Chromatographic Separation of Alkaline Earth Metals Using Alpha-Hydroxyisobutyric Acid

AN ABSTRACT OF THE THESIS OF

JOHN ARTHUR HAUSCHILD for the MASTER OF SCIENCE (Name) (Degree) in CHEMISTRY (ANALYTICAL) presented on (Major) Title: CHROMATOGRAPHIC SEPARATION OF ALKALINE EARTH

METALS USING ALPHA-HYDROXYISOBUYRIC ACID Abstract approved: Redacted for Privacy Max B. Williams

A systematic study of the elution of magnesium and calcium from Dowex 50 X 8 resin using a-hydroxyisobutyric acid (a-HIBA) at various pH values and concentrations, indicated that the difference in the equilibrium distribution coefficients of these two elements was large enough for a good separation.This fact was applied to develop a chromatographic procedure for the separationof milligram quantities of magnesium, calcium, strontium, and barium.After magnesium was eluted with 0. 22M a-HIBA at pH 4. 5, thethree remaining elements were eluted by varying the concentration and pH of a-HIBAduring the course of the elution (exponential gradient elution).After its respective elution, each alkaline earth metal was directly determined by atomic absorption spectroscopy.Using this method, several successful analyses of synthetic samples (similar to the composition of sea water) were performed.Yield determinations of the alkaline earth metals from these analyses were consistently greater than 93%, with the overall average yield being 98%. Chromatographic Separation of Alkaline Earth Metals Using Alpha-Hydroxyisobutyric Acid

by John Arthur Haus child

A THESIS submitted to Oregon State University

in partial fulfillment of the requirements for the degree of Master of Science

June 1971 APPROVED:

Redacted for Privacy Professior of Chemistry in charge of major

Redacted forPrivacy Chairman of Department of Chemistry

Redacted for Privacy

Dean of Graduate School

Date thesis is presented )6,7)/q2) Typed by Mary Jo Stratton for John Arthur Hauschild ACKNOWLEDGEMENTS

There are numerous people that have made it possible for my having the opportunity to pursue the program that has led up to this dissertation.It would be impossible to express my thanks to all of these individuals but there are a few that should be singled out for their help. I would like first to thank Dr. Max B. Williams, my major professor, for having the courage and patience to continue to support me throughout this entire program.Without his help and encourage- ment I would probably have never completed this program for the M. S. degree.

Secondly,I would like to express my thanks to Captain Jack Terry, a fellow Air Force officer stationed at McClellan AFB, Sacramento, California.It was with his help that data for some of the experiments performed in this study were accumulated. Also I would like to thank Dr. Bert E. Christensen, former chairman of the chemistry department, and Dr. Stephen J. Hawkes for their assistance and guidance in my course work. Finally,I wish to express my sincere gratitude to the United States Air Force for making it possible for me to undertake this study at Oregon State University.The opinions expressed by me, however, are my own and do not reflect the policies or opinions of the United States Air Force or the Department of Defense. TABLE OF CONTENTS

Page

INTRODUCTION 1

Survey of Separational Methods 1 Purpose of This Study 11

PRINCIPLES OF ION EXCHANGE 14

Resins 14 Column Mechanics 16 Column Operation 17 Factors Affecting Separation 20 Column Parameters 23 Distribution Coefficient 23 Separation Factor 23 Volume Distribution Coefficient 24

SEPARATION FACTOR DETERMINATION 28

Objective 28 Apparatus 29 Preparation of Solutions 31 a- HIBA 31 Calcium and Magnesium Stock Solutions 31 Pretreatment of Cation Exchange Resin 32 Procedure for Peak Elution Volume Determination 32 Column Method 33 Concentration Determination 33 Results and Discussion 35

ANALYSIS OF SYNTHETIC SAMPLE 41 Preparation 41 Procedure 41 Results and Discussion 43

BIBLIOGRAPHY 56 LIST OF TABLES

Table Page

1 Treatment Procedures for Cation Resin Columns. 32

2 Operating Conditions for Calcium and Magnesium. 34

3 Effect of Eluting Agent on Peak Elution Volumes. 38

4 Analysis of Synthetic Sample. 47

5 Operating Conditions of Perkin-Elmer Model 403 Atomic Absorption Spectro- photometer. 50 LIST OF FIGURES

Figure Page

1 The separation of magnesium, calcium, strontium, barium, and radium with ammonium lactate eluent. 6

2 The separation of beryllium, magnesium, calcium, and strontium with 1. 5M hydrochloric acid eluent. 7

3 The separation of magnesium, calcium, strontium, and barium with ammonium citrate eluent. 8

4 Separation of alkaline earths with zirconium molyb date. 10

5 System with constant-volume mixing chamber for gradient elution. 30

6 Influence of alpha-hydroxyisobutyric acid concentration and the effect of pH upon the elution of magnesium. 36

7 Influence of alpha-hydroxyisobutyric acid concentration and the effect of pH upon the elution of calcium. 37

8 Separation of alkaline earths and sodium with ammonium alpha-hydroxyisobutyrate eluent. 44

9 Calibration curve for magnesium, using the Perkin-Elmer 290B atomic absorption spectrophotometer. 48

10 Calibration curve for calcium, using the Perkin-Elmer 290B atomic absorption spectrophotometer. 49

11 Calibration curve for magnesium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 51 Figure Pa e

12 Calibration curve for calcium, using the Perkin-Elmer 433 atomic absorption spectrophotometer. 52

13 Calibration curve for strontium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 53

14 Calibration curve for barium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 54 CHROMATOGRAPHIC SEPARATION OF ALKALINE EARTH METALS USING ALPHA-HYDROXYISOBUTYRIC ACID

I.INTRODUCTION

While numerous methods for the separation of trace amountsof the alkaline earths are known, none of these methods aresimultane- ously practical and quantitatively successful in theseparation of Large amounts of these elements. A survey of the literatureshows that cation-exchange methods have been used almost exclusivelyin the few papers which have reported quantitativedata (7, 22, 25, 29), and this technique appears at the present time to be the mostpromising one for separating larger quantities of the alkaline earths.

