The Separation of Rhodium from Other Platinum-Group

THE SEPARATION OF RHODIUM FROM OTHER PLATINUM-GROUP

METALS BY ION EXGHANGE

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

EDWARD STEPHEN MCKAY, B. S., M. S.

******

The Ohio State University

1956

Approved by:

I_ Aclvi sers Department of Chemi stry DEDICATION

In memory of my sister, Jean

ii AC KNOWLEDGMENT

The author wishes to express his deep appreciation and gratitude to Professor William M. MacNevin for his guidance* encouragement, friendship, and wise counsel during the course of this study.

Also, acknowledgment is made of the financial assistance pro­ vided by the Graduate School in the form of the various teaching and assisting positions that it has been my privilege to hold during my stay at The Ohio State University.

Finally, the author expresses his thanks to Mr. Harold D. McBride for his assistance with the spectrographic analysis of the platinum m e t a l s . TABLE OF CONTENTS I

Page INTRODUCTION...... 1

The Platinum-Group Metalsi Their History and Occurrence. . . . 1

Economic Importance of the Platinum Metals ...... 2

Relationship of the Platinum Metals to Other Metals in the

Periodic Table...... 4

Separation of the Platinum Metals by Gravimetric Methods . . . 6

Separation of the Platinum Metals by Ion Exchange...... 8

Objectives and Approach to the Problem ...... 1 0

The Ion Exchange Process ...... 1 1

EXPERIMENTAL . 15

Preparation of Standard Solutions of the Platinum Metals . . . 15

Qualitative and Quantitative Methods for the Platinum Metals . 16

Properties of the Ion Exchange Resins Used in This Work. . . . 17

Preparation of Ion Exchange Columns...... 18

The Behavior of Platinum and Palladium Chlorides on Ion

Exchange Resins ...... 21

Conclusions ...... 2 8

The Behavior of Iridium Chloride on Ion Exchange Resins. . . . 28

Conclusions ...... 31

Some Properties of Rhodium Chloride and its Behavior on Ion

Exchange Resins ...... 32

Conclusions ...... 36

The Elution of Rhodium Chloride from Cation Exchange Resin3 . . 37

The Separation of Rhodium Chloride from Platinum Chloride. . . 38

iv The Separation of Rhodium Chloride from Palladium Chloride • 41

The Separation of Rhodium Chloride from Iridium Chloride • • 45

The Separation of Rhodium Chloride from Platinum and

Palladium Chlorides ...... 47

The Separation of Rhodium Chloride from Palladium and

Iridium Chlorides ...... 48

The Separation of Rhodium Chloride from Platinum and

Iridium Chlorides ...... 51

The Separation of Rhodium Chloride from Platinum,

Palladium, and Iridium Chlorides...... 52

DISCUSSION...... 54

SUMMARY ...... 55

BIBLIOGRAPHY...... 57

AUTOBIOGRAPHY ...... 61 ► INTRODUCTION j

The Platinum-Group Metals: Their History and Occurrence

The metals of the platinum-group consist of the siac transitional

elements: ruthenium, osmium, platinum, palladium, rhodium, and iridium. — 5 Of these, platinum is the most abundant and makes up 2 x 10 per cent

of the earth's crust. It was known to pre-Columbian Indians of Ecuador

(26,36,44) and first recognized as a metal by Scalinger of Italy in

1557. Early discoveries of platinum were made in South America where

it was alloyed with gold. Because of the alloy's resistance to melting

by fire, it was often separated and thrown back into streams as waste.

Platinum occurs chiefly in the metallic state in granular alloys con­

taining the other platinum metals and gold, copper, nickel, iron, and

cobalt. Numerous deposits of platinum occur in the Ural Mountains of

Russia; Columbia, South America; Abyssinia; and in the Sudbury district

of Ontario, Canada.

Palladium, the second most widely used metal of the group, was

known as a metal to Brazilian miners in 1700, (21), but it was not until

1803 that Wollaston, (48), contemporary of Berzelius, separated it from

platinum and identified it as a new element.

Rhodium, another metal discovered by Wollaston in 1804 (47), repre­

sents about 1 x 10“ 6 per cent of the earth's crust. It ocours chiefly

in a gold-rhodium ore known as rhodite and is found alloyed with plati­

num and osmiridium.

1 Iridium, present in the same abundance as rhodium, was discovered in 1803 by Tennant (40,41), an English contemporary of Wollaston* It is most often found alloyed with osmium in iridosmine and siserskite.

Iridium is also alloyed with platinum and gold in platiniridium and aurosmiridium, respectively.

Osann, a professor at Dorpat, Russia, is credited with the discovery

of ruthenium in 1828, but it was not until 1845 that Claus prepared the pure metal*

Osmium was discovered simultaneously with iridium by Tennant in

1804 (40,41). It represents 5 x 10" 6 per cent of the earth's crust.

It oocurs alloyed with iridium as osmiridium in platinum sands of North

and South America and in the Ural Mountains of Russia.

Economic Importance of the Platinum Metals

The platinum-group metals are of interest because of their unique

physical and chemical properties. Table 1 illustrates the more common

properties of these elements. Because of their great stability and low

reactivity, they are termed "noble” metals.

The platinum-group metals are utilized.as pure metals, combined,

or alloyed with other metals or as coatings in the chemical and electrical

industries, in dentistry and jewelry, and for numerous miscellaneous

purposes. Their chemical applications arise from their catalytic

activity and resistance to corrosive chemical action, even at high

temperatures. Platinum oatalysts are used mainly for hydrogenation

and dehydrogenation reactions; in oxidations and reductions; as reforming

catalysts, and in many types of syntheses (17). Pure platinum and

iridium-platinum alloys are used as insoluble anodes in various elec- Table 1. Physical Properties of the Platinum Metals (20)

Property Ruthenium Rhodium Palladium Osmium Iridium Platinum

Atomic Number 44 45 46 76 77 78

Atomic Wei ght 101.7 102.91 106.7 190.2 193.1 195.23

Color G-Wf GWf S-WS B-G6 G-Wf B-W6

Density 20°C 12.2 12.44 12.02 22.48a 22.5 21.45

Melting Point °C 2400 1966 1554 2700 2454 1774

Boiling Point °C 4900 4500 3980 5500 5300 4530

Crystal Lattice HCPd FCCb FCCb HCPd FCCb FCCb

Common Oxidation 2,3,4 2,3,4 2,4 2,3,4,8 2,3,4 2,4 States

a At 18°C b Face-centered cubic c Silvery-white d Hexagonal close-packed e Blue-white f Grey-white g Blue-grey 4 troplating prooesses. Platinum-gold and platinum-rhodium alloys have been used many years for preparation of spinnerets for making rayon fiber from viscose. Fiber glass in rapidly increasing quantities is produced in a somewhat similar manner by forcing molten glass through banks of platinum nozzles, whence it emerges in fine streams that are blown into fine diameter filaments. New combinations of the platinum- group metals are constantly being sought in the glass fiber industry in order to reduce contamination of the spinnerets with base metals.

Relationship of the Platinum Metals to Other Metals in the Periodic Table

The platinum-group elements, as members of the transition series, have many properties in common with the other members of this group.

Table 2 indicates the electronic configurations of several of the transi­ tion elements as recorded by Pauling (55). Ordinarily the orbitals

in each shell of an element are filled before those of the succeeding

shell begin to fill. In such cases, of which the alkali metals and

halides are examples, the combining properties of the atoms are largely

controlled by the electrons in the outer shell. However, in the case of the transition metals, the outermost shell has acquired some electrons

before the d orbital of the preceding shell has been filled. The change

in the electron configuration in an inner shell has less effect upon the

properties of an element than a similar change in the outermost shell.

This accounts for the horizontal similarities in properties of the

transition elements. Because of such electronic configurations and a

combination of such favorable factors as small cation size, comparatively

large nuclear or ionic charges, and multiple oxidation states, the

ability to form complexes is at a maximum. 5

Table 2. The Electronio Con.figuratione of the Group VIII Elements

SHELL K L M N 0 P Atom ORBITAL Is 2s,2p 3s,Sp,3d 4s,4p,4d,4f 5s,5p,5d 6 s

Fe 22626626

Co 2262672

Hi 2262682

Ru 2 2 6 2 6 10 268 1

Rh 22626 10 268 1

Pd 2 2 6 2 6 10 2 6 10

Os 2 26 26 10 26 10 14 2 6 6 2

Ir 2 2 6 2 6 10 2 6 10 14 2 6 9

Pt 2 26 26 10 26 10 14 269 1 The platinum-group metals are characterized by a multiplicity of oxidation states which are difficult to systematize. Few compounds containing uncomplexed species are known, the only real exceptions being

oxides, sulfides, and a few halides and sulfates. Parallels may be

found among the complexes of iron, ruthenium, and osmium, of cobalt,

rhodium, and iridium, and of nickel, palladium, and platinums, but

complications are introduced by the great number of oxidation states

among the platinum metal derivatives. In general, reduction of any

oxidation 3tate to the free metal is rather readily effected.

Separation of the Platinum Metals by Gravimetric Methods

Berzelius (7,9) published a method of separation of the platinum

metals in 1828 which was based on the action of aqua regia on the metals

and the precipitation of platinum and iridium with potassium chloride.

In 1854, Claus (9) published a method which depended on the insolubility

of the chloroplatinate, chlororuthenate, and the chloroiridate in a

saturated solution of ammonium chloride, and the reducibility of the

ammonium chloroiridate to ammonium chloroiridite through the use of

hydrogen sulfide gas. In 1877, Deville and Stas (13) found that on

fusion with lead, platinum, palladium, rhodium, and copper formed alloys

with the lead, whereas iridium, ruthenium, iron, and osmium formed an

alloy in themselves which separated from the lead alloy. At the

present time the most widely used scheme for the analysis of the platinum-

group metals is that reported by Gilchrist and Wichers (19) of the U.S.

