Citation
Activation analysis of trace elements in serum

Material Information

Title:
Activation analysis of trace elements in serum
Creator:
Varcoe, Frederick Turner, 1944- ( Dissertant )
Bates, Roger G. ( Thesis advisor )
Winefordner, James D. ( Reviewer )
Harrahn, Robert J. ( Reviewer )
James, John H. ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1974
Language:
English
Physical Description:
ix, 203 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Ashes ( jstor )
Chelates ( jstor )
Chelation ( jstor )
Dietary trace elements ( jstor )
Gas chromatography ( jstor )
Geometry ( jstor )
Ions ( jstor )
Irradiation ( jstor )
Ligands ( jstor )
Sodium ( jstor )
Blood -- Analysis ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Trace elements -- Analysis ( lcsh )
Trace elements in the body ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
This research deals with radiochromatographic separations in neutron activation analysis of trace elements in serum. Understanding the metabolic roles of trace elements as well as their toxicological effects in biological systems has been limited by the sensitivity and selectivity of existing analytical techniques. The lack of human serum concentration data for many trace elements prevents optimum clinical utilization of this extremely powerful diagnostic tool. Neutron activation analysis has extremely low limits of detection for a large number of elements. However, interference from those elements present in human serum at relatively high levels must be eliminated before even a high-resolution solid state detector can be effective. The utilization of high sensitivity sodium iodide detectors requires still more separation. Chromatographic separations of metal chelates were developed for both pre- and post-irradiation separations. Quantitative elution from the gas chromatographic column for fluorine substituted beta-diketonate complexes of Cr, Cu, Mn, Fe, Be, Lu, and Gd was studied; recoveries between 52 and 92% were found. Extensive studies of decomposition and adsorption losses in the system are reported. Construction and evaluation of various sampling systems, transfer lines, and counting geometries are included. A completely reversible counting geometry for radioactive metal chelates using wide-range temperature and carrier gas control to trap and subsequently release eluate fractions is described. Also included is an apparatus for carrying out chelate separations from solutions whose large volumes preclude gas chromatographic injection. The successful use of a chelating ion exchange resin in human serum neutron activation analysis is reported. Operating conditions which optimize sodium removal with respect to sample loss are described.
Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 185-202.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022918040 ( AlephBibNum )
14175125 ( OCLC )
ADB3708 ( NOTIS )

Downloads

This item has the following downloads:


Full Text












ACTIVATION ANALYSIS OF TRACE
ELEMENTS IN SERUM












By

FREDERICK TURNER VARCOE, JR.


A THESIS PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY












UNIVERSITY OF FLORIDA


1974













ACKNOWLEDGEMENTS

The author would like to thank his research director, Dr.

Stuart Cram, for suggesting the research problem, for valuable

advice, suggestions, and support, and for a wealth of practical

experience. Thanks also go to the other members of the super-

visory committee, Dr. Roger Bates, Dr. James Winefordner, Dr.

Robert Hanrahan, and Dr. John James for their confidence and

help. The project could not have been completed without them.

Thanks also go to the machine shop personnel, in partic-

ular to Mr. Arthur Grant, Mr. Chester Eastman, and Mr. Frank

Ottinger for their expert fabrication of many pieces of equip-

ment used in this work. The glassblowing skills of Mr. Truman

Robbins, Jr., are also greatly appreciated.

The reactor operators and other personnel, in particular

Mr. Henry Gogun at the University of Florida and Dr. Milton

McLain, Jr., at Georgia Tech deserve heartfelt thanks for pro-

viding the irradiations necessary for these experiments.

The author also appreciates the contributions made by the

other members of his research group. Dr. David Cottrell deserves

special recognition for the design and assembly of the computer

interface and software with which all the high-resolution gamma-

ray spectra were recorded. Dr. Jack Leitner's expert help with

the operation of the system and his encouragement and advice

are also appreciated.

The author is deeply grateful to his parents for their sup-

port, encouragement, and sage advice throughout his education.

iii















TABLE OF CONTENTS


Page
ACKNOWLEDGEMENTS.............. ....................... iii

ABSTRACT..................................... ..........

CHAPTER I

Introduction. Trace Analysis........................ 1
Trace Elements in Serum............................ 1
Elements Identified................................ 2
Function. .......................................... 2
Disease Relationships Established................. 7
Heart disease ..................................... 7
Cancer. .......................................... 8
Other diseases................. .... .............. 9
Serum Trace Element Analysis...................... 10
Flame Spectroscopy................................ 11
Atomic absorption................................ 13
Atomic emission................................. 14
Atomic fluorescence.............................. 18
Arc-Spark Emission Spectroscopy................... 18
Colorimetry and Fluorimetry....................... 19
Colorimetry..................................... 19
Fluorimetry .................................... 21
X-Ray Techniques.................................. 21
X-Ray fluorescence... ............................ 21
Electron probe and laser probe X-ray microanalysis 23
Mass Spectrometry................................. 23
Spark source mass spectrometry.................. 23
Isotope dilution mass spectrometry.............. 24
Electrochemical Methods............................ 25
Gas Chromatography................................ 26
Metal chlorides.................................. 27
Organometallic compounds ........................ 28
Metal chelates................................... 28
H(tfa).......................................... .31
H(hfa) ... ........................................ 38
H(fod)................................... ........ 39
Instrumentation................................... 43
Injection ........................................ 43
Doctors. ......................................... 44
Serum application................................ 45
Neutron Activation................................. 46
Instrumental neutron'activation analysis ........ 47
Destructive activation analysis................. 50
Ashing. .......................................... 52









Table of Contents Continued.....
Page

Precipitation.................................... 53
Distillation....................................... 53
Electrochemical methods........................... 54
Dialysis........................................... ....... 54
Gel filtration.................................... 55
Szilard-Chalmers effects.......................... 55
Solvent extraction............................... 56
Inorganic separations .............................. 57
Ion exchange....................................... 58
Paper chromatography.............................. 61
Gas chromatography ................................ 63

CHAPTER II

Experimental.......................................... 65
Physical Facilities and Equipment..................... 65
Nuclear Reactors...................... ........... . 65
UFTR.............................................. 65
GTRR.............................................. 65
Separation Equipment ................................ 65
Low-temperature dry asher......................... 65
Gas chromatograph................................. 65
Vapor-phase separation apparatus.................... 68
Counting Geometry ................................... 68
Stainless steel spiral............................. 70
Charcoal traps.................................... 70
Reversible counting geometry...................... 73
Counting and Data Processing Equipment.............. 76
High-resolution y-ray spectrometer................ 76
NaI(Tl) scintillation spectrometer................ 77
Lead shielding..................................... 78
Procedures............................................ 78
Ashing.............................................. 78
Chelation from Solid Ash .............................. 79
Irradiation........................................ 79
Gamma-ray counting................................. 79
Chelate Extraction from Ash Solution................ 80
Carrier preparation............................... 80
Irradiation........................................ 80
Chelation......................................... 80
Vapor-phase separation.............................. 83
Gamma-ray counting................................. 84
Annealing Studies................................... 84
Chelates with no treatment........................ 84
Chelates with treatment............................ 85
Gas Chromatographic Studies......................... 86
Materials......................................... 87
Chelate synthesis .................................. 87
Cr (tfa)3.......................................... 87
Cr (fod) ..........................................