Survey of Separational Methods

In recent years many investigators havereported separating the alkaline earths from each other and from severalother elements by methods other than precipitation procedures.One method where progress has been made in theseparation of the alkaline earth metals from each other is paper chromatography.Gordon and Hewel (11) reported quantitatively separating microgram amountsof several alkali and alkaline earth cations, and believed that acomplete separation of alkaline earths and beryllium was possibleby their procedure. Another scheme, which describes the quantitativeanalysis of microgram amounts of alkaline earthmixtures, employs paper strips with 2 an eluting solvent consisting of methanol, isopropanol, formic acid, ammonium formate and water (35). Application of solvent extraction to the analysis of alkaline earths is not extensive.Few procedures are specific for these elements, and the applications have mainly been to the separation of these elements from one another.Morrison and Freiser (30) reviewed extraction procedures up to 1956.For the separation of calcium from strontium, resort must be made to the extraction of the anhydrous nitrates with such solvents as absolute alcohol and ether, acetone, nitric acid, or glacial acetic acid (3), which dissolves calcium nitrate and leaves most of the strontium and barium nitrate. This procedure has been applied particularly for the separation of a little strontium from much calcium. Extraction of calcium, barium, strontium and many other elements with organic solutions of 2-thenoyltrifluoroacetone (TTA) used in radiochemical separations has also been summarized (37). Stary and co-workers have made several systematic studies of many metals, alkaline earths included, by solvent extraction.In one study, the extraction of alkaline earths by 0. 500M oxine (8-hydroxyquinoline) in chloroform, in the absence of a complexing agent and in the presence of 0. 010M oxalic acid, was discussed (46).In another study, the extraction of alkaline earth metals by 0. 100M acetylacetone, benzoylacetone, and dibenzoylmethane solutions in benzene as a 3 function of pH was mentioned (47).These investigators were not able to extract the alkaline earths with chloroform in the presence of an excess of cupferron (0. 05M) throughout the pH region studiedin another investigation (48). The development of ion-exchange methods has been very rapid. At first ion exchangers were mostly used for water softening but they soon became widely employed in many other fields.These methods assumed great importance in chemical research, in analysis, in preparative work as well as in technology.The appearance of ion exchangers had a tremendous impact on analytical chemistry also. Their use gave analysts new methods which led to the solution of previously insoluble problems, such as separating the rare earth metals.The need for an accurate and fairly rapid procedure for the analysis of mixtures of alkaline earth metals also led to the application of ion-exchange chromatography to this problem. Trace amounts of alkaline earths have been separated from each other on anion exchange resin pretreated with 0. 05M ammonium citrate at pH 7.5, elution order being barium, strontium, and calcium (32).In another anion exchange study, Nelson, Day and Kraus (31) separated trace quantities of a number of elements using ethylenediaminetetraacetic acid (EDTA) solutions. A few other procedures utilizing anion exchange columns have also been reported for separating trace amounts of individual alkaline earth metals from other 4 contaminants. Numerous separations of alkaline earths from each other have been made with cation exchange resin, utilizing many different eluting solutions.Tompkins (51) separated radioactive strontium, barium, and radium on a column of colloidal agglomerates of Dowex-50, with 0. 5M ammonium citrate at pH 7.8, but only a nearly complete separation was achieved. A modification of this procedure, that is by eluting with 0. 32M ammonium citrate at pH 5. 6, gave a better separation of milligram quantities of barium and radium (41). A separation of magnesium from other alkaline earths, copper, nickel, and zinc was accomplished by using different concentrations of ammonium acetate (14). A similar technique using radioactive tracers of the alkaline earth metals was achieved by utilizing a gradient elution of ammonium acetate (10).Khopkar and De (16) did a very comprehensive study on the cationexchange behavior of milligram amounts of barium using ammonium acetateand several other eluants.Their data showed the following order with respect to increased eluting efficiency:tartaric acid < EDTA < citric acid < ammonium acetate < sodium chloride < ammonium chloride < hydrochloric acid < sodium nitrate < nitric acid. A similar study on the behavior of milligram amounts of calcium was completed (25). Eluting agents employed were arranged in order of increasing efficiency as:citric acid < sodium nitrate = ammonium chloride < 5 hydrochloric acid = ammonium acetate = nitric acid.Acetic acid- ammonium acetate mixtures were reported as not satisfactory in the latter study.Li lova and Prevbrazhensky (23) reported achieving a better separation of the alkaline earths at 20 C than at 80 C using ammonium acetate; but they achieved a still better separation by using ammonium lactate at 80 C. Milligram quantities of the alkaline earth metals have been quantitatively separated from each other using 1M (36) and 1. 2M (22 ) ammonium lactate.The latter procedure was later applied to the quantitative determination of the alkaline earths in glass (21).Milton and Grummitt (29) comprehensively studied the cation-exchange separation of milligram amounts of all the alkaline earth metals at elevated temperatures, and applied this to the analysis of Sr90in milk ash samples.They quantitatively eluted magnesium, calcium, strontium, barium and radium in that order with 1. 5M ammonium lactate at pH 7 and at 78 C.They also obtained good separations with 1. 5M and 4M hydrochloric acid eluants and with an eluant of 5% ammonium citrate at pH 5.Temperatures of 60 and 78 C were used for the hydrochloric acid and ammonium citrate elutions, respectively. However, ammonium lactate was confirmed to be superior to ammonium citrate and hydrochloric acid.Some of their results are reproduced in Figures 1-3. Separation of milligram amounts of magnesium and calcium using Interstitial Column Volumes 2 5 10 20 50 100 200 0, 6 0, 8 1 , . 1 I I Mg 10 Ca

Sr 0 Ra 1.0 100

_0 10 20 50 100 200 500 1000 Eluate Volume (ml)

Figure 1.The separation of magnesium, calcium, strontium, barium, andradium with ammonium lactate eluant -1. 5M at pH 7.1. 1 x 8 cm column, flow rate 1. 0 T = 78 C.(Milton and Grummitt, 29,p. 547) o\ Interstitial Column Volumes 6 8 10 20 30 40 60 80 100 10 T

cu 4 3

74) Ci) 4_41 . 0 0

O cd Sr 4a) a)

O U No Ba to 650 rrl 10 20 30 50 70 100 200 500 Eluate Volume (ml)

Figure 2.The separation of beryllium, magnesium, calcium, and strontium with 1. 5M hydrochloric acid cluant.1. 1 x 8 cm column, flow rate 0 ml /min, 60 C.(Milton and Grummitt, 29,p. 547) Interstitial Column Volumes 0.5 1 2 3 4 6 810 20 40 6080 100 200 10 1 1 1

Sr Mg tto

0 1.0 0 0 4-1 (r3

0j CU

0 U

I 1 , t I . I 1 I it ti 2 4 6 810 20 40 60 80 100 200 400 700 1000 Eluate Volume (m1)

Figure 3.The separaticn of magnesium, calcium, strontium, and barium with ammonium citrate eluant5% at pH 5.1. 1 x 8 cm column, flow rate 1.0 m1 /min, T = 78 C. (Milton and Grummitt, 29,p. 548) 00 9 1. 21M hydrochloric acid had been reported previously; but only partial separation had been achieved using a 2. OM hydrochloric acid solution

(7).Griswold and Nello (12) reported separating microgram amounts of calcium and magnesium from each other using 3M hydrochloric acid as eluting agent. Kraus and co-workers (18, 34) also used hydrochloric acid and ammonium chloride in separating radioactive tracers of the alkaline earth metals several years ago when they began a re-examination of the adsorptive properties of inorganic materials.One of the most attractive features of the inorganic adsorbents they tested, compared with standard organic ion-exchange resins, was their unusual and sometimes unique selectivity.Some of the "acid salts" (salts such as zirconium phosphate, zirconium tungstate, and zirconium molybdate) were found to have excellent selectivity for the alkaline earths. As demonstrated in Figure 4, the elements from calcium to radium were completely separated from each other with a relatively small column of zirconium molybdate.Because of the large differences in adsorbability, a stepwise elution technique was chosen. In a comprehensive study, Maeck, Kussy and Rein (24) did for four inorganic exchangers in nitrate solutions the same kind of thing that Kraus et al. did for anion-exchange resins in chloride solutions. Several investigators have reported separating carrier-free radioisotopes of the alkaline earths (4,9,38), while others (6,53, 55) Sr

0. 1M 0. 5M NH4 Cl 0. 2M NH4C1 1M NH4 Cl Sat'd NH4 Cl a- NH49 Ca 4-) ). 005mO. 005M HC1 b. 005M 0.005M HC1 0.01M HC1 j3 HC1 HC1 cd Ra 4-)

;-1 ai 2 Ba 0 ca a)

O 1

J11.11 . . La.* a anma L. 5 15 20 25 30 Number of Column Volumes

Figure 4.Separaticn of alkaline earths with zirconium molybdate (drying temp. 25 C, 2 0. 19 cmx 10. 0 cm column, flow rate 1. 1 cm /min). (K. A. Kraus et al. , I- 18,p.10) 11 have accomplished separating milligram amounts of these elements from cation exchange resin with ethylenediaminetetraacetic acid

(EDTA).Other eluants reportedly utilized in separating the alkaline earths on cation exchange columns include potassium chloride (49). ammonium formate (52), and diaminocyclohexanetetraacetate (39, 40,

50). The last decade has seen no radically new developments in ion- exchange chromatography, but rather a steady exploitation of existing methods. An outstanding example of this exploitation surfaced when, in the separation of the rare earths,it was demonstrated that ammonium a-hydroxyisobutyrate gave superior results compared with ammonium lactate and ammonium citrate.Since that time, several investigations have shown that this eluant is also extremely useful for the separation of calcium, strontium, and barium from each

other and from many other elements (2,8,33, 43, 54).Two procedures for the radiochemical separation of strontium-90 from fission

products are based upon the use of this eluant (5,17).