Bureau of Standards in 1935. This scheme is illustrated in the flow

sheet in Figure 1. Osmium and ruthenium, being volatile, are removed

by distillation, leaving behind the four metals platinum, palladium, Pt, Pd, Rh, Ir (As Chloro-Salts)

NaBrCh Na 2 C0 3

PcLOg, RhOg, Ir02 PtCl4

HCl h J S i PdCl2 , RI1CI 3 , IrCl4 Pts

Dimethyl ;lyoxime £>

Pd(DMG) 2 RhCl3, IrCl4 Pt°

Ii2 (S04 )s

Pd

Rh Ir(S04 )g, Ti 2 (S04 ) 5

H 2E0 4 Cupfeirron

I 2 4 ) 3 Rh (S0 Ti-Cupferrate Ir( SO^ ) z

H 2S KaBr03 Na2C03

RhgSg IrO*

H. Hr

Rh Ir

Fig. 1-Flow sheet for Gilohrist-Wichers method of separation. rhodium, and iridium. Th© mixture is oxidized with sodium bromate and, upon addition of sodium carbonate, palladium, rhodium, and iridium, are precipitated as the hydrous dioxides while platinum remains in solution. Platinum is then precipitated as the sulfide and ignited to the metal in air. The dioxides are dissolved in concentrated hydrochloric acid and dimethylglyoxime is added to precipitate the palladium. The precipitate is filtered, dried to constant weight, and the amount of palladium metal is calculated by the use of a gravimetric factor. Up to this point the separations are quantitative and the procedure is rapid and efficient. Before rhodium can be separated from iridium, the excess of dimethylglyoxime must be destroyed with sulfuric and nitric acids. Rhodium is then precipitated as the metal but only by adding titanium sulfate, a reagent which must subsequently be removed. Rho­ dium is re-precipitated in the form of a sulfide and ignited to the metal. The excess titanium sulfate is removed by precipitation with

Cupferx'on. The Cupferron, in turn, must be destroyed with nitric and sulfuric acids and only then can iridium be hydrolytically precipitated as the hydrous dioxide. The operation must be repeated at least once in order to recover any iridium which may have eo-precipitated. The iridium dioxide is then ignited to the metal. The method becomes impractical when small quantities of rhodium and iridium are present.

In 1949, MacNevin and Tuthill (31) separated rhodium from iridium electrolytically using a controlled cathode potential. The method is now widely used but requires extensive instrumentation.

Separation of the Platinum Metals by Ion Exchange

In 1953, MacNevin and Crummett (28), in the first application of ion exchange to the separation, of* the platinum metals, succeeded in

separating 95 per cent of rhodium from iridium by anion exchange but were unable to recover any of the iridium. In the same year a brief

paper by Stevenson, et al, (39) described a separation of the four

platinum metals, platinum, palladium, rhodium, and iridium, by adsorbing

the perchlorates of these metals on a cation exchange resin and then

selectively eluting them. A patent was obtained on this process in

1955. In 1955, Berg and Senn (6 ) separated small quantities of rhodium

and iridium by reacting the chloroacids with thiourea, and passing the

mixture through a cation exchange resin. Although they recovered both

metals quantitatively, the method is subject to extensive pre-treatment

of samples and requires the decomposition of the complexes and excess

of thiourea afterward. They did not demonstrate whether the n© thod

could be applied to the separation of larger quantities of the metals.

In the same year Cluett, et al, (10) used an anion exchange resin to

separate the chloroacids of rhodium and iridium. The chlorides were

treated with a 2 per cent sodium chloride solution in 0.1M hydrochloric

acid containing 5 per cent by volume of saturated bromine water. The

rhodium was eluted with the same sodium chloride-bromine mixture while

iridium was retained on the resin. An elutant containing 5M aqueous

ammonia and 1M ammonium chloride followed by either SM nitric or 6M

hydrochloric acid removed the iridium. These investigators demonstrated

the feasibility of using their method for analytical amounts of the two

metals by successfully separating a mixture consisting of approximately

50 mg. each of both metals in a total volume of 950 ml. There are,

however, three disadvantages to the methods (1 ) the preparation of the 10 sample requires a large volume of solution due to the salt content;

(2) the elution of the iridium is done in two stages, an ammonia-ammonium chloride wash and an acid wash. This amounts to approximately 1 liter of elutants for 10 mg. of iridium and requires several hours for elution.

Since much heat is generated inside the column on addition of acid

after the ammonia, there apparently is a neutral washing period with water which the authors do not mention; and (3 ) hoth fractions contain

large amounts of salt. The rhodium is freed from sodium ions by

passing over a cation exchanger. The authors do not mention subsequent

treatment of the iridium eluate which also has a great excess of

ammonium salts.

Objectives and Approach to the Problem

The separation of rhodium and iridium is the most trying step of

the entire platinum-group analysis. As has been pointed out, the existing

methods are inadequate, both in time consumed and results achieved.

Hence, a simple method of separation is desirable. A simpler method would

be one which requires no special apparatus, and the minimum of operator

time. It should be flexible and yet be capable of achieving separations

on an analytical scale. A study of various complexes of the platinum

metals and their retention by different resins could point the way

toward easier elution and the means toward other separations. A study

of conditions necessary for separations, such as temperature effects,

resin particle size, length of resin bed, and distribution coefficients

would be most helpful in predicting whether separations are possible.

This dissertation concerns itself with such a series of developments. 11

The Ion Exchange Process

Although ion exchange was discovered in 1850 by Way and Thompson

(42,45), its first laboratory application did not take place until

1906 when the German chemist Gans (16), after studying the colloidal chemistry of the aluminosilicates, realized that ion exchange could be applied to water softening. Except for this application, not much had been done with this new tool until 1935 when Adams and Holmes (l) introduced synthetic organic ion-oxchange resins. Today ion-exchange takes its place among the more common unit processes, such as distilla­ tion, extraction, and filtration.

Ion exchange is a process whereby a mixture of ions in solution is adsorbed on a solid phase, known as the ion exchange resin, and then selectively removed so as to affect a separation. Exchange resins may be of the anion or cation exchange type and are illustrated as follows!

Anion exchanger 11R 3NOH + X “n ^ ± 1 (R3 N)n X + nOH“

Cation exchanger M + hRH R q M + nH+

■where R is the non-reactive part of the exchanger.

All uses of the ion exchange method depend upon the competition of

various ions for the available exchange positions. Ion exchange may be

divided into two main divisions and one related division. The divisions

are: break-through or displacement ion exchange, ion exchange chromato­

graphy, and ion exclusion.

Break-through or displacement ion exchange is a method employed

to remove one ion completely from a solution and substitute another for

it. For example, in the softening of water the calcium and magnesium

ions are removed from water and replaced by sodium ions. Likewise, 12 in the de-ionination of water the solution is first passed through a cation exchanger which removes all cations and then over an anion exchanger which removes all anions. The result is essentially pure water.

Ion exchange chromatography is a process whereby a mixture of ions in solution, essentially anions or cations only, are adsorbed on the ion exchange resin and then selectively removed so as to effect a separation. Since the various ions differ in their affinities for the resin, they will move down the column at different rates and thus tend to separate into distinct bands. Examples of ion-exchange chromatography include the separation of the rare earths (38), the separation of the alkalies (ll), etc. In this process the separation of similar elements or familite of elements, which have similar chemical properties, are treated.

In the ion-exclusion process (14) there is no chemical exchange of ions, but rather a physical separation of ionic from non-ionic material.

The terms ionic and non-ionic are relative rather than limiting terms; such molecules as acetic acid would be considered non-ionic. Separa­ tions might include ammonium chloride from amino acids, sodium chloride from boric acid, etc.

The degree of separation of two or more ions depends upon such factors as: (1 ) the quantity and nature of the solutes, ( 2 ) the ionic strengths and flow rates of the solutions passing through the column,

(3) the nature and rate of flow of the elutriant, (4) type and mesh size of exchanger, (5) column length, and (6 ) temperatures. 13

To make a complete study of any separation when so many variables are involved is time consuming, particularly if all of these studies are made by column experiment. It is possible, however, to oaloulate the necessary conditions for any degree of separation from equilibrium experiments and application of the plate theory. The plate theory was proposed by Martin and Synge (32) in 1935 and it was primarily applied to chromatographic work. In 1947, Mayer and Tompkins (33) applied it to the separation of the rare earths. They have demonstrated that the plate theory gives a close approximation to the experimental curves,

even for short columns, provided that Henry's law is obeyed so that there is no change in the distribution coefficient as the solute con­

centration varies.

The plate theory assumes that an ion-exchange column behaves as

though it were divided into a number of segments or theoretical plates,

each of which operates on the solute in a manner similar to that of a

batch, equilibrium separation. This theory also assumes that the action

of the column as a whole can be interpreted as though each increment of

solution were acted upon by a number of batches of resin.

From equilibrium experiments one can determine the distribution

coefficients of ions and ultimately calculate the separation factor

which gives an indication of the ease of separation of the two or

more ions from one another. The distribution coefficient is the ratio

of the quantities of an ion in any two phases under consideration. If

the resin is one phase and the solution the other phase, each containing

a quantity of a particular ion, A, then the distribution coefficient

for ion A, K d (A) iss 14 Kd(A) = Concentration of solute in the resin Concentration of solute in the solution

Similarly, a distribution coefficient K^(B) can be determined for another ion, B. The ratio of these distribution coefficients, K^CAj/K^CB) is the separation factor (c< ) and relates the ease with which one ion can be separated from another. When this ratio is greater than unity a relatively simple separation is indicated. If the separation factor is less than one, the separation takes place with difficulty and requires chromatographic elution or elution by complex formation of

one or both of the ions. EXPERIMENTAL

Preparation of Standard Solutions of the Platinum Metals

The stock solutions of each of the metals, platinum, palladium, rhodium, and iridium, were in the form of the chlorides. Since all of these salts are hygroscopic, no attempt was made to prepare the standard solutions by weighing alone. Instead, each salt was made up to a definite volume and the solution was standardized gravimetri- cally as indicated below.

Platinum} Approximately 1 gram of chloroplatinic acid

(H2 PtCl0 *6H 2 O) obtained from the Coleman and Bell Company, Norwood, Ohio, was dissolved in distilled water and 3 ml. of 3N hydrochloric acid were added. The solution was then diluted to 500 ml. with water. Aliquot portions of this stock solution were standardized by reducing the chloroplatinic acid with formic acid and weighing the metal after ignition in air.

