 85 ml 1 M aqueous H,SO,

9 22 8 96

9 28 8 99

8 94 9 13 8 66

9 06 9 06 8 66

Anion-exchange TABLE III

separatrons (contd)

157

Found,

Taken,

Metal mtxtures (column dtmenstons) and elutmg agent --

mG?

-.

“g

N~Ii_ThIV_MnII_CoIJ_Cu”_Fe’T’

(6 x 22cm) 8 ml 95 % EtOH-0 3M HCl NI 73 ml 48 % EtOH-48 %M OH 0 2M HCI 85 ml 96 % MeOH- 2M HCI Th 85 ml 3M HCl-55 % 2-PrOH Mn 55 ml 1 3M HCl-55 % 2-PrOH co Cu-Fe8 50 ml 0 1M HCl-55% 2-PrOH

CU Pe Carr-MS Ca

896 906 8.95 951

8 99 P-06 8 96 956

8 94 913

896 9 03

(5 x 2 ml 56 ml 80 ml

2 2 cm) 95 % EtOH-0 3M HCl 96 % MeOH2M HCI 3M HCl-55 % 2-PrOH

8 33 906

840 9 02

Ntrr-Mnrr (5 x 2 ml NI 56 ml 80 ml Mn

2 2 cm) 95 % EtOH-0 3M HCl 96 % MeOH- 2M NC1 3M HCl-55 % 2-PrOH

445 9.05

4 51 9.07

Mn

8 &-Fe eluted together with EDTA as tltrant.

and analysed

by a photometrlc

tItratio@

This scheme gives sharp separatrons and virtually ehmmates taihng of the bands. In Fig. 15 the elutron curves are compared for the separation of nickelrx and cobalta by the isopropyl alcohol and the previous ethyl alcohol elutron schemes. In the ethyl alcohol scheme, elution with 0*3M hydrochloric acrd in 82% ethyl alcohol was continued after elutron of the nickel untrl the cobalt break-through. In the isopropyl alcohol scheme, elution wrth 3M hydrochloric acid in 55% rsopropyl alcohol was discontinued shortly after elutron of the nickel was complete. At thus point the cobalt band was still 4 cm from the bottom of the column, and the estimated cobalt breakthrough wrth the 3M eluent would be around 250 ml. The isopropyl alcohol scheme does not offer a convenient separation of nickelrr or calciumrr from manganese rr. However, rt was found that an excellent separation of nickelrr and manganeseI can be achieved through the use of 0=2&f hydrochlonc acrd m 96% methyl alcohol (see Frg. 10). Although the drstnbution coefficient for manganeseI is only 27 m 0*2M hydrochlonc acid-96 % methanol (compared to a D of around 2000 m 0*3M hydrochloric acid-95 % ethyl alcohol), column experiments showed that manganese rr is still strongly retamed by a column. An elutron curve for the column separation of nickelrr and manganeserl is given m Fig. 16. The curve shows sharp bands wrth very little tailing, and an excellent separation factor. The recommended separatron method IS first to separate nickelrr, calcmmrr or dysprosmmrrr from manganeseI (and other metal ions) using 0.2M hydrochloric acrd in 96% methyl alcohol. The metal ions remaining on the column are then separated using the 55 y0 isopropyl alcohol scheme outlmed above. Fig. 17 shows elutron curves for a multi-component mixture separated by tins method, and Table III presents results

JAMESS FRITZ and DONALD 3 PIETRZYK

158

,

loot,

I

I

I

m 0

i6

6

I

I

ii&RcENrtpL

$otKk

26

I

I

I

66

76

86

Id

~~T~~~ FIG 13-Dtstnbutlon

concentration 200

1

90 0

6

I

coefhents

of metal chlorides

of methyl alcohol

I

I

m 0 2M hydrochlorz and ethyl alcohol totalhng 96 %.

I

I

I

acid us

I

l-j

-A---Ai I6

FIG 14-Dlstnbutlon concentratron

26

i%

46 METHVL?WHCL 66 76 86 CENT coefficients of metal chlorides m 0 2iU hydrochlonc aad vs. of methyl alcohol and isopropyl alcohol totalhng 96 %.