Purpose of This Study

The purpose of this study is to determine whether milligram quantities of magnesium, calcium, strontium, and barium, all combined in a single sample, can be quantitatively recovered, fairly rapidly, from an ion-exchange column by using a gradient elution 12 technique with a-hydroxyisobutyric acid.The basis for this investigation is the considerable interest shown in the quantitative determination and separation of these elements because of their extreme importance in environmental and biological substances (28). The significance of the precise determination of the alkaline earths has been justly emphasized by the bewildering number of papers which have been published on this subject, describing numerous methods and techniques which can be applied to a variety of situations. Therefore, many well-established techniques exist, such as flame and plasma emission, atomic absorption, colorimetry, direct and inverse polarography, mass spectrometry, and neutron activation analysis, which have the needed sensitivity for the determination of the alkaline earth elements and most of the other elements as well.However, most of these techniques have been devised for applications to specific samples. When these techniques are applied to a wide variety of environmental or biological specimens, many difficulties are encoun- tered in obtaining accurate values of concentration.To circumvent these difficulties in many cases, separations that could be avoided in rapid, approximate analyses have been found necessary for analyses of maximum precision and reliability, even though the final measurement is usually made by a selective technique such as atomic absorption. At the same time, the advent of modern chemical instrumentation 13 has itself necessitated the development of new and sometimes highly specific techniques of separation for most of the elements, so that these instrumental techniques can be fully exploited.Consequently, the goal of this study is to quantitatively separate and determine milligram amounts of the alkaline earth elements by means of cation- exchange and atomic absorption spectroscopy. 14 ILPRINCIPLES OF ION EXCHANGE

Resins

All resins can be thought of as consisting of tiny, highly porous, plastic balls, the surfaces of which (both inside and outside of the pores) are covered with chemically active groups--either acidic or basic.The plastic itself can be either a phenol-formaldehyde resin or any of a multitude of polymerized hydrocarbons containing sulfonic, carboxylic, or phenolic groups for the cation exchangers or some type of basic amine for the anion exchangers. A common form of cation resin is illustrated below.

CH=CH2 CH2 -CH- CH2 -CH- CH2 -CH- H2504 CH=CH2 styrene divinylbenzene 503H SO3H -CH -CH -CH -CH -CH-CH - e 1 2 2 2

In this case, styrene is polymerized utilizing divinyl benzene as the cross-linking substance.The benzene rings are then sulfonated leav- ing the surface of the plastic covered with highly reactive sulfonic acid groups.Since the bonding of the active hydrogen is predominantly ionic, there exists a situation which fosters easy exchange between the active hydrogen and some other cation.These ions are ionizable and 15 readily replaced.The resulting compound would be some group appended to a large organic aggregate. A monovalent cation would replace a single hydrogen ion, while a divalent cation would replace two nearby hydrogens. As might be suspected, a stronger charge is bound to a greater degree and retained better by the resin.Similarly, an ion with a large hydrated radius is somewhat hindered by itsphysi- cal size in adsorbing strongly, and consequently, is retained somewhat poorly.Both of these considerations tend to counteract each other and affect the relative retainability of the ion for the resin, together with other considerations such as the polarity of the molecule, its shape, and the affinity it has for the medium or vehicle in which it is dissolved while the exchange is taking place. The use of divinyl benzene as a cross-linking substance introduces some practical considerations in resin choice.Both the swelling of the resin and its capacity are affected by the degree of cross-linkage. When ion exchange resins are immersed in different media (acids, water, etc. ) they tend to take up some of the solution and physically increase in size; i. e. ,swell.Alternatively, in some cases, they shrink considerably, sometimes to the point where they break their continuity and prevent column operation.The swelling effect is directly dependent on the degree of divinylbenzene used in the resin. Low swelling effects are associated with a high degree of cross- 16 linkage, while high swelling effects are characteristic of a small percent of divinylbenzene.This is just one of the factors that must be considered in the selection of a resin. The capacity of a resin is simply the amount of any substance that can be absorbed per gram of resin--the amount of ion that a gram of resin will retain.The usual capacity range for most resins is from three to six milliequivalents of a compound for each gram of resin (dry weight).Increasing the amount of divinylbenzene in a resin decreases the number of available active groups or sites, and consequently, reduces the capacity of the resin and makes for a slow equilibrium exchange between the resin and the ion as it passes through the resin bed.Therefore, low cross-linkage resins have far greater capacity than high cross-linked resins, but they also have bad swelling characteristics.

Column Mechanics

Columns are generally prepared by inserting a small plug of glass or quartz wool in a glass column, slurrying an appropriate amount of resin into the glass column, and then carefully putting a small plug of glass wool on top of the resin to prevent disturbing the resin bed as the various solutions are passed through the column. Usually, the next step is to pre-condition the resin by passing a specific solution through the resin bed.The substance to be absorbed 17 is then loaded on the resin by letting the load solution pass down through the column bed.The column is washed, usually with distilled water, to remove the remainder of the loading solution from the glass wool plug and column walls above the resin bed.Finally, the absorbed species is eluted from the resin with a suitable eluent.

Column Operation

Columns operate in a manner similar to a precipitation reaction where the equilibrium among the hydrogen ions (from the active group on the cation resin), resin anions, cations in the load solution, and the anions in the load solution all react together to form a very insoluble compound, namely, the ionic compound consisting of the cation and the resin bead acting as the anion.This new compound then possesses a characteristic similar to a solubility product.If this solubility product is considerably small, it shows that the compound formed is only minutely soluble and that only extremely low concentrations of its ions will exist together in solution.If the solubility product is relatively large, the solubility is greater, and the cation tends to enter into the liquid phase of the solution.The proper resin favors the "precipitation" on the column of the desired cation, thereby forming an insoluble compound.Left in the solution will be the hydrogen ion removed from the sulfonic acid group, and the original anion from the solution.This reaction is illustrated by the following 18 equilibrium:

zR-H +Cz+ Rz-C + where hydrogen (H) ions of valence one and C ions of z valence are exchanged on a cation exchanger in contact with an electrolyte solution.In the equation Rz- is the ion exchanger anion. Like all equilibrium reactions, the one illustrated is constantly exchanging ion species, with a few of the insoluble molecules dissolving while an equal number of those in solution precipitate. Neglecting the theoretical approach, it is common to introduce a selectivity coefficient, K.For exchange of ions of equal ionic charge, the selectivity coefficient is defined as:

[C]r [H] KH (1) [H]r [C] but for exchange of ions of different ionic charges, it is defined by:

[F]y C [C]r KH (2) [ H]Yr[C]x Brackets [] represent the concentration (mg equiv /m1) in the external solution (no subscript) and in the resin phase (subscript r).The values of the charges of ions C and H are denoted by x and y, while C and H represent the ions being exchanged.Formally, the selectivity coefficient corresponds to the mass action product derived from the classical law of mass action, but it should not be interchanged with the thermodynamic equilibrium constant. 19 In a constantly flowing stream of eluent solution it is apparent that during that period of the equilibrium reaction when the ions are in solution, they will be carried down the column and re-precipitated. Since different metal cations will have different affinities for the resin, and hence, different equilibrium constants, it is apparent that some cations will be absorbed more strongly thanothers;i. e., their cations will be in the resin phase for a longer period of time.This sequential solution and re-precipitation of the metal cations of two different species, in a moving stream of eluent, will serve to mechanically separate the two cations in direct relation to the magnitude of the difference between their respective equilibrium constants or affinities for the resin.This reaction, occurring count- less times as the eluent moves through the column, serves to amplify even small differences in affinity, therebyproviding a useful means of separation.The ions most strongly held by the resin will move down the column at the slowest rate as the absorption and "desorption" processes continue.However, the removal of a cation of interest from such a column can be speeded up by reducing the concentration of the metal cation while in solution.This can be accomplished by using, as an eluent, some solution containing anions which strongly complex the released cations and prevent them from recombining with the resin. 20 Factors Affecting Separation