Palladium: Approximately 1 gram of palladous chloride (PdC^), obtained from the Coleman and Bell Company, was dissolved in 0.1N hydrochloric acid, heated on the hot plate to affect complete solution, cooled to room temperature and diluted to 500 ml. with 0.1N hydro­ chloric acid. Aliquot portions of this stock solution were precipitated by the addition of a solution of dimethylglyoxime. The precipitate was dried at 110°C for one hour, weighed, and the palladium content calcu­

lated by use of a gravimetric factor.

Rhodium: Approximately 1*5 grams of rhodium chloride (RhClg),

obtained from the American Platinum Works, Newark, New Jersey, was 16 I | dissolved in 0.1N hydrochloric acid and diluted to 500 ml. The solu-

\ tion was standardized by precipitating the hydrous dioxide according

to the procedure of Gilchrist and Wichers. The dioxide was reduced to

the metal with hydrogen gas by heating in a Rose crucible.

Iridium? Approximately 2 grams of iridic chloride (IRCI4 ),

obtained from the American Platinum Works, was dissolved in 200 ml.

of 0.1N hydrochloric acid in a 600-ml. beaker. The mixture was

heated on a hot plate at 80°C and chlorine gas passed through the solu­

tion for one hour. This procedure oxidizes all of the iridium to the quad­

rivalent state. The excess of chlorine was removed by gentle boiling

on a water bath. The solution was cooled to room temperature, trans­

ferred to a 500 ml. volumetric flask and diluted to volume with 0.1N

hydrochloric acid. Measured aliquots of this stock solution were

precipitated hydrolytically. The hydrous dioxides were ignited to the

metal in a Rose crucible in the presence of a stream of hydrogen gas.

Qualitative and Quantitative Methods for the Platinum Metals

For quantities greater than 1 mg., platinum was separated from

the other metals and determined according to the method of Blackmore,

et al. (8 ). Traces of platinum in the presence of rhodium were deter­

mined by spectrographic methods. The Applied Research Laboratories

1.5 meter spectrograph was used.

Palladium was determined in any mixture of other platinum metals

by precipitation with dimethylglyoxime, drying the precipitate at

1 1 0 °C and weighing.

Rhodium, in the presence of iridium, was determined spectrophoto-

metrically with stannous chloride according to the method of Ayres, 17 et al. (4). Minute quanities were determined by use of the spectro­ graph. Rhodium, if present, was first separated from platinum by the

Gilchrist scheme.

The detection of iridium in the presence of much rhodium was performed spectrographically. For quantitative analysis the procedure of Ayres and Quick (3) using a mixture of nitric, perchloric, and phosphoric acids was employed. This method, however, gave such varia­ ble results, even on duplicate runs, that in later work it was used only as a semi-quantitative determination. 'When only small amounts of rhodium were present, the (ethylenedinitrilo) tetraaoetic acid method of MacNevin and Kriege (30) was preferred. High results were found when excess rhodium was present, but the actual concentration of iridium

could be calculated accurately since the amount of rhodium present was previously determined and known.

Properties of the Ion Exchange Resins Used in This Work

Cation exchange and anion exchange resins were used. Anion exchange

resins used were Dowex-1 and Dowex-2. The cation exchange resins used were the Amberlite resins IR-IOO(H), IR-120, and IRC-50, and various

grades of Dowex-50.

Both Dowex-1 and Dowex-2 are synthetic exchange resins containing

quaternary ammonium groups as their functional groups which are

attached to a styrene-divinylbenzene copolymer matrix. Both are

strongly basic resins and have good stability towards alkalies and

acids. Since the resins are light colored, adsorption of the platinum

metals was followed visually. r

18 I Amberlite IR-IOO(H) is the analytical grade of cation exchange j resin which contains methylene sulfonic acid groups and phenolic groups as its functional groups. Amberlite IR-120 is a strongly acidic cation exchange resin of the sulfonic acid type. Amberlite IRC-50, on the other hand, is a weakly acid cation exchanger which has carboxylic acid groups for the ion active group. Amberlite IR-100 and IR-120 differ from Amberlite IRC-50 in their relative affinities for the hydrogen ion. Amberlite IR-100 and IR-120 resins are highly ionized and show a low affinity for hydrogen ion. Amberlite IRC-50 vhich is only slightly ionized, shows high affinity for hydrogen ions.

Dowex-50 resins contain nuclear sulfonic acid groups attached to a styrene-divinylbenzene matrix. They are strongly acidic and have stability to strong acids and bases, and moderate oxidizing agents.

They can be used at elevated temperatures and operate over the entire pH range. They differ in varying degrees of divinylbenzene cross- linking and particle size.

Preparation of Ion Exchange Columns

Cation exchange resins were used in the hydrogen form. The resin was placed in a beaker and covered with 8 M nitric acid. The mixture was covered with a watch glass and digested on a hot plate for several hours. The mother liquor was decanted and more 8 M nitric acid was added. The procedxire was repeated. This removed any residual iron and

oxidizable organic matter, and also separated the fines from the normal mesh of resin. The resin was washed with distilled water until neutral, air dried, and stored for use in salt mouth bottles. 19

Anion exchange resins were used in the chloride or hydroxyl form.

The treatment of these resins was similar to that of the cation ex­ changers except that in place of nitric acid, 6 M hydrochloric acid was used to complete the conversion to the chloride form while a 5 per cent solution of sodium hydroxide was used to fix the resin in the hydroxyl form.

Two types of columns were used in this work. The first consisted merely of a buret, Figure 2, the stopcock of which was replaced by a capillary tube, joined with Tygon tubing between which was a screw-type pinch clamp. This column was approximately 50 cm. in length and 1.4 cm. in inside diameter. A wad of glass wool placed at the bottom of the column kept the resin from passing through and made a convenient

^window" for observing the beginning of elution. The resin was usually placed in a beaker and slurried with water. The slurry was then poured into the column which previously had been filled with distilled water.

This procedure eliminated the possibility of any air being trapped in the resin which might later cause channeling of the elutriant. As the resin was settling, the column was vibrated with a mechanical vibrator in order to cause tight and uniform packing of the resin. After the resin was seated properly, another wad of glass wool was firmly packed on top of it and this was followed with a layer of fine sea sand approxi­ mately 2 cm. in depth. The purpose of the sarid was to act as an automa­ tic cut-off valve and thus maintain the liquid level above the resin.

This enabled the operator to give his attention elsewhere. After the packing was oomplete the column was washed with distilled water until the eluate was neutral to litmus paper. The column was then ready for use. Sand Glass wool

cm. Resin

Glass wool

Tygon tubing Clamp

Capi Mary

Fig. 2.— Ion exchange column 21

Another type of column, Figure 3, consisted of a buret which was

surrounded by a jacket, the purpose of which was to permit the circula­ tion of warm water in order to study the effect of temperature on the

elution process. This column was conveniently made out of a water

condenser. The condenser tube was cut short and oonstricted in order

to attach a capillary tube about 3 cm. in length. Another capillary

tube of the same diameter but 4 cm. in length was joined by a piece of

Tygon tubing and served as a delivery tip of a buret. A screw-type pinch

clamp between the tip and body of the column served to regulate the flow

rate of elutriant. The inlet tube of the condenser was connected by

means of Tygon tubing to a spiral of copper tubing through which water

was circulated from a cold water tap. Thespiral was heated with a

Bunsen or Fischer burner. By regulating the size of flame, the tempera­

ture of the column could be maintained fairly constant. A constant

temperature bath with immersion heaters and pumps was available for

use. However, exact temperature control was not necessary and it was

decided that this added instrumentation would only detract from the

simplicity of the experimental work.

The Behavior of Platinum and Palladium Chlorides on Ion Exchange Re sins

MacNevin and Crummett (28) have shown that the chloro salts of the

platinum metals are quantitatively adsorbed on the anion exchangers

Amberlite IR-4B, Dowex-1, and Dowex-2 but not at all by the cation

exchange resins Amberlite IR-100 and Dowex-50. Crummett (12) has

demonstrated that ruthenium and osmium were reduced to the metals on

exchange column. This phenomenon was also observed during the present 22

mm.

8 cm Sand

Glass wool

Water jacket cm. Resin

Tygon tubing j L

Spiral heater Glass wool

Tygon tubing U K - Clamp Flame

Capillary

Fig. 3 . — Ion exchange column with heating jacket 23 study, and for this reason (and the fact that osmium and ruthenium are readily separated due to their volatilities) these two metals were not investigated any farther. Kraus, Nelson, and Smith (23) have determined the distribution coefficients for platinum and palladium chlorides on

Dowex-1. They have observed that the distribution changes with pH and have concluded that platinum and palladium chlorides were not separable by hydrochloric acid on Dowex-1 Similar work was done during the course of this investigation; however, Dowex-2 was the resin employed and the distribution coefficients were determined by an equilibrium study on adsorption of platinum and palladium chlorides by this resin. A fixed amount of the prepared, air dried resin, usually 250 mg., was weighed into a 50 ml. Erlenmeyer flask and approximately 40 mg. of palladium in the form of chloropalladous acid was added to the flask. The pH was adjusted to 3.5 with sodium hydroxide and the volume diluted to

20 ml. The flask was stoppered and shaken for 12 hours. A 15-ml, portion was withdrawn through a glass wool filter. Palladium was deter­ mined in this portion by precipitation with dimethylglyoxime. The resin was then filtered and washed well with water, dried and ignited at a low temperature. Palladium was extracted from the ash by repeated heatings with aqua regia. The acid was evaporated to dryness and the residue taken up in water. Again palladium was determined by precipi­ tation with dimethylglyoxime. Similar experiments were done with

chloroplatinic acid and Dowex-2.

The following data, obtained by the procedure outlined above, give the distribution coefficients, K^, and separation factors, (K^(Pt)/

Xd(Pd)), at various pH levels for palladous chloride and chloroplatinic 24 acid when adsorbed on the anion exohanger, Dowex-2j

Metal Found in Metal Taken Liquid Phase pH K,j(Pt) K,j(Pd) Mg. Mg. Pt Pd Pt Pd

56.1 41.0 3.50 2.50 1.0 1320 1890 0.70

56. 1 41.0 2.41 0.16 3.5 870 1250 0.69

56.1 41.0 6.10 5.95 6.0 470 730 0.64

The data show that the separation factor is much less than unity and indicate that the separation of platinum and palladium chlorides on Dowex-2 could be accomplished only by ion exchange chromatography and possibly only by complexing the chlorides with a reagent which might form complexes whose affinity for the resin would be much less than that of the corresponding chloride. A column experiment, in which palladium chloride and platinic chloride were adsorbed on Dowex-2, alone and in mixtures of the two chlorides, confirmed the assumption that the chlorides were neither completely removed from the resin, nor were they completely separable. Distribution coefficients were not determined above pH 6.0 because palladium hydroxide precipitates at the higher pH values.