Taole of Contents Continued...
Page

Be(tfa)2 ........................................... 88
Fe(fod) .......................................... 88
Cu(tfa) .......................................... 89
Gd (fod) .......... ................................ 89
Lu (fod) .......................... ................. 90
Gas chromatography.................................. 90
Cr(tfa) .......................................... 90
Cr(fod)............................................ 90
Be(tfa) .......................................... 90
Fe(fod)3 ........................................ 93
Cu(tfa)2 ........................................... 93
Gd (fod) .......... ................................ 93
Lu (fod)3.......................................... 94
Cu(hfa)2 and Cr(hfa)3 ............................ 94
Mn(hfa) 2 .......................................... 94
Chelate residue exchange....... ..................... 94
Silanization........................................ 95
Counting geometry operation......................... 96
Stainless steel spiral............................. 96
Charcoal traps.................................... 96
Simple reversible geometry......................... 96
Improved reversible geometry...................... 97
Ion Exhange......................................... 97
Postirradiation separation.......................... 98
Sodium separation optimization.................... 98
Trace element determination......................... 98
Preirradiation separation........................... 98
Short irradiation ................................ 98
Resin blank ...................................... 99
Solution blank.................................... 99
Serum solution ................................... 99
Irradiation and counting.......................... 100
Long irradiation.................................... 100

CHAPTER III

Results and Discussion................................ 101
Objectives............................................ 101
Chelation Studies .................................... 102
Chelation Procedure................................... 109
Chelation from solid ash............................ 110
Advantages.......................................... 110
Procedures........................................ 110
Preirradiation separation......................... 111
Postirradiation separation........................ 121
Extraction from solution............................. 129
Advantages.......................................... 129
Procedures........................................ 129
Carrier Preparation................................... 133
Additional Separation ................................. 139
Interferent Removal................................... 142









Table of Contents Continued...
Page

Annealing Studies................... ................... 145
Untreated chelates.............. .................. 145
Treated chelates................................. 147
Gas Chromatography.................................... 148
Incomplete Elution.................................... 148
Residue measurement.............................. 148
Injection procedures................................ 151
Counting Geometry..................................... 153
Stainless steel spiral............................. 158
Charcoal trap ..................................... 163
Reversible counting geometry...................... 163
Ion Exchange.......................................... 170
Postirradiation Separation............................ 170
Sodium separation optimization................... 170
Serum analysis.................................. 173
Preirradiation Separation............................. 173
Short irradiation................................. 173
Long irradiation................................. 174
Conclusions......................................... 179
Sequence of Operations................................. 179
Preirradiation separation......................... 179
Postirradiation separation........................ 179
Separation Techniques ................................... 181
Chelation......................................... 181
Chelate separation................................ 181
Ion exchange....................................... 183
Future Work........................................... 183

REFERENCES............................................ 185

BIOGRAPHICAL SKETCH................................... 203










Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



ACTIVATION ANALYSIS OF TRACE
ELEMENTS IN SERUM

By

FREDERICK TURNER VARCOE, JR.

June, 1974

Chairman: Roger G. Bates
Major Department: Chemistry

This research deals with radiochromatographic separa-

tions in neutron activation analysis of trace elements in

serum. Understanding the metabolic roles of trace elements

as well as their toxicological effects in biological systems

has been limited by the sensitivity and selectivity of

existing analytical techniques. The lack of human serum

concentration data for many trace elements prevents optimum

clinical utilization of this extremely powerful diagnostic

tool.

Neutron activation analysis has extremely low limits

of detection for a large number of elements. However,

interference from those elements present in human serum at

relatively high levels must be eliminated before even a high-

resolution solid state detector can be effective. The

utilization of high sensitivity sodium iodide detectors

requires still more separation.


viii









Chromatographic separations of metal chelates were

developed for both pre- and postirradiation separations.

Quantitative elution from the gas chromatographic column

for fluorine substituted beta-diketonate complexes of Cr,

Cu, Mn, Fe, Be, Lu, and Gd was studied; recoveries between

52 and 92% were found. Extensive studies of decomposition

and adsorption losses in the system are reported. Construc-

tion and evaluation of various sampling systems, transfer

lines, and counting geometries are included. A completely

reversible counting geometry for radioactive metal chelates

using wide-range temperature and carrier gas control to trap

and subsequently release eluate fractions is described. Also

included is an apparatus for carrying out chelate separations

from solutions whose large volumes preclude gas chromato-

graphic injection.

The successful use of a chelating ion exchange resin

in human serum neutron activation analysis is reported.

Operating conditions which optimize sodium removal with

respect to sample loss are described.














CHAPTER I

INTRODUCTION. TRACE ANALYSIS


Trace Elements in Serum

Most of the elements in the geochemical environment can

be detected in animal and plant tissue if a sufficiently

sensitive technique is available (1). Elements are present in

bulk and trace concentrations. The trace elements, usually

considered to be those present at concentrations below 100 ppm,

can be divided into two groups: those which serve an essential

biochemical function without which subnormal functioning or

illness results, and those which are nonessential. Abnormally

low concentrations of an essential element can cause a defi-

ciency disease, while the lack of a nonessential element has

no ill effects. Excess of either essential or nonessential

elements can be toxic, however. As our knowledge of the

biochemistry of trace metals grows, more elements may come to

be considered essential. It is already known that many of

those not now considered essential have very important roles

in metabolism.

The concentrations of essential trace elements within

the organism are maintained within very narrow limits, which

is one reason why so much attention has been given to the

relationships of these elements to normal and abnormal metabolic

functions (2). Normal concentration ranges for nonessential









elements also exist, but are much broader than those of

essential elements. The trace elements must be present in

the diet. Their only sources are the earth's crust and the

ocean. Life cannot exist without them (3).

Elements Identified

Those elements considered essential for healthy human bio-

chemical functioning (1,4) are listed in Table 1, along with

serum concentrations which have been found in normal human

subjects. Also included are elements considered nonessential

environmental contaminants, often having toxic effects.

Reports dealing with elements whose serum concentrations are

relatively high, or which for other reasons can be determined

using less sophisticated analytical techniques, have existed

for some years. Among these are Cu (5-48), Zn (19-59), Mn

(14-27, 56, 59-62), Cr (13, 41-44, 57-71), I (47, 48, 58, 72-

79) and Fe (11, 24-27, 40-42, 56, 58, 80-83). Recently new

techniques have allowed identification of as many as 14

elements simultaneously in blood and serum samples (26, 40,

53), while V (84), Ba (85), Co (86), Sr (85, 87), U (88, 89),

Ca (90-95), Mg (95, 96), Hg (97-99), As (99), Pb (100-102),

Cd (101), Se (103-105), Au (106-108), Te (109), Li (110), W

(111), Be (70, 112, 113), Br (114) and F (115), have been

determined individually. Several review articles have been

written on trace element analysis in biological systems (116,

117, 118).