There are two principal factors which can produce a separation of ions on a resin; these are:

1. Charge of the ion; the greater charge is more strongly held;

2. Hydrated ionic radius; the smallest is held more strongly. Obviously both affect the relative affinity of the ion for the resin. This led to the description of various series of elements, where the relative affinity of a cation (in strong acid eluents) for the resin was listed in increasing order of strength. Another consideration affecting separation is the presence of a complexing anion which might serve to remove the free metal cation from solution and prevent its recombination with the resin.In this case two competing equilibria are present:

R-C +zH+ Cz++ zR-H z

and

Cz++ xA-Y (CAx)z-xy

Where AY is the complexing anion.The stronger the affinity of the cation for AY the less will be deposited on the column. Frequently, the concentration of the anion in solution of these organic complexing agents is extremely dependent on thehydrogen ion

concentration.Such acids ordinarily ionize in the following manner: 21 - -n H H++ (H A) 2H++(Hn-2A)=7n-----anH++ A nA = n-1 From this sequence it is apparent that relatively minor changes in pH would produce large differences in the amount of available complexing anion AY. The elution step represents the actual stage where separation occurs. As the eluent flows down the column, absorbed ions are desorbed and reabsorbed, depending on the stability of the resin compound, until ultimately, they are carried into the column effluent. The repeated desorption and absorption process, hopefully, has amplified the slight difference in the stability of the resin complex between species S and species T into an elution pattern that looks something like the following illustration:

103 T

I le

o 8 I 102 // e

10 5 10 15 20 2530 35 Number of Column Volumes'

1Column Volume: Signifies the resin-bed volume; i. e. ,length times area.In many references, column volume is used to signify the interstitial column volume, defined as iarea- length, where i is the interstitial space of the column (i. e. ,fraction of resin bed not occupied by resin).This interstitial space is approximately 0. 4/ cm3 for Dowex 50x8 resin. 22 In the elution sequence pictured above it appears that relatively pure species S could be obtained by collecting the effluent fraction in column volumes 5 through 15, and relatively pure species T by taking volumes 25 through 35.The slight tailing effect of each species into the other, shown for volumes 15 through 25, indicates incomplete separation.This overlap or tailing can generally be reduced by designing the ion exchange column to take advantage of any factor favoring the resin-cation exchange equilibrium.Like many equilibria, this exchange is favored by:

1. Increasing the temperature,

2. Slowing down the flow rate of the eluent to provide more time for the reaction to take place,

3. Providing a high ratio of resin to cation,

4. Using a fine mesh resin. In most procedures, ultimate selection of resin and column conditions is a compromise between speed, purity, and yield of the desired cation or anion.The completeness of the separation depends to considerable extent on the actual geometry of the column, particularly on the length of the column relative to the amount of resin necessary to accomplish the original absorption.If the column is too short, the supposedly absorbed ions may appear in the effluent even while loading; if too long, the ions will elute in diffuse bands. As a general rule, the length of the resin bed should be about 20 times 23 the amount of resin needed to absorb all the absorbable ions in the load solution.Area is of secondary importance, but should not be so large as to produce band diffusion by virtue of varying hydrodynamic flow characteristics.

Column Parameters

Distribution Coefficient

It is frequently desirable to devise separation procedures from published data relative to ion exchange experiments.Tables of distribution coefficients are the most valuable since they permit the application of several empirical formulae useful in such trial procedures.The distribution coefficient, symbolized by D,is defined by the following relationship: concentration of ion /gram of drresin D=concentration of ion /milliliter of solution (3)

Separation Factor

The separation factor, symbolized as a, is the ratio between two different distribution coefficients.For example, using the two species S and T:

D(T) [Tir [s] a = (4) D Fri [s] (S) r If a is large, species T tends to remain on the resin and species S will 24 be the first to elute.However, if a is close to unity, there will be a very poor separation, since the two species will tend to come off together.If a is less than one, species T will elute first.These relations are true, however, only where the eluent is constant, as in a steady flow of an inorganic acid (or constant molarity complexing agent with constant pH), where the distribution coefficients have been determined for that specific reagent.

Volume Distribution Coefficient

The distribution coefficient can be related to column geometry and column behavior by use of the volume distribution coefficient, usually abbreviated as D . v

Dv = D P (5) Rho ( p) in this relationship is the bed density expressed as grams of dry resin per cubic centimeter of resin bed.The bed density depends in turn on several factors, namely:

a. Mesh: This value merely indicates the physical size range for the resin beads,

b. Packing: Bed density depends in a small degree to the actual manner in which the column is physically packed,

c. Cross-linkage: Bed density is extremely dependent on the cross-linkage. The value of P can be found in resin tables prepared by the 25 manufacturers. The volume distribution coefficient, which is more suitable for chromatographic calculations, can also be conveniently determined by means of a column.On the basis of its definition, the volume distribution coefficient can be written as: vmax Dv A L where vmax is the volume (milliliters) of eluting solution, which moves the maximum concentration of the band down the total lengthof the column (L),in centimeters, of cross section A cm2and of void fraction i.On the basis of this equation, the value of Dv can be calculated from measurable data. Once the volume distribution coefficient is known it is possible to predict the performance of an experimental column, to a fair degree of accuracy, by the use of the following relationships:

1 d A E Dv + (7)

and

1 = V (8) E max=i+Dv=i+D p where E is the elution constant, Vmax is that column volume of the effluent in which the ion of interest appears in greatest concentration, and d is the distance an absorption band travels down a column on addition of v milliliters of eluent to a column of cross-sectional area

A. 26 The only difficult parameter to determine in the above equations is the interstitial volume or void fraction.Various experimental procedures, however, have been used to approximate this volume. These procedures are based principally on measurement of the effluent volume before some non-absorbed ion or dye appears to come off a prepared column of measured bed volume. A column volume as used here is simply AL (cross sectional area times length) of the resin bed.Most experimental data on ion exchange is published showing graphs of Dv vs molarity of eluent used. An example that applies to species 5 and T on a specific resin is illustrated below:

15 Dv 10

1 2 3 4 5 6 Motarity of Eluent (eluent specified)

Selecting a 2M eluent solution, and reading from the graph, the separation factor D /D is found to be 15/5, or 3.Therefore a (T) (S) good separation should be obtained.If a five cm3resin bed is assumed,i would be approximately two.Using equation (8), 27 V = 7, and Vmax(T) = 17, meaning one could expect approxi- max(S) mately a ten column volume separation between peaks in our elution sequence. It must be stressed that these empirical considerations apply only to a steady state elution sequence, where the eluent is not changing in any manner.Similarly, these relations will, at best, give only an approximation of the final performance of any real elution.All columns must be calibrated by experimentation, since the very best resins frequently vary from batch to batch, and since every column varies somewhat in size, packing, flow rate, temperature, etc.Only an extremely precise technique, under highly controlled conditions, will result in identical performances over a period of time. For detailed treatment of ion exchange theory and applicability, several books are referenced in the bibliography (13,15,19, 20, 27, 42, 44). 28 III.SEPARATION FACTOR DETERMINATION

Objective

The objective of this portion of the investigation was to study complexing behavior of magnesium and calcium with alpha- hydroxyisobutyric acid (a -HIBA) using Dowex 50 X8, 100-200 mesh, as the cation exchanger.Ko et al.(17) and Wish (54) have shown that calcium, strontium, and barium can be separated fairly rapidly by using various concentrated solutions of a -HIBA.Apparently, however, no one has investigated the feasibility of separating magnesium from calcium and the other alkaline earths using this complexing agent.Consequently, this study was conducted to ascertain what influence various concentrations of a -HIBA had upon the elution of magnesium and calcium from Dowex 50 X 8 resin, and, since complex equilibrium is dependent upon the pH of the solution, what influence pH had on the elution of these two elements.To restrict the study only to the differences in complex formation induced by changes in pH and molarity of a-HIBA, all other column variables such as flow rate, temperature, cross-linkage, and mesh size were held constant. It was hoped that such a study would lead to a good separation procedure for mixtures of the alkaline earths. 29 Apparatus

1. For the spectrophofometric determinations of magnesium and calcium, a Perkin-Elmer, Model 290-B, atomic absorption spectrophotometer fitted with a Perkin-Elmer premix air-acetylene burner was used. A calcium-magnesium multi-element lamp (Perkin-Elmer) was used as the radiation source.