Many of the base metals may be removed as insoluble hydroxides or hydrated oxides from a solution in which the platinum metals exist as

soluble nitrito complexes. If the chlorosalts of platinum, rhodium, and iridium are treated with sodium nitrite and heated, there are

formed the corresponding soluble nitrito complexes (18) Na.^Pt(N02)e] >

Hag [ph-CNG^g] , and Nag [lr(H02)6j • The palladium complex, Na » 25 is stable at pH 8 but it begins to decompose by precipitating at a pH of about 10. Now if these complexes could be made and adsorbed on an anion-exchange resin, it might be possible, due to differences in adsorption, to remove them selectively. The method was not successful because gas bubbles, presumably oxides of nitrogen, due to decomposi­ tion of the sodium nitrite with acid, broke up the resin and resulted in considerable channeling.

In 1954, Kriege (25),investigating some of the reactions of the platinum metals with (ethylenedinitrilo)tetraacetic acid (Versene), found that palladium (2) and iridium (4) form stable complexes with this reagent while platinum (4) and rhodium (5) did not. According to him these complexes are more stable than the corresponding chlorides. Hence, if palladium and platinum were adsorbed as the chlorides on an anion exchange resin and a solution of Versene passed through, It might be expected that the palladium would be eluted. This procedure was attempted by adsorbing approximately 15 mg. of palladium in the form of palladous chloride, on 3 g. of Dowex-2 and eluting with a 0.5 per cent solution of Versene. The elutriant was later increased to 3 per cent in Versene but the yellow band of palladium remained stationary on the column.

Only 35 per cent of the palladium was eluted. A variation of this pro­ cedure was to add Versene to the chloropalladous acid in order to form the complex. The complex was then adsorbed on the resin. Elution with hydrochloric acid precipitated Versene in the column and caused the flow of elutriant to stop after a short period of time,

Palladous oxide (PdO) is insoluble in water, but soluble in acids and excess of bases with the formation of palladous ion, Pd+^. The 26 hydroxide (Pd(O0)2) reacts similarly (27). This observation would indicate that when palladium hydroxide is dissolved in acids, or excess

+2 of bases, the cation, Pd , is present, probably as an intermediate, before the more stable anion, Pd(OH)^, is formed. If this is so, then it should be possible to remove palladium from an anion exchange resin such as Dowex-2 with potassium or sodium hydroxide.

Palladous chloride was adsorbed on the anion exchanger Dowex-2 and washed until the washings were neutral. Elution of the palladium chloride from the resin was done with potassium hydroxide at room temperature. The following data illustrate the per cent of elution of palladium chloride alone, platinum chloride alone, and a mixture of the two chlorides.

Elution of Palladous Chloride with Potassium Hydroxide

Volume of Molarity of Metal Form of Metal Taken Elutriant Elutriant Recovered Resin Mg. Ml. Mg.

27.3 250 2.0 M 24.9 Cl

13.6 200 0.5 11.8 Cl

13.6 200 0.5 11.2 OH

Elution of Platinic Chloride with Potassium Hydroxide

Volume of Molarity of Metal Form of Metal Taken Elutriant Elutriant Recovered Re si n Mg. Ml. Mg.

37.4 200 2.0 M 3.6 Cl

37.4 200 0.5 1.4 Cl 27

Elution of a Mixture of Platinic and Palladous Chlorides •with Potassium Hydroxide

(37.4 mg. Pt and 27.3 mg. Pd)

Volume of Molarity of Metal Form of Run No. Elutriant Elutriant Recovered Resin Ml. Mg. Pt Pd 1 1st 200 0.5 M 0.00 15.0 Cl 2nd 200 10.5 5.54

2 1st 50 2.0 M 8.2 12.2 Cl 2nd 50 3.9 3.8 3rd 50 2.5 0.38 4th 50 0.0 0.0

These runs indicate that palladium is eluted partially hy a base, presumably because the palladous hydroxide which is formed is soluble in excess of the base and dissociates to yield the palladous ion,

Pd+^, and the hydroxyl ion, 0H~. Platinum, on the other hand, is eluted to a lesser extent. When a mixture of the two is adsorbed on the exchange resin, considerably more of the platinum is eluted. This indicates that the presence of more than one of the platinum metals causes effects which act either in a positive or a negative manner with respect to the removal of each ion. Similar observations were observed by Crummett and other investigators (12,43). To trace the movement of the metals through the exchanger, the resin was extruded from the column and cut up into three sections--top, center, and bottom. Each section was dried, ignited at low heat, the ash extracted with aqua regia and the palladium content determined. Of 9.6 mg. of palladium found in the resin 47 per cent was present in the top section, 31 per cent was in the center section and 22 per cent was found in the bottom section. 28

Conelusions

(1) Tha separation factor for palladium and platinum adsorbed

on the anion exchange resin, Dowex-2, is less than unity

and indicates that a complete separation of these metals

by anionic displacement with chloride ion is not possible.

(2) Elution of the nitrito complexes causes the exchange resin

to break up mechanically due to the formation of decompo­

sition products of nitrous acid.

(3) Versene does not readily displace the chloropalladous ion

on Dowex-2.

(4) Potassium hydroxide is a good elutant for palladium but

poor for platinum.

(5) In mixtures of platinum and palladium no selective elution

by potassium hydroxide takes place. Instead, an eluted

mixture in the ratio of approximately 2 parts palladium to

1 part of platinum is found.

(6) Complete separation by elution with potassium hydroxide of

a mixture of platinum and palladium chlorides has not been

achieved.

The Behavior of Iridium Chloride on Ion Exchange Resins

When iridium chloride (irCl^) is treated with hydrochloric acid,

ohloroiridic acid (H^IrC^) is formed. The acid ionizes to form hydro­

gen ion and the hexachloroiridate anion. Iridium hydroxide (ir(OH)^)

can be precipitated from this anion by heating with a base such as

sodium hydroxide or by hydrolysis with a salt such as borax or sodium carbonate. If only small amounts of hexaohloroiridic acid are present

(o.i 2£i. or less) the precipitate becomes grey and slowly changes to ml. dark blue. Continued heating dissolves the precipitate with the forma­ tion of a blue solution, possibly a colloidal form of iridium hydroxide.

Iridium in this form will be retained partially by a cation exchange resin. It can be removed by eluting with dilute hydrochloric acid.

A greater portion of iridium in this particular blue form will be adsorbed by an anion exchanger from which it cannot be removed by hydro­

chloric acid.

When greater concentrations than 0.1 mgy^il. of chloroiridio acid

are present, the precipitated iridium hydroxide does not dissolve in

• t excess base. The precipitate can be dissolved in hydrochloric acid to

give an unstable blue solution which gradually turns green. The green

species of iridium is retained completely by an anion exchanger. Once

again its elution with hydrochloric acid is nil. The changes in color

of the solutions may be due to various oxidation states or degrees of

hydration of the ions.

Tetravalent iridium can be reduced to tervalent iridium by addition

of a reducing agent, such as ferrous sulfate, hydroxylamine, or oxalic

acid. This reduced state of iridium is much less resistant to precipi­

tation by bases or basic salts, and it is possible to heat the trivalent

form for some length of time before any precipitation of iridium oxide

takes place. Once the iridous chloride is oxidized and iridic oxide

is formed it is most difficult to redissolve. Approximately one hour is

required to dissolve 5 mg. of the oxide with hot concentrated hydrochloric

acid. This procedure precludes its use for ion-exchange separations. so f , If iridic chloride is reduced to iridous chloride with hydroxylamine, ! borax added and the solution is boiled, no precipitate will form. When

this basic solution is passed through an anion exchange resin, the iridium

passes through. If this same solution is passed over a cation exchanger

the iridium is retained. It can be removed from the column with dilute

hydrochloric acid. Similarly, if iridium chloride is reduced to the

trivalent state with hydroxylamine hydrochloride, ammonium hydroxide

added, the solution boiled to expell excess ammonia and then acidified

with hydrochloric acid, a quantitative retention of the iridium on a

cation exchange resin takes place. The reduced iridium presumably forms +3 ammine complexes with ammonia of the type IrChHg^ (22) which are

retained by cation exchangers but can readily be removed with hydrochloric

acid. These complexes are different from those formed with hydroxylamine

in basic solution. The former can be decomposed only with sulfuric and

nitric acids while the latter are readily transformed to iridium chlo­

ride merely by heating with hydrochloric acid. Kraus, et al. (23),

have shown that trivalent iridium can be removed from anion exchangers

more readily than quadrivalent iridium. It was found in the present

study that quadrivalent iridium could be removed from anion exchangers

by the following methods Iridium in the quadrivalent state is adsorbed

on an anion exchanger, such as Dowex-2. Hydroxylamine is then passed

over the resin. This procedure reduces the chloride to the tervalent

state. Ammonia is then passed through the resin and allowed to remain

for a few minutes in contact with the adsorbed iridium. The ammonia

forms the cationic complex of the type mentioned previously. Hydrochloric 31 acid is then passed over the resin to remove the iridium. This proce­ dure, however, is more difficult on the exchange ooluxnn since quadriva­ lent iridium once adsorbed on the exchange resin is difficult to reduce.

The method of Cluett, et al. (10) is lengthy because of the slow reduc­ tion of iridium(4) to iridiuxn(3) . Their procedure does not depend upon a reducing agent but merely upon the action of ammonia which slowly reduces iridium and forms an ammonia complex.

Conclusions

(1) Strong bases precipitate iridic hydroxide from iridic chloride.

If the concentration of iridium chloride is less than 0.1

mg./ml., the precipitate redissolves in excess of base.

If greater than 0.1 mg./ml., the precipitate of iridic

hydroxide does not dissolve completely.

(2) The dissolved iridium hydroxide dissociates partially

into positive ions whioh are retained on cation exchange

resins.