96

Anion-exchange

separations

159

for the quantitatrve separation of several mixtures. It should be noted that a column of larger diameter was used for these separations than for those using 0*3M hydrochloric acid in ethyl alcohol (Table II). The larger column diameter allows greater loading and faster ffow rates, but of course requires larger volumes of eluent for separation. The distrrbution coefficient for thoriumrv is fairly high in 0.2M or 0.3M hydrochloric acid in 95% ethyl alcohol or isopropyl alcohol. Under these conditions the dist~butlon coefficients of nickel=, calcium=, etc. are a little too high to permrt a convenient separation from thorium Iv. However, Figs. 13 and 14 show that mixing TABLE IV.-SEPARATION

AND ANALYSISOFSYNTHETTC

METALION

QUANTITIES

MIXTURES ON

(5 X 2.2 CM)

DOWJZX 1 X 8,2m-TOM-MESH,CHLORlDE-FORMRESIN

AND TYPEOFELUTINGAGENTARESIMILARTOTHOSE IN TABLE IIf

Metal mrxture

Taken,

Found,

v

“?g

CO”“Cu” co Cu

38 46 4 46

38 40 4 45

NP-Co11 Ni CO

35 46 4 82

35 48 4 82

CO Fe

38 46 4.63

38 49 4-64

MnI-CoI’ Mn co

3622 4 82

36.39 4 83

Carl-WP Ca Mn

33.34 4.51

33 36 4 60

NIII-MnlI Nl Mn

35 46 4 51

35 5.5 4 65

COILFe111

methyl alcohol with isopropyl alcohol or ethyl alcohol provides a way to lower gradually the distribution coefficients of metal eons. In 0*2M hydrochloric acid-48 % ethyl alcohol-48% methyl alcohol, a quantitative column separation of thorium” and nickel11 is obtained, although some tailing is observed in elution of the nickel. It appears that the use of mixed alcohols may also permit a fair separation of a nickeldysprosium-thorium mixture. In the mixtures analysed above, the sample constituent are present in approximately equal molar amounts. In Table IV, results are reported for the successful analysrs of mixtures where two components are present m approximately a 10 to 1 ratio. So long as the column 1s not overloaded, a high ratro of one or more sample constituents to the others does not appear to cause any difficulty.

160

JAMES S. FRITZ

I

I

I

I

DONALD J PIETRZYK

and

1

I

I

I

I

I

I

BPXETHYL 4 03M

I

52ml

IlOml

\

NI (II)

Co(II)

c

I

IO

I

20

I

30

I

1

40

I

50 60 EFFLUENT

I

I

I

70 90 90 VOLUME (ML)

I-1

I

IYL ALCom r, I I I i Co(lI) I I : I I I I I

‘, IL-i

I

72% 1 *9.3M

HCI

I

/-I

I

\

\ i

1

I

loo

‘4.

i

__C--.-C

I

120

II0

I

130

IS-Elutton curves of a N&Con separatton Sohd lure represents elutton by the varymg hydrochlorm acid-55 % tsopropyl alcohol elutmg scheme Dotted lme represents elutton by the varying percent of ethyl alcohol-O 3M hydrochlortc acrd elutmg scheme Flow rate of l/4 to l/3 ml/mm and column drmensrons of 5 Y 2 2 cm ( J, = last detectable trace of metal ion ) FIG

4-

I

I

I

I

I

I

I

I-

ALCOHOL 31-

&

Ni(lI)

n (II)

2-

I-

O-

-J I 10

I

I

I

I

I

20

30

40

60

70

EFFLUENT

VOLUME

I.+ I

90

(ML)

FIG. 16-Elutton curve of a NP-Mnn separatton Flow rate of l/4 to l/3 ml/mm and column drmensrons of 6 x 2 2 cm. ( J = last detectable trace of metal ton )

I

I4

Anion-exchange

161

separations

0 t

I IO

I 20

I

I

30

40

I

I

60

70

EFFLUENT

I

80

I

I

NO

120

I 130

I 140

I

Is0

I 160

I 170

IJ 180

VOLUME (ML)

FIG I7---Eluttton curve of a Ni~-C~t-CuIL~~ and column dtmenslons of 6 x 2 2 cm

separation. Flow rate of 114to l/3 ml/mm ( 4 = last detectable trace of metal ion)