2.A Beckman Zeromatic pH meter equipped with a combination electrode and a Heath Malmstadt-Enke pH Recording Electrometer equipped with calomel and glass electrodes were used for pH measurements.

3.Columns were constructed by drawing the end of a 28 cm length of 10-mm I. D. Pyrex glass tubing to a tip (I. D. approximately 1. 5 -mm). A slight flare of the tube diameter was made 3 cm from the other end of the column for the insertion of a number 1, one-hole rubber stopper. A small plug of glass wool was placed in the drawn out tip, and the column resin bed made by a distilled water slurrof cation resin.Slurry was added until the height of the resin bed (after resin had settled) was 4.2 cm, then another plug of glass wool was carefully placed on top of the resin.2Above the column is a gradient- elution system which gives the system a pressure head of about 85 cm from the top of the packed column (see Figure 5).

2Atop plug not only keeps the resin bed from being disturbed, but it also helps in achieving even absorption of a sample. 30

2 Liter Aspirator Bottle 0 MINIMAVAIPAMPARIMA:_warAIAIVAIAVAPAIAAVAIAIMPI/AWALIMIIIVIVAIIAVAIr 3-Way Stopcock

1 Liter Aspirator Bottle Magnetic Mixer

Screw Clamp

Glass Column

Figure 5.System with constant-volume mixing chamber for gradient elution. 31 4.A Research Specialties automatic fraction collector (time type) was used to collect 3. 5 ml samples of column effluent.The flow of the eluent through the columns by gravity feed was adjusted to 1. 2 ml /min by screw clamps.

Preparation of Solutions

a -HIBA

Alpha-hydoxyisobutyric acid was purchased from Continental Chemical Company, North Sacramento, California, as an aqueous solution of 50% a -HIBA and was used without further purification. For buffered solutions of various concentrations, the acid was adjusted with concentrated (-,30%) NH4OH to a specific pH and diluted with distilled water accordingly.

Calcium Stock Solution (2. 000 mg /m1) and Magnesium Stock Solution (2. 000 mg /m1)

These solutions were separately prepared by oven-drying their respective analytical-reagent grade carbonates, dissolving in a minimum quantity of 1M hydrochloric acid, and diluting to volume with distilled water.Both of these solutions were standardized by EDTA titration. All other solutions were prepared using distilled water and analytical grade reagents. 32 Pretreatment of Cation Exchange Resin

Dowex 50 X 8,100-200 mesh, analytical grade, hydrogen form (Bio-Rad Laboratories, 32nd and Griffin Avenue, Richmond, California) was treated as indicated in Table 1, after being slurried into a glass column.After the indicated treatment (the remarks column in Table 1 contains brief comments on the purpose of each step) the resin column was ready for use.

Table 1.Treatment Procedures for Cation Resin Columns. Volume of Remarks Solution Solution Used distilled H2O 35 ml washing 2M HC1 35 ml washing distilled H2O 30 ml removal of excess HCl 3M NH4OH 100 ml conversion of resin to ammonium form distilled H2O until column effluent is neutral to pH paper

Procedure for Peak Elution Volume Determination

The peak elution volume vmax, i, e. ,the volume of eluent required to elute the maximum concentration of an elution band, can be calculated from a known distribution coefficient, which in turn is obtained from equilibrium data.However, since the distribution coefficients of the species involved in this study were not known, vmax 33 was determined by a simple column method.

Column Method

On completion of preparing a column and pretreatment of the resin (p. 29, 32), a 2 ml sample (4. 00 mg) of either magnesium or calcium is quantitatively transferred to the top of the resin bed.After the load solution has drained down to the resin bed, the column walls are rinsed with 4 ml of distilled water.The absorbed species is then eluted by passing a known concentration of a -HIBA solution, buffered to a specific pH, down through the column.The column effluent is collected in 3. 5 ml fractions (one column volume fractions) in test tubes which had been previously placed in the automatic fraction collector.Finally, 1 ml aliquots are taken from each collected fraction and diluted to a suitable concentration for determination of the metal species by atomic absorption spectroscopy.The shape of each elution band (i. e. ,position of the peak and the width of the band) is observed by plotting concentration of collected fraction versus collected fraction (number of column volumes) or volume (ml) of column effluent.

Concentration Determination

The instrument used for concentration determinations of the collected fractions was a Perkin-Elmer atomic absorption 34 spectrometer operated under the following conditions:

1. The fuel gas was commercial acetylene in cylinders.The cylinder pressure was not allowed to fall below 75 psi, since there is danger of acetone (solvent for acetylene) being swept into the flame.The presence of acetone in the flame gives enhanced values for the absorption of alkaline earth metals.

2. Air was normal compressed air in cylinders. A cylinder was used until the pressure fell below 200 psi, after which instability of the flame was sometimes noticed.

3. The operating conditions used for concentration determinations of calcium and magnesium are given in Table 2.All of the conditions are those suggested by the manufacturers, except the fuel conditions which were established using 10 µg /ml and 2 µg /ml of calcium and magnesium solutions, respectively.

Table 2.Operating Conditions for Calcium and Magnesium. Calcium Magnesium

Wavelength 4227 X 2852 X Slit width 20 X 20 X Lamp current 6 mA 6 mA Fuel: Acetylene flow 14.00 14.00 Air flow 13.20 13.20 35 Distilled water was first aspirated through the flame followedby the calcium or magnesium solutions.This operation was repeated after each absorption measurement.

Results and Discussion

Peak elution volumes, vmax (in column volumes) for theelution of magnesium and calcium are plotted as a function of a-HIBA concentration at four specific pH values in Figures 6 and 7,respectively.3 As shown in Table 3, the separation factor remainsnearly constant for the molarities listed, despite the rapidchange in equilibrium distributions at the higher molarities. One noticeable point in Table 3 is the apparent changein a due to a change in pH.This can be explained by remembering that the

3Generally, the distribution coefficient (D) is plotted as afunction of concentration.From such a plot the separation factor (a), defined as D2 /DP is easily obtained.If distribution coefficients are not known, one can still obtain the separation factor for twospecie3 by eluting each species separately from a column, recordingits respective peak elution volume, and calculating a by: vmax L A=Dp +i If the length of the resin bed (L), cross-sectional area (A)and void fraction (i) are kept constant for the elution of both species, vmax will be proportional to D and, consequently, vmax 2 a vmax 1 .5

. 4

. 3

0 0 10 20 30 40 50 60 70 80 90 100 Peak Elution Volume (column volumes)

Figure 6.Influence of alpha-hydroxyisobutyric acid concentration and the effect of pH upon the elution of magnesium. Column:10 x 40 mm; Dowex 50 X 8 (NH4+; 100-200 mesh) Temperature: 22 C Flow rate:1.2 ml /min .8

.7 .6 ci 0 5 cd 4 G.) 'o.3

a) .2

00 20 40 60 80 100 120 140 160 180 200 Peak Elution Volume (column volumes)

Figure 7.Influence of alpha-hydroxyisobutyric acid concentration and the effect of pH upon the elution of calcium. Column:10 x 40 mm; Dowex 50 X 8 (NH4 '100-200 mesh) Temperature: 22 C Flow rate:1.2 ml /min 38