(3 ) Quadrivalent iridium chloride precipitates more readily

than tri valent iridium chloride on the addition of strong

bases.

(4) Addition of hydroxylamine and borax to iridium chloride

produces a cation which can be converted readily to the

chloride by heating with hydrochloric acid.

(5) Addition of hydroxylamine and ammonia produces iridium ammine

complexes which are retained quantitatively on cation ex­

change resins. 32

(6) Trivalent iridium chloride is adsorbed to a small degree on

cation exchangers.

(7) Most cation exchange resins reduce some iridic chloride to

iridous chloride which may then be adsorbed and retained.

Some Properties of Rhodium Chloride and its Behavior on Ion Exchange

Resins

Claus observed that borax did not give a precipitate with cold or hot solutions of platinum chloride but palladium and rhodium chlorides precipitated almost quantitatively and iridium chloride reacted only slowly to give a precipitate similar to that observed with strong bases. This observation allows one to separate partially palladium, rhodium and iridium from platinum, and it was probably because of this that Wichers (46) in 1924, attempted a quantitative separation of rhodium from platinum by precipitation with bases. Wichers found that rhodium would not precipitate completely when treated with borax but when a freshly prepared suspension of barium carbonate was added to the mixture an almost quantitative separation could be made.

In 1929, Kraus and Umbach (24) observed that rhodium sulfate exists in two forms, a yellow ionic form and a red non-ionic form. They have shown that the yellow rhodium sulfate cori’esponds to the violet chromium sulfate Cr(HgC)g g(SO^)g, while the red rhodium sulfate is similar to the green amorphous chromium sulfate (^^(SO^)^. Meyer and Kawczyk (34), pointed out that rhodium chloride also exists in two forms, a yellow ionic form and a red non-ionic form. They postulated that the yellow rhodium chloride may correspond totally to the gray-violet chromium chloride, Cr(HgO)Q Clg, while the brown-red form may be compared to 33

■the still unknown chromium chloride CrClg(HgO)g. By conductivity measurements they concluded that the formula of the yellow rhodium

chloride is Rh(HgO)gClg. and that it ionizes to give the hydrated cation

of rhodium. The red-brown rhodium chloride remains un-ionized as

RI1CI3 ‘SllgO. They have furthei* shown that silver nitrate precipitates

silver chloride from the yellow form but does not precipitate silver

chloride from the red-brown form. They were unable to isolate the yellow

form of rhodium chloride.

One of the earliest attempts in the present investigation was to

heat to boiling a solution of the freshly prepared red rhodium chloride

and pass it over a cation exchange resin. If any yellow form of rhodium

chloride were present, as reported by Meyer and Kawczyk, it would be

adsorbed on a cation exchange resin. Elution of the resin with hydro­

chloric acid would yield a yellow solution of rhodium chloride. The

experiment was carried out in the following manner; Approximately 5

mg. of freshly prepared rhodium chloride was boiled for several minutes

and the hot red solution was passed over the cation exchange resin,

Dowex-50. The effluent was collected and again boiled for several

minutes and passed over the column for the second time. The resin was

then washed with water until the eluted solution became neutral to litmus

paper. Dilute hydrochloric acid (6 M) was passed through the column and

the effluent was collected in a beaker. The effluent was concentrated

by evaporation to a small volume on the steam plate and its rhodium

chloride content wa.s determined colorimetrically. Approximately 0.2 mg.

of rhodium was found indicating only a 4 per cent retention. This method,

therefore, was of no value in producing the cationic form of yellow

rhodium chloride. 34 \ A literature search revealed that rhodium sulfate could be prepared I by dissolving rhodium sesquioxide in sulfuric acid. Similarly, rhodium

nitrate can be prepared by dissolving the sesquioxide in nitric acid.

According to Friend (15) and Schoeller (37) the sulfates and nitrates do

not appear as single salts but are double salts of the type

RhgOrj’Rhg^O^)^ and NaNOg'RhXROgJg^xHgO. Since these salts are not

very soluble, other forms which may yield the yellow cationic rhodium

were sought.

The compound obtained when rhodium chloride is precipitated with

borax is rhodium hydroxide (Rh(OH)g) or hydrated rhodium sesquioxide

(RhgOs'-^HgO). It was believed at this stage of the investigation that,

since borax precipitates rhodium almost quantitatively to give a hydrated

oxide, it might be possible to obtain rhodium cations by dissolving the

oxide in water. This assumption proved to be correct. Five mg. of

rhodium chloride was heated in a beaker in 10 ml. of water and preci­

pitated as the sesquioxide by addition of 0.5 g. of borax. The yellow

precipitate was filtered through a porous crucible and washed well with

water. The precipitate was dissolved in approximately 5 ml. of 6 N

hydrochloric acid and the resulting yellow' solution was passed over the

cation exchange resin, Dowex-50. The column was washed with water

until the washings were neutral to litmus paper. The effluent was

tested for rhodium with hypochlorite according to the method of Ayres

and Young (5). The test was negative indicating that all of the rhodium

was in the cationic form and was quantitatively retained by the resin.

The resin was then eluted with dilute hydrochloric acid and a yellow

eluate was obtained. Upon concentration this solution turned to rose,

the color of the original rhodium chloride. 35

After some months of observation, the following facts were assembled on rhodium chloride. Rhodium chloride (RhClg) is red-brown in color.

When dissolved in hydrochloric acid it gives a rose to red colored

solution (depending on concentration) of chloro-rhodic acid (HgRhClg).

On standing a month or more it gradually turns orange and on prolonged

standing for a year or more the solution becomes almost completely yellow. When freshly prepared, chlororhodic acid is retained completely

in a narrow orange band at the top of anion exchange resins. When an aged solution of chlororhodic acid is passed over the anion type of

exchanger only a portion of the solution is retained in a narrow band

at the top of the column. The greater part passes through the resin to

give a yellow effluent. When a freshly prepared solution of rhodium

chloride is heated to boiling it turns orange in color. When the heated

solution is passed over an anion exchange resin some rhodium passes

through. This indicates that some conversion to the hydrated cation

has taken place. Solutions of rhodium chloride have different spectra

depending on their method of preparation (2).

'With concentrations of rhodium chloride less than 0.1 mg./ml. a

yellow solution is formed but no precipitate appears upon addition of

basej at higher concentrations, a yellow precipitate appears. The same

type of precipitate can be produced by the addition of sodium carbonate,

sodium hydroxide or potassium hydroxide. The only difference observed

was in the color of the precipitate; that with borax is light yellow,

those with strong bases tended to have a tint of orange.

Vihen a mineral acid such as hydrochloric, nitric, perchloric or

sulfuric, is added to the hydrated sesquioxide of rhodium, a yellow 36

solution is formed in which rhodium is present as a cation. The yellow

solution is adsorbed completely by a cation exchanger but not at all

by an anion exchanger. It can be removed from the resin by elution with any of the mineral acids mentioned above. The yellow solution of

rhodium chloride migrates slowly toward the cathode terminal when sub­

jected to an applied potential.

Another method found for producing yellow rhodium chloride is to

precipitate rhodium hydrolytically according to the method of Gilchrist

and Wichers. The hydrous rhodium oxide is dissolved in hydrochloric

acid to give a green solution in which rhodium is in the quadrivalent

state. Addition of hydroquinone or other reducing agent produces

yellow rhodium chloride in which rhodium is present as the cation.

When yellow rhodium chloride is heated a gradual conversion to the

rose f o m takes place. Even though there i s no evident color change,

some of the rhodium is converted to the anion, Rh Cl^. This can be

demonstrated with column experiments. This phenomenan may be due to

(1) catalytic action, (2) an equilibrium between the two forms, (3)

temperature effects only, or (4) concentration effects.

Conclusi ons

(1) Rhodium hydroxide is precipitated in 99fe yield from

rhodium chloride by strong bases or basic salts.

(2) Rhodium hydroxide is soluble in mineral acids with the

formation of yellow rhodium chloride in which rhodium

exists as a cation. 37

(3) Yellow rhodium chloride is retained quantitatively by

cation exchange resins and not at all by anion exchange

resins.

(4) Yellow rhodium chloride can be obtained by dissolving the

hydrated dioxide, precipitated according to the method of

Gilchrist and Wichers, in acid and reducing the aoid solu­

tion with hydroquinone or other reducing agent.

The Elution of Rhodium Chloride from Cation Exchange Resins

A study of the conditions necessary for the retention and elution of the yellow form, of rhodium chloride from various cation exchange resins was undertaken. The effects of column length, particle size of resin, temperature changes and concentrations of elutriants were consi dered.

Amberlite IR-100, IR-120 and various forms of Dowex-50 retained yellow rhodium chloride quantitatively. Amberlite IRC-50, a weakly acidic resin containing carboxylic acid functional groups, retained only about 10 per cent of the influent, rhodium chloride. Since small quantities of rhodium were used in most instances, column length was not important. Five mg. of rhodium was retained by as little as 3 cm. of resin when adsorbed individually in the absence of other platinum- group metals. Similarly mesh size had no great effect. The resins employed had particle sizes ranging from 20 to 100 mesh. Each retained rhodium chloride equally well. Flow rates of the influent were at the maximum rate the column would deliver and depended, therefore, only on the particle size and the depth of the resin bed. 38

A graph showing the effect of the elution of rhodium chloride at various temperatures is represented in Figure 4. Figure 5 illustrates the effect of concentration of the eluting agent on the elution of rhodium chloride. The resin employed for these measurements was

Dowex-50, 20-40 Mesh, H form. The depth of the bed was 22 cm.

The Separation of Rhodium Chloride from Platinum Chloride

As pointed out previously, Claus observed that borax did not give a precipitate with cold or hot solutions of platinio chloride, but rhodium chloride under the same conditions was precipitated almost quantitatively. In 1924, Wichers (46) attempted a separation of rhodium from platinum by using either borax, sodium hydroxide, or sodium carbonate.

He found that rhodium chloride is most nearly precipitated at a pH close to the neutral point but not completely by any of the three. He did find, however, that two or three precipitations with a fresh suspen­ sion of barium carbonate did bring down all of the rhodium chloride as rhodium hydroxide but did not preoipitate any of the platinum chloride.