Zusammenfassung-Die Vertellungskoeffizlenten fur Metalhonen zwlschen emem Amonenaustauscher und emer flusslgen Phase (bestehend aus emer Mlschung von Wasser und orgamschem Solvent), die Salzsaure enthalt, wurden gemessen Dte Gegenwart von Salzsaure bewlrkt, dass die Metalhonen be1 germgerer Saurekonzentmtlon adsorblent werden In vlelen Fallen smd die Ve~ellungsk~~ienten erhebhch grosser als m Systemen mit wassriger Salzsaure Be%sonst glelchen Bedmgungen 1st die Relhenfolge der Koeffizlenten m Alkohol-Wasser-Salzsaure wle folgt* Isopropanol > lithyl- > Methylalkohol Saulentrennung von Metallionenmlschungen konnen durchgefuhrt werden mlttels Elutlon durch Alkohol-Wasser-Salzsaure-Mischungen verschledener Zusammensetzung Erfolgrelche Trennung emlger Gemlsche wlrd mltgetellt R&um&Les auteurs ont mesure des coefficients de partage d’lons m~tal~ques entre une r&me &changeuse d’amons et des m&anges eau-solvant orgamque contenant de l’aclde chlorhydrjque La presence d’un solvant orgamque provoque la fixation des ions m&talhques B des concentrations d’aede chlorhydrlque plus falbles. Dam de nombreux cas, les coefficients de partage sont beaucoup plus 6lev6s que dam les systtimes eau-aade chlorhydrlque Les autres condltlons &ant comparables, l’ordre des coefficients de partage dans les melanges alcool-eau-aclde chlorhydrlque est. le suivant’ alcool lsopropyhque, alcool bthyhque, alcool mkthyllque Les skparatlons sur colonne de mClanges d’lons m&alhques peuvent dtre r&alrs&espar &utton avec des mblanges alcooI-e.au-aclde chlorhy~ique de drffkrentes compositions Des separations d’un certam nombre de melanges r&h&es avec suc& sont mentlonn&es REFERENCES 1 K. A Kraus and F. Nelson, Symposzum on Ion Exchange and Chromatography zn Analytzcal Chemzstry ASTM Special Techmcal Pubhcatlon No. 195, American Society for Testmg Materials, Ph~ladelphla, 1958, p 27-57. 2 Idem, Proceedmgs, Intern Conference on Peaceful Uses of Atomic Energy, 7, 113, 131, Urnted Natlons, 1956. a D. H. Wdkms and L E Hobbs, Analyt Chzm Acta, 1959, 20,427. 4 D H Wlllcms, zbzd 1959, 20, 271 6 Idem, zbzd, 1959,20,273.

162

JAMESS. FRITZ and DONALDJ. P~?~RZYK

a Zdem, Ibid, 1958, 18, 372. ’ Zdem, ibid., 1957, 16,449 8 D. Jentzsch, Z annfyt. Chem., 1956, 152, 134. s L. I Katzm and E Gebert, J. Amer Chem. Sot , 1953,75, 801. lo C W Davies and B D R. Owen, J Chem. Sot, 1956, 1676 I1 D G. Tuck and G. A. Welch, J Znorg. Nuclear Chem., 1959,9,302. I8 A. Iguchl, BUN Chem Sot Japan, 1958,31,600, Chem. Abs ,53,6906g. 1s Zdem, rbrd., 1958, 31, 748; Chem Abs , 53, 11085a I4 F. H Burstall, P J Forrest, N F. Kember and R. A Wells, Znd. Eng Chem , 1953, 45, 1648 I6 J. Korknch, P Antal and F Hecht, Z anaZyt. Chem , 1960, 172,400 I6 E W. Berg and J. T Truemper, Analyt Chem , 1958 30, 1827 I7 M. KoJlma, Japan Analyst, 1957, 6, 369 I8 Zdem, rbld, 1958,7, 177 I9 Y. Yoshmo and Y Kurlmura, BuIl Chem. Sot. Japan, 1957,30,563 a0J E Powell, J S Fritz and D B. James, Analyt. Chem , 1960, 32, 954 e1 J S Fritz, W J Lane and A S Bystroff, rbzd, 1957,29, 821. ra A. J. Barnard Jr, W. C. Broad and H Flaschka, The EDTA Tztratron’ Nature and Methods of End-Pomt Detectron. J. T Baker Chemical Co , Phdhpsburg, New Jersey, 1957 pBJ. S Fritz and M A Payne, unpublished results. u J S. Fritz, R. T. Ohver and D J. Pletrzyk, Analyt Chem , 1958, 30, 1111. SKJ. S. Fritz, zbid, 1954,26, 1978 se J S. Fritz and M Johnson, zbzd, 1955, 27, 1653. 27 R. Pilbl, Analyst, 1958, 83, 188 ae H Dlehl and G F. Smith, The Copper Reagents: Cuprorne, Neocuprome, Bathocuproine. The G. Frederick Smith Chemical Co., Columbus, Ohlo, 1958 ID E A Johnson and W. Jablonskl, Zzncon Hopkm and Wdhams Ltd , Chadwell Heath, Essex, England, 1959, Monograph No 37 5o E A Johnson and E J Newman, Thorin. Hopkm and Wdhams Ltd., Chadwell Heath, Essex, England, 1956, Monograph No 29. a* J. S Fritz and M. Johnson-Richard, Analyt. Chzm. Acta, 1959, 20, 164. 88F. D BneU and C T Snell, Colorzmetric Methods of Analyszs. D. Van Nostrand Company, Inc , New York, 1949, Vol II. 88H Dlehl and G. F. Smith, Qrruntrtatme Analysts. John Wiley and Sons, Inc., New York, 1952. 84F. W Comlsh, Analyst, 1958,83, 634 *s A. L. Underwood, Analyt Chem , 1953, 25, 1910.