Table 3.Effect of Eluting Agent on Peak Elution Volumes. Eluting Agent (a-HIBA)Ca/Mg Ratio of Peak Elution Volumes Concentration H 4.0 H 4.5 H 5.0 H 6. 0 0. 20M 2. 08 2. 15 1. 80 1.50 0.22 1. 42 O. 25 2. 00 2. 28 2. 10 0. 30 2.08 2. 24 2. 17 0. 40 2. 00 . 0. 44 -- - 2.00 effect of complex formation on ion-exchange equilibriais governed principally by two factors. First, the extent of complexformation by a given cation is determined by thecomplex strength (i. e. ,the stability constants) and the concentration of the freecomplexing anion. Second, the separation factor of two cations depends onthe free-anion concentration.Consequently, the optimum separation factor could theoretically be obtained by properly adjusting this concentration. In column separations the complexing anion isintroduced as the developing agent; therefore, the free-anion concentrationin the column depends on the composition of this agent and is a directfunction of pH when the anion forms a weak acid. Usually the free-anion concentration is adjusted toobtain the best separation factor, but this can be accomplishedonly within certain limits since the free-anion concentration alsodetermines whether displacement development or elution developmentis taking 39 place.This point is illustrated by the curves shown in Figures 6 and 7 for the pH values of 4. 0,4. 5, and 5. 0.Within this pH range, calcium and magnesium are changing their preferences forthe ion exchanger to that of complex formation with the complexing anion. Apparently, the high a -hydroxyisobutyrate ion concentration at pH 5. 0 causes extensive complex formation and thereby reducesthe affinity of calcium and magnesium cations for the ion exchanger so much that now,NH4 is preferred.Therefore, displacement development is the prevailing technique occurring in this situation. At pH 4. 0 the concentration of free a-hydroxyisobutyrateions is much lower because a -HIBA is only weakly dissociated(pKa is approximately 4. 6).Here, calcium and magnesium cations are less extensively complexed and are therefore preferred by the ion exchanger, even though a-hydroxyisobutyrate ions are present.In this situation the separation is being accomplished predominantlyby elution. Nevertheless, the peak position and width of the separate elution bands were considered the paramount factors in achieving the separa-

tion.It was apparent from the data accumulated that several a-HIBA concentrations (dependent on pH conditions) could be used to success- fully separate magnesium from calcium.While in the process of determining the optimum conditions for such a separation, adecision was made (1) to extend the proposedmethod to include strontium and 40 barium, and (2) that the method should be applicable for the separation and determination of the alkaline earths from many materials. As a result of this decision, a synthetic sample containing approximately known concentrations of the alkaline earths and sodium, similar to those concentrations found in sea water (45), was used in determining a procedure for the separation of the alkalineearths. Exploratory experiments were made to determine the position of the peak and width of each elution band, using the Perkin-Elmer atomic absorption spectrophotometer for identification.From these experiments a gradient elution method was developed to quantitatively separate the alkaline earths from each other. 41 IV. ANALYSIS OF SYNTHETIC SAMPLE

Preparaticn

A solution containing 1322 fig /m1 magnesium, 399.2 rig /m1 calcium, 194.0 µg /m1 strontium, and 885.1 µg /m1 barium was prepared by oven-drying individually the alkaline earth carbonates, and dissolving each in a minimum quantity of IM hydrochloric acid.These were mixed together along with a measured volume of 4M NaOH standard solution (to give 1.058 x104µg /m1), pH adjusted to 2.5 with concentrated NH4OH, and then diluted to volume with distilled water. Apparatus, resin, and other solutions used, along with the procedure for pretreatment of resin and concentration determination, were the same as previously described.

Procedure

A 10.00 ml aliquot of the synthetic sample solution was trans ferred to the top of a 110 x 10 mm resin bed.This load solution was allowed to drain down to the level of the resin at a flow rate of approximately 3 ml per minute.Ten ml of distilled water was added to rinse the column walls and the top glass wool plug.It also was allowed to drain to the resin bed at the same rate as the load solution. The column was then connected to the gradient elution system (see 42 Figure 5), which contained 350 ml (percolumn)4 of 0. 22M a- hydroxyisobutyric acid buffered to pH 4.5 in the lower reservoir (one liter mixing chamber).The three-way stopcock was open to one upper reservoir which contained 2000 ml of the same solution as in the lower reservoir.The elution procedure was as follows:

(1)The eluting solution was passed through the column at a flow rate of 1.2 ml per minute (1. 5 cm per minute), being controlled by a screw clamp.

(2)Three hundred twenty nil (exclusive of the 20 ml correspond- ing to the load solution and water rinse) of this solution was allowed to pass through the column during which time sodium was eluted between 40-140 ml (col. vol. 5-14).

(3)The next 700 ml (col.vol. 33- 102) contained magnesium and was collected in a 1000 ml volumetric flask.

(4)After discarding the following 50 ml (col. vol. 103-107), a 200-m1 volumetric flask was placed under the column, the stopcock was turned open to the other upper reservoir which contained 2000 ml of 2M a-HIBA at pH 6. 0, so that this solution would start dripping into the lower reservoir

Each experiment was in duplicate,i. e. ,one column was called the control column and the other was called the standard column. Effluent from the control column was collected in 10 ml fractions in test tubes placed in the automatic fraction collector.Consequently, each fraction (column volume) could be checked by atomic absorption spectrometry to determine if any species was present, and if so, how much.Effluent from the standard column was collected in various volumetric flasks for yield determinations. 43 and gradient elution allowed to take place.Calcium was eluted in the first 170 ml (col. vol. 108-123).

(5)Ten ml was discarded (col. vol. 124), a 200-m1 volumetric flask was placed under the column and another 140 ml (col. vol. 125- 137) which contained strontium was collected.

(6)Ten ml was discarded (col. vol.138), another 200-ml volumetric flask was placed under the column and the barium was collected in the next 90 ml (col. vol. 139- 147). Finally,1 ml aliquots were taken from each fraction collected, diluted to a suitable concentration and estimated by atomic absorption spectrometry.The type of separation obtained is shown in Figure 8. A constant eluting concentration is necessary for the separation of magnesium from calcium, while gradient elution must be employed thereafter, otherwise calcium, strontium, and barium remain on the column too long and elute in diffuse bands. The columns are regenerated by washing with distilled water, followed by 3M ammonium hydroxide solution, and finally with distilled water, after which the column is again ready for use.

Results and Discussion

Because a good separation between sodium-magnesium and between magnesium-calcium was achieved with 0. 22M a -HIBA at pH 4.5, this solution was selected as the initial eluting solution.As 0. 22M HIBA, pH 4, 5

2, OM HIBA, pH 6.0 into 0. 22M HIBA, pH 4. 5

3 Mg Ba

0 0 10 20 30 40 50 60 70 80 90 100110 120130140 150 Fluent Volume (column volumes)