This same principle with slight modifications was applied to the quantitative separation of rhodium from platinum by ion exchange In the following manners

Approximately 100 mg. of sodium chloride was added to 20 ml. of a solution containing platinum and rhodium chlorides. The mixture was evaporated on a steam bath to approximately 5 ml. and then cooled to room temperature. A strong solution of sodium hydroxide (30 g. in 100 ml. water) was added dropwise until the precipitation of rhodium hydroxide was evident, and one drop was then added in excess. The mixture was Concentration of effluent (mg. Rh/50 ml.) i. . eprtr dpnec o rhodium of dependency Temperature 4.— Fig. 14 12 10 8 6 4 2 O lto wt 6 hdohoi acid. hydrochloric 6M with elution . lto at Elution A. B. lto at Elution 100 oue f flet (ml.) effluent of Volume 39 0C. 30°C 5C. 65°C 200 0 0 3

Concentration of effluent i. . Efc o cnetain f ltn aet on agent eluting of concentration of Effect 5.— Fig. (mg. Rh /50 ml.) O lto o rhodium. of elution . lto wt 6 yrclrc acid hydrochloric 6M with Elution A. . lto wt 3 hdohoi acid hydrochloric 3M with Elution B. oue f flet (ml.) effluent of Volume 100 ho 200 0 0 3

41 allowed to stand for ten minutes, and 3 N hydrochloric acid was added until the pH was 3.5. As the solution became acidic the precipitated rhodium hydroxide dissolved and a yellow solution was formed. The mixture was passed over 25 cm. of the cation exchange resin, Dowex-50-X8

(50 - 100 Mesh). Fifty ml. of water was then passed through the column to complete the elution of platinum chloride. The elution of rhodium

chloride was carried out in the heated column at approximately 65° C with 200 ml. of 6 M hydrochloric acid. The eluted rhodium chloride was evaporated to drynessand the residue was examined for traces of

platinum. It was found to be spectrographically free of platinum.

The following data are representative of this separation;

Metal Taken Metal Unsorbed Metal Eluted Mg. by Column from Column Mg. Mg. Pt Eh Pt Eh Pt Rh

5.20 7.06 5.20 0 . 1 0 0 . 0 0 6.95

5.20 7.06 5.20 0 . 1 2 0 . 0 0 - 6.90

5 . 20 14.12 5.20 0.28 Trace 13.9

5.20 14.12 5.20 0 . 1 0 0 . 0 0 14.1

le Separation of Rhodium Chloride from Palladium Chloride

Method 1.

This separation was carried out in exactly the same manner as the

separation of rhodium chloride from platinum chloride. After elution

of the rhodium chloride from the resin, the effluent was evaporated to

dryness to remove hydrochloric acid. The residue was then diluted to

approximately 150 ml. with water and 3 ml. of 6 M hydrochloric acid

were added. Approximately 5 ml. of a 1 per cent solution of dimethyl- 42 glyoxime was added and the mixture allowed to stand overnight. Any palladous chloride present in this solution precipitated as the dimethyl- glyoxime salt. The following data are typical of the results obtained by this method of separation;

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg• Mg. Pd Rh Pd Rh Pd Rh

8.32 14. 12 8.17 0.27 Trace * 13.9

8,32 7.06 8.32 0 . 1 0 0 . 0 0 6.90

8.32 7.06 8.30 0 , 1 0 0 . 0 0 6.90

8.32 14.12 8.25 0 . 2 0 Trace * 14.0

*The actual amount ofpalladium was not determined here as the amount of precipitate given by dimethylglyoxime appeared negligible.

Method 2

MacNevin and Kriegev (29) have shown that palladous chloride forms a complex with (ethylenedinitrilo)tetraacetic acid (Versene). As a means of improving the separation of rhodium chloride from palladium chloride it was decided to treat palladous chloride with Versene and then add sodium hydroxide and proceed as in method 1. The following procedure and data illustrate the results;

Five ml. of a 1 per cent solution of Versene was added to 10 ml. of a solution containing palladium and rhodium chlorides. The mixture was heated to boiling and kept at the boiling point for approximately two minutes in order that the reaction with Versene would go to comple­ tion. The solution was then cooled to room temperature. A strong solution of sodium hydroxide (30 g. in 100 ml. water) was added dropwise until the precipitation of rhodium hydroxide was evident; then one drop 43

was added in excess. The mixture was allowed -bo stand for ten minutes,

and 3 H hydrochloric acid was added until the pH became 3.5. The mixture

was passed over 25 cm. of cation exchange resin, Dowex-50-X8 ( 50 - 100

Mesh). Fifty ml. of water was then passed through the column to com­

plete the elution of palladium chloride. The elution of rhodium chloride

was carried out in the heated column at approximately 65° G with 200 ml.

of 6 M hydrochloric acid. In order that tire eluates be free of excess

Versene, chlorine gas was passed through both effluents for 10 minutes.

The rhodium solution was then evaporated to dryness and diluted to

approximately 150 ml. and 3 ml. of 6 M hydrochloric acid were added.

Five ml. of a 1 per cent solution of dimethylglyoxime were added and the

mixture allowed to stand overnight. A. yellow precipitate indicated the

presence of palladous chloride.

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg. Mg.

Pd Rh Pd Rh Pd Rh

8.32 7.06 8.30 0.10 Trace * 6.90

8.32 14.12 8.20 0.15 Trace * 14.0

* The actual amount of palladium was not determined here as the amount of precipitate given by dimethylglyoxime was negligible.

Method 3

Earlier it was pointed out that hydrous rhodium dioxide can be

precipitated according to the Gilchrist Wichers scheme and upon dissolu­

tion of the precipitate with hydrochloric acid a blue-green solution

,is obtained. This solution will pass through a cation exchange resin.

If, however, the blue-green solution is reduced to the trivalent oxidation

state a yellow solution is obtained which is completely retained by a 44 cation exchanger. This solution has the same characteristics as the yellow solution obtained when rhodium hydroxide is dissolved in hydro­ chloric acid. Undoubtedly, both solutions are identical.

With the above thought in mind the following procedure was applied

to the quantitative separation of rhodium chloride from palladium chloride.

A mixture of rhodium and palladium chlorides was precipitated according to the method of Gilchrist and Wichers. The precipitate was filtered by suction through a porous crucible and washed twice with a one per oent solution of sodium chloride and once with water. The precipitate was dissolved in 2 ml. of 6 M hydrochloric acid by passing the acid dropwise through the filter crucible and washing with hot water.

The filtrate was collected in a small test tube and its volume was approximately 15 ml. after the dissolution of the precipitate. The

solution was transferred to a beaker and its pH adjusted to 2.8 by the addition of sodium hydroxide.- The reduction of rhodium from the

quadrivalent to tervalent state was carried out by adding a 1 per cent

solution of hydroquinone dropwise until no further color change was

observed. The mixture was then passed through the ion exchange resin.

The remaining procedure was exactly the same as that described previously

under method two.

It was found that the washing of the precipitate after its disso­

lution was incomplete since the recovery of rhodium amounted to only

92 per cent although it was entirely free of palladium. The volume of

acid was kept small during the dissolution of the precipitate in order 45

-fco keep "the volume of sodium hydroxide to a minimum -when the pH was

adjusted; otherwise, much salt would be formed. The sodium chloride would be retained by the column and eventually eluted with the rhodium.

The solution to the problem was to use a Goooh crucible fitted with a glass fiber mat. After filtration the paper thin mat could be

removed and broken up with a stirring rod. The minimum amount of acid

is then required to dissolve the precipitate. When this procedure was

carried out the following results were obtained*

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg. Mg. Pd Rh Pd Rh Pd Rh

8.32 7.06 8.32 0.00 0.00 7.10

8.32 14.12 8.32 0.00 0.00 14.0

The Separation of Rhodium Chloride from Iridium Chloride

As previously pointed out, the separation of rhodium from iridium

is the most difficult of the entire platinum-group analysis. Earlier

in this investigation it was found that the yellow form of rhodium

chloride was adsorbed quantitatively on cation exchange while iridium was not adsorbed at all when in the quadrivalent state. However, when

a mixture of the chlorides is treated with sodium hydroxide and then acidified to obtain the yellow form of rhodium chloride, considerable iridium is retained on the cation exchange resin. The iridium is sub­

sequently eluted with the rhodium and the separation is incomplete.

The problem was resolved in the following way* 46

A 1 per cent solution of hydroquifcone was added to a mixture of rhodium and iridium chlorides until no further change in color was observed. The hydroquinone reduces all of the iridium to the tervalent state. A strong solution of sodium hydroxide was added until precipi­ tation of rhodium hydroxide was evident and one drop was added in excess.

The solution was then acidified with 3 N hydrochloric acid until the pH was 2.8. The solution was allowed to stand 10 minutes at room tempera­ ture. Chlorine gas was then bubbled through the mixture for approximately

10 minutes, after which time the mixture was passed over a cation exchange column paoked with Dowex-50-X8 (50 - 100 Mesh) to a depth of

25 cm. After the sample had passed through the column completely, the resin was washed with 50 ml. of 10 per cent chlorine water. The effluent contains all of the iridium.

The elution of rhodium was carried out at room temperature or at approximately 65° C with 200 to 300 ml. of 6 M hydrochloric acid. The eluted rhodium is speotrographically free of iridium. The following data are representative of this separation.

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg. Mg. Rh Ir Rh Ir Rh Ir

7.06 7.14 0 . 1 0 7.14 6.90 0 . 0 0

7.06 7.14 0 . 1 0 7.14 6.95 0 . 0 0

14,12 7.14 0.30 7.14 13.8 0 . 0 0

14.12 7.14 0.28 7.14 13.9 0 . 0 0

14.12 7*14 0 . 1 0 7.14 14-.0 0 . 0 0

14.12 14.28 0.15 14.28 13.9 0 . 0 0

35.3 14.28 1.5 14.28 33.0 0 . 0 0

33.5 35.70 1 . 8 35.70 31.6 0 . 0 0 47

The Separation of Rhodium Chloride from Platinum and Palladium Chlorides

The separation of rhodium chloride from platinum and palladium chlorides was affected considerably by the pre-treatment of the sample before it was passed through the ion exchange column. Samples which were made basic with sodium hydroxide and then heated to boiling precipitated both palladium and rhodium hydroxides. After the precipitates were dissolved and passed through the column in the usual manner considerable palladium was found in the rhodium eluate. This was shown both by precipitation with dimethylglyoxime and by spectrographic analysis.