Function

The importance of trace elements lies in their roles in











Table 1. Essential and Nonessential Elements in Human
Biochemistry

Element Serum Concentration Reference
pg/ml


Essential
Bulk

Ca

Cl

K

Mg

Na

P

S

Essential
Trace

Co

Cr

Cu

F

Fe

I

Mn

Mo

Ni

Se

Si

Sn

V

Zn


(116)

(116)

(116)

(116)

(116)

(116)

(27)


2.8

1,000

1,700

13

3,250

0.50

5,400




0.00018

0.02-0.05

1.16

0.028

1.14

0.0029-0.11

0.0083

0.4

0.0042

1.1

7.0

0.033

1.0

0.98


(117)

(69)

(117)

(117)

(117)

(117)

(117)

(117)

(117)

(117)

(27)

(117)

(117)

(117)









Table 1. continued...

Element Serum Concentration Reference
Vg/ml


Nonessential


0.28

5

0.19

0.03

2

0.059-0.066

<0.004

830

<0.1

0.05

0.0027-4.0

18

0.031

0.046

14

0.054

0.04

0.057

0.4

0.048-0.072


(116)

(27)

(117)

(116)

(27)

(116)

(117)

(116)

(117)

(27)

(116)

(116)

(117)

(117)

(27)

(117)

(116)

(117)

(27)

(116)









biological functioning. They take part as active components

of enzyme systems, in transport mechanisms, in tissue syn-

thesis, and in bone formation (2).

The concentrations of trace elements in marine organisms

are fixed by the concentrations of trace elements present in

the surrounding water. In terrestrial animals, highly

developed homeostatic processes are required to maintain trace

element concentrations at fixed levels in the face of widely

varying amounts in the environment. When, for one reason on

another, an essential trace element's concentration lies out-

side the normal range, a serious metabolic upset must also

be present.

Trace element concentration variations have been impli-

cated in diseases ever since it became known that these

elements were an integral part of metabolism. It has become

obvious that knowledge of the functionof trace elements in

metabolism should help the diagnosis and treatment of disease

(2).

Most diseases of any consequence are accompanied by

changes in the concentrations of one or more trace elements

in some tissue or body fluid. However, it must be kept in

mind that only a few diseases have abnormal metal metabolism

as a primary underlying cause, even though correcting the

imbalance will sometimes alleviate the symptoms, e.g.,oral Cr

dosage will improve abnormal glucose tolerance, although

abnormally low serum Cr probably does not cause diabetes (119).

Even though many changes in the normal tissue concentration

of a trace element are nonspecific indications of an abnormal








condition, the analysis of body fluids and tissues for trace

elements is becoming an important clinical test in humans.

Blood is both the easiest to obtain for analysis and

the most pervasive of the body's tissues. It travels every-

where in the body, providing an exchangeable reservoir of

elements and compounds for all the other tissues. The com-

ponent of blood which is most useful for trace element studies

is the serum which remains after the blood cells and clotted

fibrin have been removed by centrifugation.

Only a small fraction of the trace elements in serum are

present as free ions. Most are in the form of complexes.

They may be loosely bound to a ligand such as albumin, or

tightly bound as in the iron-porphyrin compounds (2, 120).

This provides a high degree of selectivity in the combinations

available for metal complex formation (2, 120), and thus

gives the body the means to select and accumulate the required

metals from circulating body fluids. These complexes assist

homeostatic mechanisms in maintaining constant trace ele-

ment concentrations by forming or dissociating when necessary.

Also, heavy metals can be maintained in the system at non-

toxic levels while still providing the mineral reservoir

needed for normal metabolism. The critical balance between

metal ion and metal complex is easily upset by any of several

factors which affect the stability constants of the complexes.

Today scientists need to learn more about the biochemical

phenomena that depend on these metals and more about the

stability constants and specificity of the biochemical ligands.

The serum concentrations of trace elements have the broadest









range of diagnostic applicability and therefore are of the

most interest at this time from the standpoint of improving

disease diagnosis and etiological knowledge, and understand

ing trace metal metabolism in the human body.

Disease Relationships Established

Heart disease.-Heart disease is responsible for nearly

half of the deaths in the United States each year. For this

reason, it has received much study. Recently the possibility

that trace elements play a part in cardiovascular diseases

has become evident. Studies exploring this possibility show

how trace element concentration information can play an impor-

tant part in understanding and diagnosing a disease.

Injured heart tissue from patients who die from heart

attacks (acute myocardial infarction) show abnormally low

concentrations of Co, Cs, K, Mo, P, Rb and Zn and elevated

concentrations of Br, Ca, Ce, La, Na, Sb and S (121). Serum

levels in patients with myocardial infarction are also abnormal.

Increases in serum Cu (11, 14, 15), Mn (14, 15), Ni, Mo and B

(122) have been found, as well as a decrease in serum Zn

(2, 123). Serum Cu determination has been suggested as a

diagnostic test for patients with myocardial infarcation (11).

Trace element intake in drinking water has been linked

with death from myocardial infarction. A highly significant

negative correlation has been found between water hardness

expressed as Ca ion concentration and the death rate in the

local population due to degenerative cardiovascular disease

in several countries, including the U.S. (124-126). Calcium









may only be a benign indicator of the presence or absence of

another trace element which takes an active part in degenera-

tive cardiovascular disease. For example, Ca is thought to

inhibit the transfer of divalent metals across intestinal

membranes (93). Perhaps this inhibition controls the presence

of a trace element which takes an active part in the disease.

Serum cholesterol has been linked to cardiovascular disease.

The presence of trace quantities of V and Fe decreases choles-

terol synthesis, while Cr and Mn have the opposite effect (127).

Administration of CaNa2EDTA has been shown to reduce serum

cholesterol and to increase urinary output of certain trace

elements, especially Zn (123).

Cancer.-Trace metal studies have also been made in con-

nection with cancer, another major cause of death in the U.S.

Brune found no variation in the serum concentrations of 12

elements in leukemia patients (26), while other workers have

found abnormal Fe (80) and elevated Zn (52) concentrations.

High serum Cu concentrations are present with breast (12)

and prostate (15) cancer, chronic and acute leukemias (2),

and Hodgkin's disease (2). Hrgovcic as cited by McCall (2),

gives an example of how determination of trace element levels

can aid in evaluating the treatment of disease. In cases

where malignancies are being treated, changes in the serum

Cu concentration can appear long before there are any clinically

apparent changes.

Studies attempting to correlate the death rate from

cancer with various types of trace element environments in









different geographical areas have been made. Berg, in his

study of carcinogenic trace metals in water supplies,states

that although the literature contains some speculation,

almost no convincing data exist with respect to these cor-

relations (128). Pories, however, states that the large

differences in death rate from cancer in different parts of

the U.S. suggest that environmental factors may be the principal

and preventable causes of cancer (129). He also states that

elements can inhibit or foster tumor growth. Selenium and

platinium interferewith tumor growth; iodine prevents thyroid

neoplasia. Zinc metabolism may be the key for limiting tumor

growth by manipulating necessary protein synthesis (129).

Other diseases.-The relationships between serum trace

elements and other diseases have also received attention.

Protein bound iodine determination in serum has been used for

diagnosing abnormal thyroid activity for some time (72-79).

Abnormal serum Cu and Zn concentrations have been linked with

rheumatoid arthritis (31). In sera from arthritic patients

Cu, Ba, Cs, Sn and Mo concentrations were elevated, Zn, Fe

and Pb concentrations were depressed, and Al, Ni, Sr, Cr and Cd

concentrations were normal (40). Hepatitis and liver cirrhosis

are accompanied by elevated serum Zn and Mn (15, 49, 52).