Figure 8.Separation of alkaline earths and sodium with ammonium alpha- hydroxyisobutyrate eluent, pH 4.5. Column: 110 x 10 mm; Dowex 50 X 8 (NH4+; 100-200 mesh) Temperature: 22 C Flow rate:1.2 ml /min 45 shown in Figure 8, the magnesium elution band is very broad and exhibits some tailing.Yet, even when 0. 30M a-HIBA at pH 4. 5 or 0. 20M a -HIBA at pH 5.0 was used as the eluting agent, the elution tailing tendency of magnesium was still apparent.Also, at these latter concentration and pH values, there were only about two column volumes separating sodium-magnesium and approximately a three column volume overlap of magnesium-calcium. According to the literature, this tailing tendency of magnesium has been observed by several other investigators during their ion-exchange studies. One important advantage of gradient elution is that it reduces

tailing.By adding 2. OM a-HIBA with a pH of 6. 0 to the 0. 22M a- HIBA at pH 4.5 already in use, the effective concentration and pHis gradually increased, producing a gradient down the column sothat the rear portion of the elution bands arealways in contact with a stronger eluting solution than the front portion.Consequently, the elution bands of calcium, strontium, and barium become respectively more narrow. Any procedure devised for the quantitative determinationof a constituent in a complex medium should have its precision and accuracy established if the results obtained by it are to be meaningful.To test the reproducibility and accuracy of the procedure previouslydescribed, six equivalent aliquots of the synthetic sample (runseparately) were separated following the recommended procedure and quantitatively 46 determined by using a Perkin-Elmer Model 290B atomic absorption spectrophotometer.However, only magnesium and calcium were determined, since the hollow cathode tubes necessary to determine strontium and barium were not available.The operating conditions used in determining the yields of magnesium and calcium werethe same as those listed in Table 2.Results of these analyses (given in Table 4) were calculated from standard calibration curves, where new standard curves were prepared for each sample analysis.Figures 9 and 10 are representative of the calibration curves plotted for magnesium and calcium. Even though the elution curves had shown clean separations between calcium-strontium and between strontium-barium, it still was necessary to get some quantitative datafor the analysis of strontium and barium.Therefore, 25 ml portions of each solution containing a separated alkaline earth species from sample number6 were sent, along with prepared standard solutionsof the individual alkaline earth elements, to the 1155th Technical Operations Squadron, McClellan AFB, Sacramento, California, to be qualitatively and quantitatively determined.The instrument used there was a Perkin- Elmer Model 403 atomic absorption spectrophotometer.All data were taken in the concentration modewith no attempt being made to electronically curve correct the data.Concentration values reported (see Table 4) were calculated from the prepared standard calibration Table 4.Analysis of Synthetic Sample (all values in milligrams). Sample Number

1 2 3 4 5 6 6** Mg--present 13.22 --found 12.69 12.83 12.81 13.35 13.54 13.29 13.14 Ca --present 3.99 --found 3.79 3.72 3.78 3.93 3.87 3.94 4.01 present 1.94 --found ------1.97 Ba --present 8.85 --found ------8.31

All four elements separated, but only magnesium and calcium were determined.Values reported were determined by using a Perkin-Elmer Model 290B atomic absorptionspectro- photometer. Values reported were determined by using a Perkin-Elmer Model 403 atomicabsorption spectrophotometer. All reported values were reproducible to better than 1%. 1. 1

1. 0

. 9 .8

. 7

. 5

. 4

. 3

.2 .4 .6 .8 1.0 1.2 1.41.61.8 2.02.2 Concentration (11g /ml)

Figure 9.Cali ration curve for magnesium, using the Perkin-Elmer 290B atomic absorption spectrophotometer. 2 3 4 5 6 7 8 9 10 Concentration (p.g /m1)

Figure 10.Ca;ibration curve for calcium, using the Perkin-Elmer 290B atomic absorption spectrophotometer. 50 curves and were reproducible to betterthan 1%. A nitrous oxide burner head was used for all determinations.The other operating conditions used in each individual determination are given inTable 5.

Table 5.Operating Conditions of Perkin-Elmer Model 403 Atomic Absorption Spectrophotometer. Barium Calcium Magnesium Strontium

Wavelength 5536 X 4227 X 2852 X 4607 X Slit 3X 4X 3X 4X Source Ba lamp Ca, Sr, Ba, Mg lamp Sr lamp Mg multi- lamp Fuel C2H2 C2H2 C2H2 C2H2 Flow rate 10.8 1/min 6 limin 6 I/min 6 I/min Oxidizer N20 Air Air Air Flow rate 12 l/min 17.6 1/min 22 1/min 22 l/min

The calibration curves obtained using this instrument areplots of relative concentration versus actual concentration asshown in Figures 11-14.The striking concavity of the calibration curves for strontium and barium are due to ionization problems.Manning and Capacho-Delgado (26) have reported the ionization of barium tobe about 92% in the nitrous oxide-acetylene flame, while Amosand Willis (1) have reported the ionization of strontium tobe around 84% in the same type of flame.The most common procedure used to control this effect is to add about 1% potassium (or anyother element 2.0 1.8

1.6 1.4

1.2

1.0

I I I I I I I 1 .2 .4 .6 .8 1.01.2 1.41.6 1.8 2.0 Actual Concentration (µg /ml)

Figure 11.Calibration curve for magnesium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 10

0 0 1 2 3 4 5 6 7 8 9 10 Actual Concentration (µg /m1)

Figure 12.Calibration curve for calcium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 32

28 z 24

C ° 16 a)

1-1; 12

00 2 4 6 8 10 12 14 16 18 20 Actual Concentration (µg /ml)

Figure 13.Calibration curve for strontium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 100

00 10 20 30 40 50 60 70 80 90 100 Actual Concentration (µg /ml)

Figure 14.Calibration curve for barium, using the Perkin-Elmer 403 atomic absorption spectrophotometer. 55 which will ionize much more readilythan barium or strontium) tothe sample being determined and also tothe standards being used. One intriguing detail observedwhile doing the exploratory experiments was that any alkalineearth sample supposedlyentirely absorbed by the resin, actuallyshowed that a minute amountof magnesium and calcium was notabsorbed.Changing the pH of the load solution, varying the amountof the load solution, andchanging the resin to theH+ form had no effect in suppressing this situation. From the six samples separatedand determined, thisunabsorbed portion was found to containapproximately 0. 015% of themagnesium and calcium in the sample.Presumably, barium and strontium would show similar results. 56 BIBLIOGRAPHY

1. Amos, M. D. and J. B. Willis.Use of high temperature pre- mixed flames in atomic absorption spectroscopy.Spectro- chimica Acta 22:1325- 1343.1966.

2. Baerg, A. P. and Rosalie M. Bartholomew.Yields in U-235 thermal neutron fission.Canadian Journal of Chemistry 35: 980- 985.1957.

3. Ballczo, H. and G. Doppler.Erdalkalianalyse II.Teil: Trennung und Bestimmung unter Zuhilfenahme des Fallungsaus- tausches und der Komplexbildung durch Athylenediamintetra- essegsaure (EDTA).Zeitschrift fur Analytische Chemie 152: 321-337.1956.

4. Bovy, Mlle R. and G. Duyckaerts.Separation Ba+2-Sr+2 sur Colonne D'echangeur D'ions.Analytica Chimica Acta 28:132- 149. 1963.

5. Bryant, E. A. ,J. E. Sattizahn and Buddy Warren.Strontium- 90 by an ion exchange method.Analytical Chemistry 31:334- 337.1959.

6. Budevski, 0. and N. Krasobaeva.Separation of strontium from calcium in the presence of EDTA by the ion-exchange method.Comptes rendus de L'Acedemie bulgare des Sciences 13:67-70.1960.(Abstracted in Analytical Abstracts 8:948. 1961)

7. Campbell, Dan N. and Charles T. Kenner.Separation of magnesium from calcium by ion exchange chromatography; application to determination of calcium oxide and magnesia in limestone and dolomites.Analytical Chemistry 26:560-561. 1954.

8. Drew, E. H. and M. W. Nash. A separation procedure for a large number of metal ions by ion exchange.Paper presented at the 148th meeting of the American Chemical Society at Chicago, Illinois.1964.

9. Duyckaerts, G. and R. Lejeune.Quelques Separations de Radioisotopes au moyen L'Acide Ethylenediaminetetraacetique. Journal of Chromatography 3:58-62.1960. 57

10. Eulitz, G.Zur Trennung der Erdalkalimetalle and des Bleis. Zeitschrift fur Analytische Chemie 178:360.1961.

11. Gorden, H. T. and C. A. Hewel.Paper chromatography of alkali and alkaline earth cations.Analytical Chemistry 27: 1471-1474.1955.

12, Griswold, Robert L. and Nello Pace.Microdetermination of calcium and magnesium in tissue ashes.Analytical Chemistry 28:1035-1037.1956.