Samples in which the precipitate of rhodium chloride was allowed to age in basic solution at room temperature carried over only spectrographic traces of palladium with the eluted rhodium. As was pointed out previous­ ly, palladous hydroxide is soluble in excess of acids and forms the palladous ion, Pd , which is adsorbed on the cation exchange resin.

This accounts for its presence in the rhodium eluate. In the former treatment in which the sample was boiled, recovery of rhodium was only

92 per cent; in the latter method of treatment, recovery of rhodium was

98 to 99 per cent. The recovered rhodium chloride, although free of platinum, contained spectrographic traces of palladium.

The procedure for this separation was as follows: Approximately

1 0 0 mg. of sodium chloride was added to the solution of the three chlorides and the mixture was heated until the salt was completely dissolved. It was then cooled to room temperature and sodium hydroxide

(30 g. in 100 ml. water) was added dropwise until precipitation of rhodium hydroxide began and then 1 drop was added in excess. Occasionally the precipitation of rhodium did not take place on addition of base. 48

In these oases 3 N hydrochloric acid was added dropwise to reduce the basicity of* the solution. This caused the solution to become cloudy and this point was considered as an indication that rhodium hydroxide had precipitated. The mixture was allowed to stand for 10 minutes.

After this time 3 N hydrochloric was added dropwise until the pH was

2.8. The solution was then passed over the ion exchange column and washed with 100 ml. of distilled water. This eluate contained the palladium and platinum chlorides. The rhodium chloride was eluted in the usual manner. The following data represent typioal separations:

Metal Taken Metal TJnsorbed Metal Eluted by Column from Column Mg. Mg. Mg. Rh Pd Pt Rh Pd Pt Rh Pd Pt

14.12 8.32 5.20 0 . 1 2 8.25 5.20 14.0 Trace* 0 . 0 0

7.06 8.32 o . 20 0 . 1 0 8.15 5.20 6.9 Trace* 0 . 0 0

* A precipitate was obtained on addition of dimethyIglyoxime to the eluted rhodium chloi'ide. The amount, however, was negligible.

The Separation of Rhodium Chloride from Palladium and Iridium Chlorides

The separation of rhodium chloride from palladium and iridium

chlorides was only 95 per cent complete. The eluted rhodium was free

of iridium but contained spectrographic quantities of palladium. Three methods were employed as a means of obtaining better separations. Two methods were of equal value. The third proved un-acceptable due to the

insolubility of iridium oxide.

Method 1

The separation was carried out in the exact manner as that for the

separation of rhodium chloride from iridium chloride. The following 49 results are typical*

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg. Mg. Rh Pd Ir Rh Pd Ir Rh Pd Ir

14.12 8.32 7.14 1.3 8.25 7.14 1 2 . 8 Trace* 0 . 0 0

14. 12 8 . o2 7.14 1.5 8 . 2 2 7.14 1 2 . 6 Trace* 0 . 0 0

* A precipitate was obtained on addition of dimethylglyoxiine to the eluted rhodium chloride* The quantity, however, was negligible.

Method 2

In order that the precipitation of rhodium hydroxide be not influenced by the presence of palladous chloride it was decided to complex the latter with Versene. It was already shown that Yersene had no adverse effects on the separation of rhodium chloride from palladium chloride and in other experiments rhodium chloride had been separated from iridium chloride in its presence. By the following procedure the separation was

95 per cent complete.

A 1 per cent solution of hydroquinone was added drop-wise to a 10 ml. mixture of the three metal chlorides until no further color change was evident. Five ml. of a 1 per cent solution of the disodium salt of

Yersene were added and the mixture heated to the boiling point. The solution was then cooled to room temperature and a strong solution of sodium hydroxide (30 g. in 100 ml. water) was added dropwise until the precipitation of rhodium hydroxide was evident, and 1 drop was then added in excess. The mixtiire was allowed to stand for 10 minutes and the pH was adjusted to 3.0 by the addition of 3 N hydrochloric acid. Chlorine gas was bubbled through the solution for approximately 1 0 minutes and the solution was passed through the exchange column. The resin was washed with 50 ml. of 10 per cent chlorine water. The effluent contained 50

-the palladium and iridium chlorides* Rhodium chloride was eluted in the usual manner. Chlorine gas was passed through both eluates for several minutes in order to destroy the Versene. The following data are representative:

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg* Mg. Mg. Rh Pd Xr Rh Pd Ir Rh Pd Ir * 15.4 8.32 7.14 1 . 2 7.14 1 2 . 2 0 . 0 0 0 . 0 0 * 13.4 8.32 7.14 1. 0 7.14 12.4 0 . 0 0 0 . 0 0

♦Palladium was not determined in these effluents, but the complete absence of palladium in the eluates indicates no sorption of palladium by the resin.

Method 5

As shown in the separation of rhodium chloride from palladium

chloride, the hydrous dioxide of rhodium, as precipitated in the

Gilchrist-Wichers scheme of analysis, may be dissolved in acid and

reduced with hydroquinone to give the yellow form of rhodium chloride

which is adsorbed quantitatively by a cation exchange resin. An attempt

was made to apply this same principle to the separation of rhodium from

palladium and iridium* The method was unsuccessful due to the insolu­

bility of hydrous iridium dioxide in hydrochloric acid. Perhaps had

enough acid been used and the mixture boiled for several minutes the

dissolution of the precipitate might have been complete. It was

necessary in this separation, however, to keep the volume of the final

solution and the acid content low. It was desired to keep the volume

low in order that time be saved in passing the mixture through the

column; the acid content had to be kept at a minimum othervri.se some

rhodium chloride would be eluted with palladium and iridium. 51

The precipitate appeared to be completely dissolved but during the neutralization process with sodium hydroxide, iridium oxide was apparent­

ly re-precipitated and collected at the top of the exchange column.

This excluded the method from further consideration.

The Separation of Rhodium Chloride from Platinum and Iridium Chlorides

The separation of rhodium chloride from platinum and iridium

chlorides was only 88 per cent complete. However, the eluted rhodium

chloride contained considerable platinum and spectrographic traces of

iridium. The presence of platinum was due to the fact that some of the

platinum chloride had been reduced by hydroquinone, perhaps even down to

the metallic state. An inorganic reagent was not used because its

introduction would require the subsequent removal of its metal ion and this operation would involve an additional run through an anion exchange

column.

The separation was carried out in the same manner as the separation

of rhodium chloride from iridium chloride. A 1 per cent solution of hydroquinone was added dropwise to the sample until no further change

in color was observed. This was followed by a few drops of strong

sodium hydroxide (30 g. in 100 ml. water) to precipitate the rhodium

hydroxide. The mixture was allowed to stand for 10 minutes. After this time the pH was adjusted to 2.8 with 3 N hydrochloric acid. Chlorine

gas was bubbled through the mixture for 1 0 minutes; then the solution was passed through 25 cm. of the cation exchange resin, Dowex-50-X8

(50 - 100 Mesh). The resin was washed with 50 ml. of 10 per cent

chlorine water to complete the elution of platinum and iridium. Rhodium was eluted in the usual manner. Data for this separation are given below. 52

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg. Mg. Eh Pt Ir Rh Pt Ir Rh Pt Ir

* 14.12 5.20 7.14 1.5 12.4 0.24 Trace** * 13.4 5.20 7.14 1.4 12.0 0.31 Trace**

* Not determined in the unsorbed effluent ** By spectrographic analysis

The Separation of Rhodium Chloride from Platinum, Palladium and Iridium

Chlorides

When the four chlorides of platinum, palladium, rhodium, and iridium were combined and a separation of the rhodium attempted by ion exchange, the eluted metal contained small quantities of palladium and platinum and spectrographic traces of iridium. The amount of rhodium eluted was approximately 95 per cent. In order to improve the purity of the rhodium chloride two methods of separation were considered. The first method was similar to that used for the separation of rhodium from iridium.

Hydroquinone was added to reduce iridium; the solution was made basic and allowed to stand for 1 0 minutes. The mixture was acidified and chlorine was bubbled through for 10 minutes and the solution was passed through 19 cm. of Dowex-50-X8 (50-100 Mesh). Fifty ml. of 10 per cent chlorine water was passed through the column to complete the elution of palladium, platinum and iridium. Rhodium was then eluted in the usual manner. The data below indicate the results obtained by this method. 53

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg. Mg < Mg. Pt Pd Rh Ir Pt Pd Rh Ir Pt Pd Rh Ir

5.20 8.32 14.12 14.28 5.0 8.31 0.50 13.4 (Trace)*

* Present by spectrographic analysis ** Not determined

The second method of separation was similar to that employed in the

separation of rhodium from palladium and iridium. In addition to hydro­ quinone, 5 ml. of a 1 per cent solution of disodium versenate were added. The solution was brought to the boiling point in order to complete the formation of the palladium-versene complex. The solution was cooled to room temperature and the pH was adjusted to 2.8 with 3 N hydrochloric

acid. Chlorine gas was bubbled through the solution for 10 minutes aPter which time it was passed through the Dowex-50 exchange column. The resin was washed with 50 ml. of 10 per cent chlorine water in order to complete

the elution of platinum, palladium, and iridium. Rhodium was eluted in

the usual manner. The data for this separation are indicated below.

Metal Taken Metal Unsorbed Metal Eluted by Column from Column Mg . M g . Mg . Eh Pt Pd Ir Rh Pt Pd Tr Rh Pt Pd Ir

14.14 5.20 8.32 14.28 1.0 5.11 8.28 12.8 (Trace)

*Not determined **By spectrographio analysis DISCUSSION

The greatest difficulty encountered in the separation of rhodium chloride from iridium chloride was to prevent the precipitation of a quantity of iridium hydroxide with rhodium hydroxide. After dissolu­ tion of the precipitates with 3 N hydrochloric acid, iridium would be retained partially on the cation exchange column. Occasionally the . iridium hydroxide, once precipitated, -mould not dissolve readily except on heating. The heating, however, converts some of the yellow rhodium chloride into the red, anionic form of rhodium chloride which is not retained by a cationic exchange resin. The solution to the problem Is not to have any quadrivalent iridium present. For this reason iridium was reduced with hydroquinone to the trivalent state which is quite resistant to precipitation* Before passing the mixture tnrough the resin, iridium chloride was re-oxidized to the higher state with chlorine gas.