Serum Cr is depressed and Mn is elevated in individuals

suffering from diabetes mcllitus (15, 119). Oral Cr supple-

mentation will improve such patient's glucose tolerance (15).

Zinc is required by more enzyme systems in the human

body than any other metal (2). In general, decreases in








serum Zn are accompanied by increases in serum Mn, although

this is not true in every case. Their antagonism has been

noted in lymphoma, pneumonia, and other diseases (2).

Low Cu levels have been measured in patients,with Wilson's

disease (9), chronic and acute leukemias, aplastic anemia,

cirrhosis, viral and microbial infections and nephritis (2).

Elevated serum Zn accompanies psoriasis (30). Abnormal trace

element concentrations have even been correlated with mental

illness. Serum Zn and Cu concentrations have been measured

in schizophrenics. Both are needed for histaminase enzymic

action, variation of which has been associated with schizo-

phrenia. The Cu concentration was found to be abnormally

high, but returned to normal following effective therapy (33).

Zinc and copper are antagonists in the human body. Oral Zn

treatment has been used successfully with such patients,

apparently forcing the Cu concentration down. Penicillamine,

a chelating agent which removes trace metals from the body,

has also been used for effective treatment (33).

Data from trace element studies using new analytical

techniques can provide the basic information necessary for

a better understanding of the roles of trace elements in

diseased and healthy metabolic activity. These techniques

can then be used to obtain the necessary information for

individual diagnosis.

Serum Trace Element Analysis

Research on the metabolic roles and toxicological effects

of trace elements in biological systems is highly dependent









on the development of new analytical techniques which are

-reproducible, rapid, sensitive, specific, and suitable for

automation. The lack of such methods has thus far impaired

development of the vast potential of serum analysis for

diagnosis and treatment in the medical sciences.

Many instrumental methods for elemental analysis in

serum have been reported, but the number well suited to trace

analysis is small. The techniques reported have been limited

in scope by several considerations. Matrix interference is

severe. The Na concentration is 2-3 mg/ml, 104 times greater

than even the higher trace element concentrations. With five

or six other elements present at macro levels (Table 1), de-

termining concentrations for the large number of trace elements

present at ppm to sub-ppm levels requires advanced techniques,

especially if highly desirable simultaneous multi-element

analyses are to be achieved. The lower limits of detection

presently available have undoubtedly limited the scope of

previous determinations. The more popular trace element

techniques, their advantages, shortcomings, and potential

for serum analysis are discussed below.

Flame Spectroscopy

Flame spectroscopy is the most sensitive of the tech-

niques requiring only moderately expensive equipment. More

than 60 elements, including all the metals, can be determined

this way. Interference is less likely than when using many

other trace analytical techniques. Chemical interference,

those which prevent the atoms from reaching the free, dissoci-









ated, un-ionized state necessary for measurement in the flame,

and spectral interference due to bands emitted by the OH and

CN radicals can present problems. 'Spectral interference

from other trace elements in the sample are seldom encountered.

However, the total salt concentration in the liquid being

analyzed should be kept below 1%, which can require sample

dilution and adversely affect trace determinations (130).

Tenfold dilution of serum samples has been found necessary

by some workers to avoid clogging the burner (32). A solvent

extraction step prior to analysis can eliminate the need for

dilution, although the risk of contamination is greatly

increased. Whatever separations are required prior to deter-

mination are usually simple (131).

Probably the main disadvantage flame spectroscopic

techniques have is that except in the case of atomic emission,

only one or two elements can be determined at a time.

Research is currently under way to remedy this, but until

these systems become fully operational either another technique

must be used or one must make a separate series of measurements

for each element.

Atomic absorption (AA), atomic emission (AE), and atomic

fluorescence (AF) are complementary spectroscopic techniques.

When the limit of detection must be optimized, AA and AF are

generally preferred for elements with resonance lines below

300 nm, since there is usually insufficient energy available

in a properly adjusted flame to excite these higher energy

transitions. For elements with resonance lines above 400 nm,









there is adequate flame energy available and an external

radiation source is not required. Elements whose resonance

lines lie between 300 and 400 nm can usually be determined

equally well by any of the three techniques (130).

Atomic absorption.-Those elements for which AA has a

lower limit of detection than AE by at least a factor of 5

are listed in Table 2. The classification is based on com-

parison, as flame techniques, without regard for the fact that

in some cases where sample size is limited, and in other cases

where the nonflame technique is unusually sensitive, a lower

limit of detection may be obtained with nonflame AA.

Table 3 lists those elements for which AA and AE have

limits of detection which are comparable. Together these

lists contain 51 elements. When Table 1 is compared with

Tables 2 and 3, it can be seen that of those essential trace

elements which can be determined spectroscopically all have

AA detection limits equal to or lower than the corresponding

AE limit. With this in mind, it is not surprising that AA

has been the most popular spectroscopic technique for trace

element analysis of serum. Of course, classification as

essential requires an adequate method for detecting the

presence of the element in question, and as other techniques

are applied to trace element analysis of biological samples

the essentiality of other elements may be discovered at levels

too low for analysis by AA.

A few representative determinations of trace elements

in serum by AA include Ca (90, 94); Fe, Cu and Zn (32, 34,









50); Au (108); Li (110); and Pb (77). Two of these references,

Au (108) and Li (110), describe metal determinations in sera

of patients currently undergoing therapy with these metals and

therefore involve abnormally high concentrations. Techniques

which require only microliter samples are being used more

frequently for serum analysis. The Delves cup technique

has been applied to serum chelate extracts of Pb and Cd (102),

and the carbon rod atomizer has been used to determine Mg, Fe,

Cu, Pb and Zn concentrations in serum (35). These techniques

are especially valuable where multielement determinations are

required, since until simultaneous multielement spectroscopic

determinations are improved, a separate portion of the sample

will have to be used for each element determined.

Atomic emission.-Those elements for which AE has a

lower limit of detection than AA by at least a factor of 5

are listed in Table 4. None of the essential trace elements

appears on this list, although several of the bulk essential

elements do. Tables 3 and 4 combined list 43 elements,

including five essential trace elements, for which the AE

limit of detection is at least as low as the AA limit of

detection. AE, then, can be a valuable flame spectroscopic

tool in serum trace metal studies, both because it is best

for several elements presently known to take an active part

in metabolism, and because it can be used for studying elements

for which AA is not adequate. Besides this, it may also

provide a basic technique for multielement analysis. For

example, eight elements have been determined in blood using

plasma torch excitation of an emission spectrum (138).