13. Helfferich, Friedrich.Ion exchange.San Francisco, McGraw- Hill, 1962.624 p.

14. Honda, Masatake.Separation of alkaline earth cations by ion- exchange resins.Japan Analyst 3:132-133.1954.(Abstracted in Chemical Abstracts 48:9868.1954)

15. Inczedy, J.Analytical applications of ion exchangers. New York, Pergamon Press, 1966.443 p.

16. Khopkar, Shripad M. and Anil K. De.Cation-exchange behavior of barium on Dowex 50W-X8; separation from mixtures. Analytica Chimica Acta 23:441-445. 1960.

17. Ko, R. et al.Analytical methods for alkaline earth determinations in support of the strontium- 90 recovery program.20 p. (U. S. Atomic Energy Commission. HW-69677.1961)

18. Kraus, K. A. et al.Ion exchange properties of hydrous oxides. In:Proceedings of the 2nd United Nations International. Con- ference on the Peaceful Uses of Atomic Energy, Geneva, 1958. Vol. 28.Geneva, 1958.p. 3-16.(Paper 1832)

19. Kunin, Robert. Elements of ion exchange. New York, Reinhold, 1960.164 p.

20. Ion exchange resins.2nd ed.New York, Wiley, 1958.466 p.

21. Lerner, Morris W. and William Rieman III.Determination of the alkaline earths in glass.Analytica Chimica Acta 13:439- 441.1955. 58 22. Lerner, Morris W. and William Rieman III.Quantitative separation of the alkaline earths by ion-exchange chromatography. Analytical Chemistry 26:610-611.1954.

23. Li lova, 0. M. and B. K. Prevbrazhensky.Ion-exchange separation of elements.VI.Alkaline-earth elements.Radio- khimiya 2:731-733.1960.(Abstracted in Analytical Abstracts 9:561.1962)

24. Maeck, William J. ,Maxine E. Kussy and James E. Rein. Adsorption of the elements on inorganic ion exchangers from nitrate media.Analytical Chemistry 35:2086-2090.1963.

25. Majumbar, Santosh K. and Anil K. De.Cation-exchange behavior of calcium on Dowex 50 X 8; separation from mixtures. Analytica Chimica Acta 25:452-456.1961.

26. Manning, D. C. and Luis Capacho-Delgado.Dissociation and ionization effects in atomic absorption spectrochemical analysis. Analytica Chimica Acta 36:312-318.1966.

27. Marcus, Y. and A. S. Kertes.Ion exchange and solvent extraction of metal complexes. New York, Wiley-Interscience,1969. 1037 p.

28. Metal levels mirror metabolic disorders.Laboratory Management, p.19.March 1969.

29. Milton, Gwen M. and W. E. Grummitt.Ion-exchange methods for the quantitative separation of the alkaline earths, and their application to the determination of Sr-90 in milk ash.Canadian Journal of Chemistry 35:541-551.1957.

30. Morrison, George H. and Henry Freiser.Solvent extraction in analytical chemistry. New York, Wiley, 1957.269 p.

31. Nelson, Frederick, R. A. Day, Jr. and Kurt A. Kraus. Anion exchange studies.XXX. A number of elements in ethylenediaminetetraacetic acid solutions.Journal of Inorganic and Nuclear Chemistry 15:140-150.1960.

32. Nelson, Frederick and Kurt A. Kraus. Anion exchange studies. XIII.The alkaline earths in citrate solutions.Journal of the American Chemical Society 77:801-804.1955. 59

33. Perdue, Hazel D. et al.Chemical separations using a- hydroxyisobutyric acid solutions and both cation and anion exchange resin.Analytical Chemistry 40:1773-1776.1968.

34. Phillips, H. 0. ,Frederick Nelson, and Kurt A. Kraus.Rapid separation of 2. 6-min Ba-137 from 37-year Cs- 137 with zirconium phosphate used as a cation exchanger.In: Chemistry Division Semiannual Progress Report, U. S. Atomic Energy Commission, June 20, 1956.p.37.(ORNL-2159)

35. Pollard, F. H. ,J. F. W. McOmie and J. V. Martin.The quantitative analysis of the alkaline-earth metals by paper chromatography. Analyst 81:353-358.1956.

36. Pollard, F. H. ,G. Nickless and D. Spincer.Quantitative inorganic chromatography.Part IX.The analysis of alkalineearth metals by ion-exchange chromatography and flame spectro- photometry.Journal of Chromatography 10:215-220.1963.

37. Poskanzer, A. M. and B. M. Foreman, Jr. A summaryof TTA extraction coefficients.Journal of Inorganic and Nuclear Chemistry 16:323-336.1961.

38. Povondra, Pavel and Zdenck Sulcek.Analytische Schnellme- thoden zur Untersuchung von Metallen und Anorganischen Rohstoffen X. Anwendung der Ionenaustauscher zurBestimmung von Mangan und Alkalischen Erden.Collection of Czechoslovak Chemical Communications 24:2398-2404.1959.

39. Determination of calcium in strontium salts and strontium in calcium-rich materials.Chemist-Analyst 50: 79, 93-94.1961.

40. Povondra, Pavel et al.Untersuchung der Kationensorption aus Komplexanmedium-II; Chromatographische Trennung des Calciums und Strontiums. Talanta 8:705-710.1961.

41. Power, W. H. et al.Separation of radium and barium by ion exchange elution.Analytical Chemistry 31:1077-1079.1959.

42. Reichenberg, D.Ion-exchange selectivity.In:Ion exchange, a series of advances, ed. by Jacob A. Marinsky.Vol.1.New York, Marcel Dekker, 1966.p. 227-276.

43. Roberts, F. P.Determination of radio strontium by ion- exchange.8 p.(U. S. Atomic Energy Commission. HW- 60 69511.1961) (Microcard)

44. Samuelson, Olof.Ion exchange separations in analytical chemistry. New York, Wiley, 1963.474 p.

45. Slavin, Walter.Atomic absorption spectroscopy.In:Chemical analysis, a series of monographs on analytical chemistry and its applications, ed. by Elving, P. J. and I. M. Kolthoff.Vol. 25.New York, Interscience, 1968.307 p.

46. Stary, Jiri.Systematic study of the solvent extraction of metal oxinates.Analytica Chimica Acta 28:132-149.1963.

47. Stary, Jiri and Edward Hladky.Systematic study of the solvent extraction of metal p-diketonates.Analytica Chimica Acta 28: 227-235.1963.

48. Stary, Jiri and Justina Smizanska. A systematic study of the solvent extraction of metal cupferrates.Analytica Chimica Acta 29:545-551.1963.

49. Stevenson, D. E. and A. A. Wilson.Separation of calcium and magnesium in biological fluids using a cation-exchange resin. Clinica Chimica Acta 6:298-299.1961.

50. Sulcek, Zdenck, Pavel Povondra and Richard Stangl.Unter- suchung der Kationensorption aus Komplexanmedium-III; Chromatographische Trennung von Strontium and Barium. Talanta 9:647-651.1962.

51. Tompkins, Edward R.Separation of radium from barium by the use of an ion-exchange column procedure.Journal of American Chemical Society 70:3520-3522.1948.

52. Tsubota, Hiroyuki.Analysis of natural waters with ion- exchange resin and a formate buffer as eluent.Journal of the Chemical Society of Japan, Pure Chemistry Section 81:927-931. 1960.(Abstracted in Analytical Abstracts 8:3045.1961)

53. Wade, Marven A. and H. J. Seim.Ion exchange separation of calcium and strontium; application to determination of total strontium in bone.Analytical Chemistry 33:793-795.1961.

54. Wish, Leon.Quantitative radiochemical analysis by ion exchange; calcium, strontium, and barium.Analytical Chemistry 33:53-55.1961. 61

55. Wunsch, L.Ein neues Trennungsverfahren mit Hilfe von Ionenaustausch and Chelatbildung.Zeitschrift fur Analytische Chemie 158:364.1957.