Originally it was believed that quantitative precipitation of the rhodivun was not necessary before conversion to the yellow cationic form.

In gravimetric methods it is necessary that the precipitate be of such a nature that It may be readily retained upon the filter and have a minimum solubility. Since no filtration takes place in these procedures it should not matter whether the precipitate is colloidal or soluble so long as all of the rhodium chloride exists as rhodium hydroxide on addition of base.

Evidently this conversion with base does not take plate instantaneously for it was found that by allowing the precipitate of h odium hydroxide to "age” for 10 minutes before dissolving in acid, the amount of rhodium passing through the resin with the iridium is less than 1 per cent. If the precipitate is not allowed to age, 8 per cent of the rhodium is not adsorbed by the resin. 54 SUMMARY

The results of this research can. be summarized as follows*

(1 ) The separation factors for platinum and palladium chlorides

on the anion exchange resin, Dowex-2, have been determined

at several pH values as an aid in the estimation of their

separation on anion exchange resins.

(2) The separation of palladium and platinum chlorides on the

anion exchange resin, Dowex-2, was not possible.

(3) Trivalent iridium forms a series of cationic ammines which are

adsorbed quantitatively on a cation exchange resin and can be

removed readily with hydrochloric acid.

(4} Quadrivalent iridium can be removed from anion exchange resins

by first reducing it to the tri valent form, adding ammonia,

and then eluting with hydrochloric acid.

(5) Rh(OH)g precipitates when bases or basic salts are added to a

solution of rhodium chloride. By dissolution of the precipi­

tate with a mineral acid a yellow solution is obtained in

which rhodium exists as a cation. This solution is adsorbed

quantitatively by cation exchange resins.

(6 ) The yellow form of rhodium chloride is obtained when hydrous

rhodium dioxide, as precipitated according to the Gilchrist-

Wichers scheme of analysis, is dissolved in acid and reduced

with hydroquinone.

(7) Rhodium can be separated spectrographically free in 98-99

per cent yield from platinum, from palladium, and from iridium.

55

\ 56

(8 ) TIThen palladium and rhodium are precipitated according to

the Gilchrist-Wichers scheme of analysis and then separa­

ted by ion exchange, quantitative separation takes place*

The same method of separation was found impractical for

the separation of rhodium from palladium and iridium due

to the insolubility of iridium dioxide.

(9) Rhodium can be separated chemically free from palladium and

iridium by ion exchange if Versene is added to complex with

palladium.

(10) It was not possible to separate rhodium spectroscopically

free from mixtures containing platinum and palladium,

platinum and iridium, and platinum, palladium and iridium. BIBLIOGRAPHY

Specifio References*

1. Adams, B. A., Holmes, E. L., J. Soc, Chem. Ind., (London) 54,

1-6T (1935)

2. Ayres, G» H* , Anal. Chem., 25, 1622 (1953)

3. Ayres, G. H., Quick, Q., Ibid., 22, 1403 (1950)

4. Ayres, G. H., Tuffley, B. L., Forrester, J, S., Ibid., 27, 1742

(1955)

5. Ayres, G. H., Young, F., Ibid., 24, 165 (1952)

6 . Berg, E., Senn, W. Jr., Ibid., 1255 (1955)

7. Berzelius, J. J., Ann. Physik (Pogg.), 15, 533 (1828)

8 . Blaokmore, A. P., Marks, M. A., Barefoot, R. R., Beamish, F. E.,

Anal. Chem., 22, 813 (1950

9. Claus, G., Beitrage zur Chemie der Platinmetalle, 55, University

of Dorpat (1855)

10. Cluett, M. L . , Berman, S. S., McBryde, W. A* E., Analyst, 8 0 , 204

(1955)

11. Cohn, W. E., Kohn, H. W . , J. Am. Chem. Soc., 70, 1986 (1948)

12. Cruinmett, W. B., Ph. D. Dissertation, The Ohio State University,

1951

13. Deville, H. Ste-C., Stas, J. S., Prooass-verbaux, Comite Interna­

tional des Poids et Mesures, 1877, Annexe Ho. II

14. Dow Chemical Co., Technical Bulletin, "Ion Exclusion", Dow Chemical

Co., Midland, Mich. (1952)

15. Friend, J. N., Editor, "A Text-Book of Inorganic Chemistry" p. 171,

Vol. IX, Part I, Charles Griffin Ltd., London, 1920

57 58

16. Gans, R., Jahrb. preuss, geol. Landesanstalt, 26, 179 (1905);

2 7 , 63 (1906)

17. Gilchrist, R., Chem. Revs., 32, 277 (1943)

18. Gilchrist, R., Anal. Chem., 25, 1617 (1953)

19. Gilchrist, R., Wichars, E., J. Am. Chem. Soc., _57, 2565 (1935)

20. Hampel, C.A., Editor, ’’Rare Metals Handbook,” Chapter 15, p. 291,

Reinhold, New York, 1954

21. Hussak, Eugen, Sitzungsber. math, naturwiss. Klasse, Akad. Wiss.,

Wien, 113, 379 (1904)

22. Jacobson, C. A., Editor, ’’Encyclopedia of Chemical Reactions,”

Vol. Ill, p. 272, Reinhold, New York, 1949

23. Kraus, K. A., Nelson, F., Smith, G. W., J. Phys. Chem.Soc., 58,

11 (1954)

24. Kraus, F., Umbach, H., Z. anorg. Chem., 180, 42 (1929)

25. Kriege, 0. H., Ph. D. Dissertation, The Ohio State University,

1954

26. Lewis, W., Phil. Trans., (2) 48, 313 (1754)

27. McAlpine, R. K., Soule, B. A., ’’Prescott and Johnson’s Qualitative

Chemical Analysis,” p. 285, D. Van Nostrand, New York, 1933

28. MacNevin, W. M., Crummett, W. B., Anal. Chem., 25, 1628 (1953)

29. MacNevin, W. M., Kriege, 0. H., J. Am. Chem. Soc., 77, 6149 (1955)

30. MacNevin, W. M., Kriege, 0. H., Anal. Chem., 28, 16 (1956)

31. MacNevin, W. M., Tuthill, S., Ibid., 21, 1052 (1949)

32. Martin, A. J. P., Synge, R. L. M., Biochem. J., 3!5, 1358 (1941)

33. Mayer, S. IV., Tompkins, E. R . , J. Am. Chem. Soc., 69, 2866 (1947)

34. Meyer, J., Kawczyk, M., Z. anorg. u. allgem. Chem., 228, 297 (1936) 59

35. Pauling, L., "The Nature of the Chemical Bond and the Structure of

Molecules and Crystals," pp. 27-29, 2nd Ed., Cornell University

Press, New York, 1940

36. Scheffer, H. T., "Das WeiBse Gold in Spanine Platina del Pinto,"

Chem. Abh. konig. schw. Akad., p. 275, 1752

37. Schoeller, W. R., Pwell, A.R., "The Analysis of Minerals and

Ores of the Rarer Elements," 2nd. Ed., p. 247, Charles Griffin,

Ltd., London, 1940

38. Spedding, F. H . , Voigt, A. F., Gladrow, E. M . , Sleight, N. R.,

J. Am. Chem. Soc., 69, 2777 (1947)

39. Stevenson, P. C., Franke, A. A., Borg, R., Narvik, W . , Ibid.,

4876 (1953)

40. Tennant, S., Phil. Trans., 94, 411 (1804)

41. Tennant, S., Nicholson’s Journal, 10, 24 (1305)

42. Thompson, H. S., J. Roy. Agr. Soc. Eng., JLjL, 6 8 (1850)

43. Tuthill, S., Ph. D. Dissertation, The Ohio State University, 1948

44. Watson, W., Phil. Trans., 46, 584 (1750)

45. Way, J. T., J. Roy. Agr. Soc Eng., 11, 313 (1350

46. Wichers, E., J. Am. Chem. Soc., 46, 1818 (1924)

47. Wollaston, W. H., Nicholson’s Journal, 1 0 , 34 (1805)

48. Ibid., 10, 204 (1805) 60

General References;

1. Cassidy, H. G., "Adsorption and Chromatography. Technique of

Organic Chemistry," Vol. V, Weissberger, A., Editor, Interscience,

New York, 1950

2. Fresenius, W., Jander, G., "Handbuch der Analytisehen Chemie,"

III, Band Vllb, Berlin, 1951

3. Kunin, R., Myers, R. J., "Ion Exchange Resins," John Wiley, New

York, 1950

4. Mellor, J. W., "A Comprehensive Treatise on Inorganic and Theoretical

Chemistry, Vol. XV and XVI, Longmans, Green, New York, 1936

5. Nachod, F. C., Editor, "Ion Exchange. Theory and Application,"

Academic Press, New York, 1949

6 . Samuelson, 0., "Ion Exchange in Analytical Chemistry," John Wiley,

New York, 1950 AUTOBIOGRAPHY

I, Edward. S. McKay, was born in Pittsburgh., Pennsylvania, September

10, 1923. I received my secondary school education in the public schools of•Pittsburgh. My undergraduate training was obtained at St. Vincent

College, Latrobe, Pennsylvania, from which I received the degree

Bachelor of Science with high honors in May, 1949.

In December, 1950, I received the degree Master of Science from

The Ohio State University, where I majored in chemistry.

From December, 1950, to March, 1953, I was a chemist with the Ohio

Department of Agriculture, Division of Plant Industry. From March, 1953, to September, 1953, I was chief chemist for the Division of Foods and

Dairies in the State of Ohio. In the latter part of September, 1953, I reentered The Ohio State University and began work toward the degree

Doctor of Philosophy. While completing the requirements for this degree,

I held the following appointmentsj graduate assistant from September,

1953, to August, 1954; University scholar from September, 1954, to June,

1955; and teaching assistant from August, 1955, to March, 1956. I was appointed an assistant instructor in the analytical division of the

Department of Chemistry at The Ohio State University for the quarter,

March, 1956, to June, 1956.

61