Table 2. Elements for Which AA Has a Detection Limit
Lower than AE by at Least a Factor of 5

Lower Limit of Detection
Element Flame AA, pg/ml Nonflame AA, ng


As


0.1

0.01

0.002

0.5

0.0006

0.005

0.005

0.1

15

0.2

4

0.0003

0.005

0.03

0.5

0.03

0.07

0.1

0.1

0.02

0.1

0.1

0.002

5


0.2 (132)


0.07

0.003

0.02

0.00006

0.007

0.02


(131)

(133)

(131)

(131)

(133)

(131)

(131)

(133)

(135)

(133)

(135)

(131)

(131)

(131)

(135)

(131)

(133)

(131)

(131)

(131)

(131)

(135)

(131)

(135)


0.003

0.02

0.02

0.2

0.06

0.05

0.2

0.003

0.001

0.008

0.5

0.001


(134)

(134)

(134)

(134)

(134)

(134)





(134)



(134)

(134)

(132)

(134)

(134)

(134)

(134)

(134)

(134)

(134)

(134)

(134)











Table 3. Elements for which AA and AE have Comparable
Detection limits

Lower Limit of Detection, ug/ml
Element Technique


0.005
0.03
0.005
0.003
0.1
0.2
2
0.1
0.003
0.002
0.03
1
0.02
2
0.002
1
0.3
0.07
6
1
0.02
0.1
10 -
0.01
3
0.3
0.04


AA
AE
AA-AE
AA
AE
AA
AE
AE
AE
AA
AA
AE
AA
AE
AE
AE
AA-AE
AE
AA
AE
AA-AE
AA
AE
AE
AA
AA-AE
AA


(131)
(135)
(131,136)
(133)
(135)
(135)
(135)
(135)
(135)
(131)
(131)
(135)
(133)
(135)
(135)
(135)
(135)
(135)
(135)
(135)
(105, 108)
(135)
(135)
(136)
(135)
(135)
(135)









Table 4. Elements for Which AE has a Detection Limit Lower
than AA by at Least a Factor of 5


Element


AE Detection


Al 0.005

Ca 0.0001

Ce 10

Cs 0.008

Eu 0.003

Ga 0.01

In 0.002

La 1

Li 0.00000

Lu 0.2

Na 0.0001

Nb 1

Os 10

Sm 0.6

Sr 0.0002

Th 150


Limit, ug/ml



(137)

(336)

(135)

(135)

(135)

(136)

(135)

(135)

3 (135)

(135)

(135)

(135)

(135)

(135)

(136)

(135)









Atomic fluorescence.-AF is the newest of the atomic

spectroscopic methods. According to Winefordner et al. (139),

for a population of atoms in an atom reservoir with low back-

ground the sensitivity of AF measurements is inherently

greater than that of AA because of benefits gained from signal

amplification and increased source intensity. This greater

sensitivity has been applied to serum analysis by a few workers,

such as Kolihova and Sychra (7), who determined Cu. As source

stability and intensity problems are overcome, AF will become

more popular for trace analysis.

Arc-Spark Emission Spectroscopy

Emission spectroscopy is capable of determining up to

70 elements with absolute lower limits of detection down to

10 ng (140). For serum, sample preparation consists only of

dilution if the Stallwood Jet or similar apparatus is used,

or dry ashing if a conventional graphite cup electrode is

used. No separation steps are required. Matrix interference

is not a problem, especially if the cyanogen bands have been

eliminated by using an arc stand with an inert atmosphere.

Besides these advantages, simultaneous multielement determina-

tion is possible, an important consideration when only a

small amount of sample is available for a survey-type determina-

tion. Drawbacks include the difficulty of obtaining quantita-

tive data of even limited accuracy. Photographic emulsions

require troublesome calibration, and tedious and time con-

suming development. Direct readers eliminate this, but allow

observation of only a limited number of wavelengths. These










must be chosen before analysis, thereby limiting the technique

even further. Direct readers also add significantly to the

already large expense associated with emission spectroscopy.

Monacelli et al. (46), and Paixo and Yoe (45) determined

Cr, Cu, Mg, Ni and Zn in plasma using emission spectroscopy.

Nedermieier and Griggs have used a direct reading emission

spectrometer to determine the concentrations of 14 trace

elements in serum (40).

Colorimetry and Fluorimetry

These two techniques can be very simple, sensitive, and

require only inexpensive apparatus under certain ideal con-

ditions. Unfortunately, since hardly any of the organic

reagents used with these techniques are specific for only

one element, interference is a major problem. Complex separa-

tion schemes can be necessary, especially for complex samples

such as serum, thereby removing any benefits from the simplicity

of the technique.

Colorimetry.-Trace quantities of elements in serum

determined colorimetrically include I (78), Br (112), Fe

(81, 82), and Cu (6). In their review of trace element

analysis in clinical chemistry, Schroeder and Nason (117)

include some lower limits of detection for several elements

using colorimetry. These are shown in Table 5. Where the

limit of detection is specified in pg rather than pg/g, the

figure indicates the smallest amount detectable by difference in

two solutions, both with known amounts added (117).










Table 5. Colorimetric Lower Limits of Detection

Element Limit, (jg/g)


Essential


0.006

0.01

1.0

0.05

5.0

0.5


Nonessential

As

B


0.5

0.5

0.1

0.04

0.05

0.25

0.8









Fluorimetry.-Since the lower limit of detection of any

spectroscopic technique is more a function of the signal-to-

noise ratio than signal strength alone, the limits of detec-

tion for inorganic fluorimetry are usually much lower than

those for inorganic absorption spectrophotometry (141). The

limits of detection in interference-free media for several

elements as noted by Winefordner et al. are listed in Table 6.

These limits show that fluorimetry can be comparable to other

sensitive methods of trace analysis if interference is not a

problem. Interference in complex samples such as serum is

greater in fluorimetry than in AE, AA or AF, however.

X-Ray Techniques

X-Ray fluorescence.-Elements with atomic numbers above

15 can be effectively determined at ppm levels by X-ray

fluorescence. Recent reductions in the size of X-ray sources,

lower detector background, and improved resolution and counting

rate capabilities of detectors and associated electronics

have greatly increased the capabilities of the technique (142).

Semiconductor detectors capable of surveying an entire

spectrum of trace elements present at levels below 1 ppm

have been used in serum analysis (143, 144). Analyses of

petroleum (145) and air filter particulates (146, 147) have

demonstrated absolute detection limits down to 10 ng/cm2 for

small particles. These levels cannot be reached with serum

without first concentrating the dissolved solids to a greater

degree than is provided by drying alone. Ion exchange mem-

branes (145) and chelating ion exchange resins (148) provide










Table 6. Fluorimetric Lower Limits of Detection (141)

Element Limit, ig/ml Element Limit, pg/ml


Ag 0.004 Li 0.2
Al 0.0008 Lu 100
As 7 Mg 0.00001
Au 0.5 Mn 0.002
B 0.0005 Mo 0.1
Be 0.00004 Nb 0.1
Bi 0.5 Nd 5
Ca 0.01 Ni 0.00006
Cd 0.02 Os 0.05
Ce 0.05 P 0.0000006
Cl 0.05 Pb 5
Co 0.0001 Pr 0.5
Cu 0.0002 Ru 1
Dy 0.01 Sb 0.05
Er 10 Sc 0.01
Eu 0.005 Se 0.005
F 0.001 Si 0.08
Fe 0.0008 Sm 0.5
Ga 0.001 Sn 0.1
Gd 10 Tb 0.1
Ge 0.004 Te 0.2
Hf 0.1 Th 0.02
Hg 0.002 Tl 0.02
Ho 100 Tm 10
I 0.6 U 0.01
In 0.04 V 2
Ir 2.0 W 0.04
Y 0.02
Zn 0.002
Zr 0.02









both preconcentration and thin geometry. Cobalt determination

at 0.001 ppm (145) and Cr(VI) and/or Cr(III) determination

at microgram levels (147) have been carried out in this

manner.

Quantitative analysis in a matrix as complex as serum

requires complicated mathematical manipulation of the data

to compensate for the X-ray enhancement and absorption effects

of other elements in the sample. These interelement effects

are compensated for by solving sets of simultaneous equations

containing measured constants for each element present (149).

Losses of sensitivity also occur if higher spectrometer re-

solution is required for analysis of complex mixtures. This

becomes even more significant if the high resolution of less

sensitive dispersive systems is required.

Electron probe and laser probe X-ray microanalysis.-This

technique has been applied successfully to the analysis of

small regions in hetreogeneous samples. Its relative lower

limits of detection are not outstanding, but absolute limits

as low as 10-2 pg have been reported (145). Since the mass

of a serum sample after dry ashing and sodium removal is very

low, this technique may be successfully applied.

Mass Spectrometry

Spark source mass spectrometry.-For trace element quali-

tative and semiquantitative analysis, this technique is unsur-

passed in cases where both low limits of detection and a wide

range of elements analyzed are required. Interferences are

not severe, thus allowing lower limits of detection to range









from 0.1 to 0.001 ppm for most elements in any matrix (150).

The small sample size requirement makes this technique

especially useful for serum analysis, where the mass of the

ash to be analyzed is quite low.

The most serious drawback of this technique is its

poor quantitative accuracy. The precision is limited to

about 25% (151). The discharge is unstable, thus preventing

spectrum scanning and requiring photographic detection. Ion

yield is influenced by sample preparation, thus making

comparisons between natural samples and synthetic standards

unreliable. Lack of good quantitative data combined with the

expense associated with purchasing and maintaining the

instrument are reasons why little trace element determination

in serum has been done.

One of the few publications using this technique is by

Wolstenholme (27), who determined 25 elements in plasma. The

reported concentrations deviate from accepted values by

factors as large as 100, thus illustrating the difficulty of

accurate quantitative analysis.

Isotope dilution mass spectrometry.-Much of the uncer-

tainty in qualtitative mass spectrometry mentioned above can

be eliminated by using isotope dilution. No reference standards

are required, and the technique is sensitive and specific.

Isotope dilution allows precise quantitative determination

because only isotope ratios from lines recorded under identical

matrix, exposure, development, and emulsion conditions are

measured. All the uncertainty due to lack of reproducibility

is eliminated.










Early application of the isotope dilution technique

with electron bombardment sources was limited by element

volatility. The technique has been extended to less volatile

metals by Frew et al. (152), who determined Cr as a B-di-

ketonate.

Isotope dilution can also be used with an rf spark source

(153). This involves additional sample preparation, as does

a chelation step. The care required in the tedious steps,

as well as equipment expense keep this technique from being

popular for serum analysis, although some serum analyses have

been reported (154).

Electrochemical Methods

With the exception of potentiometry, electrochemical

methods have not been widely used for trace element determina-

tions in biological media. For example, in a recent Analytical

Chemistry review article which dealt with biological trace

analysis only one article of the 49 reviewed dealt with a

nonpotentiometric electrochemical determination (155). One

reason for this is the susceptibility of electrochemical

techniques to interference. Anodic stripping voltammetry, for

example, can be applied to extremely low concentrations and

is still sufficiently selective to allow separation from

species present in a thousandfold or greater excess (156).

Even so, multielement serum trace analysis cannot be carried

out for elements with similar unmasked electrodeposition

potentials. Because only al.5-vanodic dissolution "window"

exists between the cathodic deposition potentials and the









anodic water-salt breakdown potential limit, one cannot obtain

.separate and distinct dissolution peaks for the large number

of different ions present in serum (156).

Potentiometry with ion selective electrodes has been

applied to serum analysis. Serum calcium, a bulk constituent,

has been determined (91, 92) as has serum fluorine (115). As

sufficiently sensitive ion selective electrodes become avail-

able for more elements, more such determinations can be made.

Czaban and Rechnitz (157) have constructed ion-selective

electrodes with tip sizes in the 100-150 um range. Micro-

electrodes such as these are capable of making in situ

measurements in biological systems. Even though concentration

measurements as low as 10-5 M can be made (157), in situ

serum trace element determination may not be possible due to

the strong metal binding properties of the serum proteins.

The protein complex formation constants must be known and

the extremely low free ion concentration must be measured

before a meaningful concentration figure for a given element

can be obtained.

Coulometry has received limited use in biological trace

element analyses. Selenium in urine (158) as well as chromium

in serum (71) has. been determined. As little as 0.2 pg of

chromium was detected.

Gas Chromatography

Gas-liquid chromatography is one of the most widely used

methods of qualitiative and quantitative organic analysis, as

well as an extremely powerful tool for separation prior to










analysis by other techniques. The chemical similarity of the

compounds being chromatographed poses no problem, since separa-

tion depends only on very small differences in the volatility

and solubility of the compounds in the liquid partitioning

phase. Instrumentation can be inexpensive, yet provide

analyses which would be difficult, if not impossible using

another method. Extraordinary sensitivity, selectivity, ease

of operation, and speed of analysis are some of the character-

istics of gas chromatographic separations. The extremely high

resolution available with long columns is unsurpassed by

nonchromatographic separation methods. Optical isomers,

geometrical isomers, even isotopes have been separated by gas

chromatography.

For several years after its initial development this

powerful analytical tool was used exclusively for organic

compounds. This was due to the lack of suitable volatile,

thermally stable metal compounds. In 1959, Wachi (159) and

Freiser (160) reported gas chromatographic metal chloride

separations, and Duswalt (161) reported separation of metal

acetylacetonates. From this point metal analysis by gas

chromatography has grown to be the well established technique

it is today.

Metal chlorides.-Some metals can be separated by gas

chromatography in the pure state. Sokolov et al. (162) have

separated cadmium and zinc vapor at 850 OC. Such separations

are difficult, requiring severe conditions. The use of vola-

tile metal chlorides allows separation at much lower tempera-









tures. The chlorides of Ni, Sb, Sn, Ta and Ti are sufficiently

volatile for elution. Separations of Sn and Ti, Sb and Ti,

and the gas chromatographic behavior of ten other metal

chlorides have been reported (163). Sie, Bleumer and Rijnders

(164-166) have studied the gas chromatography of 12 elements,

all of which were found to elute satisfactorily. The boiling

points of the chlorides studied are listed in Table 7. All

but two boil below 200 OC and are eluted at 125 oC after

reasonable periods of time (164). Metal halides can require

moderately high temperatures for elution. Their reactivity,

especially their hygroscopicity, presents problems in sample

handling and choosing adequate column packing materials.

Organometallic compounds.-Most classes of volatile

metal compounds are organometallic in nature. Among these

are metal alkoxides, carbonyls, alkyls, hydrides and T-bonded

metal complexes such as cyclopentadienyls (167). Metal

alkoxides, such as Al, Ge, Si and Ti isopropoxides, as well

as Hf, Ti and Zr tertiary-amyloxides, have been eluted with

little decomposition at moderate column temperatures (168).

Most of the rest have proved unsatisfactory for various reasons.

Metal carbonyls, hydrides and alkyls rarely form in quantita-

tive or easily reproducible yields. In addition, solvolysis

during formation and in the gas chromatographic liquid phase

is often encountered with all of the classes mentioned.

Metal chelates.-Of all the various classes of metal

compounds, metal chelates are the most useful for metal

analysis by gas chromatography. This is due to the ease with










Table 7. Boiling Points of Various Chlorides

Boiling
Element Substance point, C


CCl4

SiC14

PC13

POC13

SC12

S2C12

SOC12

SO2C12

TiCl4

VC1 ,

VOC13

Cr02Cl2

Ga2C16

GeCl4

AsC13

SnCl4

SbC13

SbCl5


76.7

57.0

76

105.3

59

138

75.7

69.2

135.8

152

127

117

200

83.1

130

113

221










which they form in quantitative yields and their general lack

of hydrolysis.

There are several ligands whose chelates have been studied

using gas chromatography. The 8-hydroxyquinolate complexes

of Al(III), Cd(II), Co(II), Cr(III), Cu(II), Fe(III), Ga(III),

In(III), Mg(II), Mn(II), Ni(II), Pb(II) and Zn(II) will

sublime at reduced pressure at temperatures between 250 and

500 OC leaving no significant residue (169). However, their

chromatography is complicated by a lack of suitable high

temperature packing.

Benzoyltrifluoroacetone [H(tba)] and thenoyl-trifluoro-

acetone [H(tta)] have been used as ligands with Al, Cr, Cu,

Fe and Ga in successful gas chromatographic separations (170).

Pivaloyltrifluoroacetone [H(pta)] chelates of Y, Sc, and nine

rare earth elements have also been successfully eluted (171).

Metal chelates of the 0-diketones are the best choice

for gas chromatography of metals thus far studied. They form

with quantitative yields, several metals in a mixture being

chelated simultaneously and reproducibly with ease. In most

cases solvolysis and hydrolysis are not significant. Besides

this, fluorinated derivatives yield chelates whose vapor

pressures are low enough to permit gas chromatography without

appreciable decomposition, while taking advantage of the

extreme sensitivity of the electron capture detector for

electronegative elements. Most fluorinated B-diketonates

will rapidly sublime at reduced pressures at temperatures

from ambient up to 100 OC.










Rare earth chelates of several 0-diketone ligands have

been successfully eluted. These include 2,2,6,6-tetramethyl-

3,5-heptanedione [H(thd)] (172), l,l,l-trifluoro-5,5-dimethyl-

2,4-hexanedione [H(tapm)] (173) and 1,1,1,2,2,3,3-heptafluoro-

7,7-dimethyl-4,6-octanedione [H(fod)] (174). All three formed

anhydrous, stable chelates suitable for gas chromatography,

but the fod chelates appear to be the most useful.

The ability of several B-diketones to form efficiently

extracted, completely eluted chelates of Zn and Co has been

studied. The ligands compared were H(fod), 1,1,1-trifluoro-

2,4-pentanedione commonly called trifluoroacetylacetone [H(tfa)],

1,1,5,5,5-hexafluoro-2,4-pentanedione commonly called hexa-1

fluoroacetylacetone [H(hfa)], 1,1,1,2,2,6,6,6-octafluoro-3,5-

hexanedione [H(ofhd)], and 1,1,1,2,2,3,3,7,7,7-decafluoro-

4,6-heptanedione [H(dfhd)]. The order of extraction efficiency

from aqueous solution into benzene was H(fod)>H(dfhd)>H(ofhd)>

H(tfa)>H(hfa). Even though H(fod) extracts more efficiently

than the rest, the fact that H(dfhd) forms chelates much

more rapidly (10 min versus 4 hr for the Zn complex) makes

it more desirable for some applications (175).

Three of the most successful of the fluorinated B-di-

ketone ligands are H(tfa), H(hfa), and H(fod). Their anions

are shown in Figure 1. Their chemistry is discussed below.

H(tfa).-Trifluoroacetylacetone is easy to work with,

forming many stable chelates. H(tfa) boils at 107 OC and is

soluble in most organic solvents. It is slightly soluble in

water. The ease with which tfa chelates are formed is typical







1,1,1,5,5,5-hexaf I uoro-2,4-pentaned ionate


F E00 F
F-C- -C-C~- C-- F

tfa l,I,l-trifluoro--2,4-pentanedionate
F 080 He
F-C- CC'-C-C-H
F H H
fod I,,1,2,2,3,3-heptaf luoro-7,7-dimethyl-4,6-octanedionate

F F F 0E CH3
F- -- 9-F- C-CC- C-CH3
FF CH3
Figure 1. The Anions of the Fluoro-acetylacetone Ligands


hfa









of the 8-diketones. Solvent extraction is applicable to

most samples and yields a chelate solution which can often

be injected into the gas chromatograph without further treat-

ment.

If sample dissolution is required, either nitric or

perchloric acid should be used. This avoids adding anions

which can compete with the chelating agent for the metal ion.

Next the pH of the solution must be adjusted. For some

extractions there is a wide range of pH values within which

the optimum extraction efficiency is obtained. Genty et al.

(176) found 100% extraction of Al(tfa)3 between pH 5 and 9.

Scribner et al. (177) found essentially 100% extraction of

Cu(tfa)2 between pH 4 and 8, and of Fe(tfa)3 between pH 4.5

and 6.1. In a later paper Scribner et al. (178),report 99.9%

extraction for Be(tfa)2 between pH 5 and 7. A sharp rise in

extraction efficiency around pH 3 and a high plateau con-

tinuing to at least pH 7 is typical of tfa chelates of tri-

valent metal ions. The divalent ions of Mg, Mn, Ni and Zn

are all poorly extracted as tfa's in the pH 3 to 7 region

due to hydrate formation.

In some cases the pH adjustment is critical. This is

particularly true where variations in pH are to be used to

carry out the separation of groups of metals prior to gas

chromatographic analysis.

Some metals form insoluble hydroxides which can inter-

fere with chelate formation. Cu can be extracted efficiently

from solutions cloudy with Cu(OH)2, but Fe does not extract









cleanly when initially present as an hydroxide (177). One

way to circumvent this problem when dealing with trace con-

centrations is to keep the pH so low that the solubility

product constant for the hydroxide is not reached. If this

cannot be done within the pH range of efficient extraction,

weak auxiliary completing agents such as acetate ion can be

used. Some neutralization of the acidic H(tfa) is often

required. Acetate salts can thus bring the pH up to a suit-

able level while preventing hydroxide formation. One draw-

back is that if a large variety of ions is being extracted,

some may be sequestered by the acetate ion and not extracted

at all. Sequestration has been shown to occur with ammonia

at pH 9 and higher, as well as with EDTA at pH 5 (177).

Since there is competition for the basic sites on the

ligand anion, complex formation is favored at higher pH. The

pH also affects ligand solubility. The larger the concentra-

tion of ligand in the aqueous phase, the farther the chelate

formation reaction will proceed. This consideration is not

critical with H(tfa), since its distribution between the

organic phase and the aqueous phase can favor chelation at

pH values as low as -0.5, with Do/w = 2.9 (177).

The organic solvent chosen for the extraction should

have a high chelate solubility to avoid chelate precipitation

in the aqueous phase. If an electron capture detector is

to be used for gas chromatographic detection,a hydrocarbon

such as benzene for which detector response is slight can

be used to advantage. Chelation and extraction generally










occur immediately after the ligand-organic solvent mixture is

introduced. Excess ligand and an additional extraction step

are usually helpful in ensuring complete extraction. The

solid chelate can be obtained by evaporating the solvent.

Alternatively, an aliquot of the solution can be injected

directly into the gas chromatograph.

While solvent extraction is the oldest and best established

technique for forming metal chelates for analysis by gas

chromatography, faster and less complex methods are being

developed. Ross and Sievers (179) report a direct chelation

procedure used for the analysis of trace chromium in steel.

A metal particle to be analyzed is placed in a glass tube

with a small amount of H(tfa) and a drop of HNO3 as a catalyst.

Next the sealed tube is heated for 30 min with a microwave

generator. The use of microwave energy stimulates reaction

at the metal-solution interface without overheating the major

part of the unreacted ligand and risking decomposition. The

solution is then dissolved in benzene and the excess ligand,

along with Fe(tfa)3, is removed.

Sievers et al. have studied the suitability of various

metals for chelation and gas chromatographic analysis (180).

Their results are shown in Table 8. Class I is made up of

chelates chromatographed with no evidence of decomposition

in the injection port or column. Sharp, well defined peaks

were obtained. Class II compounds, although eluting with

sharp peaks, left residue in the injection port, indicating

partial decomposition.










Table 8. Metals Suitable for Gas Chromatography


Class I Class II


Al -Hf

Be Mn

Cr Zn

Cu Zr

Fe

Ga

In

Rh

Sc









Metal analysis by gas chromatography has fewer inter-

ference problems than most other techniques because the

analysis process involves separation. Interference in the

analysis of various tfa complexes has been studied. Ross

and Sievers (181) found no interference with ppm Be analyses

from a mixture of seven metal cations, most of which form

stable chelates with H(tfa). A mixture of five anions also

failed to interfere. It has also been shown that quantitative

extraction of some metals, including Be, can occur in the

presence of EDTA (178). If interference at high concentra-

tions should pose a problem, the interferents could probably

be effectively masked by EDTA.

Genty et al. (176) found some metals interfered to a

small extent in the analysis of microgram quantities of Al

in uranium. The interferents were present at concentrations

100 times greater than the Al, yet only errors below 25% were

encountered.

Savory et al. (67) tested various ions for interference

in the analysis of Cr as Cr(tfa)3. The ions tested were As,

Pb, Hg, Li, Mn and Sr at 50 vg/100 ml; K at 40 mg/100 ml;

Ba, Cu, Ni, and Zn at 100 vg/100 ml; Ca at 10 mg/100 ml; and

(H2PO4)-at 5 mg/100 ml. The presence of these ions had no

effect on the Cr determination.

The usefulness of H(tfa) has been shown in many analyses.

A few representative analyses include Al, Fe and Cu determined

simultaneously in alloys (177, 182); Fe and Al (183) and Al,

Ga and In (184) determined using the solvent extraction









procedure; microgram quantities of Al, Ga, In and Be determined

.with an overall relative mean error of 2% (185); and Be, Al,

and Cr (186), as well as Al in uranium, determined at levels

down to 10-11 g (176).

H(hfa).-Hexafluoroacetylacetone boils between 63 and

70 OC and is soluble in most organic solvents. Since H(hfa)

is more highly fluorinated than H(tfa), its chelates have

higher vapor pressures. This allows the hfa chelates to

elute at lower column temperatures, thereby allowing more

thermal instability. Cr(hfa)3, for example, can be eluted

rapidly at column temperatures as low as 30 OC. This tempera-

ture would have to be more than 100 OC higher to elute the

nonfluorinated chelate under similar conditions (187).

Unlike H(tfa), H(hfa) reacts with water to form a tetra-

hydroxy compound, 1,1,1,5,5,5-hexafluoro-2,2,4,4-tetrahydroxy-

pentane. This makes synthesis of hfa metal chelates from

aqueous solutions of the metals much more complicated than

the tfa synthesis and extraction described above. Although

some aqueous hfa syntheses have been reported, they are

probably best avoided. One alternative is conversion to the

metal chloride followed by direct reaction with the ligand

to yield the chelate and HC1. This is more time consuming

and awkward than the extraction procedure.

Hill and Gesser (186) successfully performed quantita-

tive analysis of Be, Al and Cr as hfa chelates. Cr has been

determined as the tfa chelate in the 10-3 10-8 g range (188),








while Cr and Al have been determined at the 10-1 g level

(189). Moshier and Sievers list Al, Be, Cr, Rh and Ti as

having been successfully eluted as hfa complexes without any

trace of decomposition (190). The hfa complexes of Ni, Ta

and Ti have been successfully chromatographed, but only with

great care-to prevent reaction with moisture in the air. It

has been shown that Cu(hfa)2 will deposit elemental Cu on

the walls of a glass tube through which its vapor is passed

at 275 OC. In a hydrogen atmosphere this deposition takes

place at 250 OC. Injection port temperatures should be chosen

with this in mind. On the other hand, Cr(hfa)3 is unusually

stable, resisting decomposition at temperatures up to 375 OC.

It can easily be steam distilled, and distills unchanged from

boiling aqua regia or concentrated sulfuric acid (191).

H(fod).-Heptafluorodimethyloctanedione boils at 33 C

(2.7 Torr) and is soluble in most organic solvents. Its

complexation reactions are similar to those of the other

8-diketones, but it can provide additional latitude in

chelation-gas chromatographic analyses. Where volatile

chelates containing a given metal are formed by several

ligands, the H(fod) and H(tfa) chelates of Cr(III) and Al(III)

being typical examples, the vapor pressures of the H(fod)

chelates are slightly lower than those of the H(tfa) chelates.

However, the primary utility of H(fod) lies in its capability

of forming chelates with metals with which no other ligands

can form useful complexes.

Anhydrous metal complexes which are coordination









saturated with the ligand have been the most satisfactory

for this type of analysis. Hydrate formation causes poor

solvent extraction efficiencies into organic solvents (175).

For example tfa chelates of divalent metal ions such as Mn,

Mg, Zn and Ni, which form hydrates to satisfy coordination

requirements, are extracted only to the extent of a few

percent (177). In addition, the presence of water has been

suggested as the cause of the thermal instability encountered

in many hydrated chelates (172). Hydrolysis at elevated

temperatures is thought to occur.

H(fod) is a much bulkier molecule than H(tfa) or H(hfa).

It forms sterically crowded chelates. This crowding results

in a decrease in the tendency for a complex to form a hydrate,

even though in some cases coordination is incomplete. An-

hydrous Co(II) and Zn(II) bis chelates of H(fod) have been

formed, extracted, and chromatographed successfully, even

though the coordination requirement of six has not been met

(172). This illustrates the usefulness of H(fod) chelates

for the analysis of divalent metal ions. In another study,

the bis fod chelates of Ni(II), Pa(II), Cu(II), and Be(II)

were also quantitatively formed and eluted without apparent

decomposition (192).

The tris chelates of the B-diketones generally have the

most desirable physical properties for gas chromatography.

Rare earth ions tend to accept extra donor groups, easily

forming complexes with coordination numbers greater than six.

Most tris bidentate rare earth complexes therefore are









hydrates (172). These hydrated chelates have been reported

to be thermally unstable (174). Hydrolysis yielding the

hydroxobis diketonate and neutral ligand is suspected to

occur, both at elevated temperatures and in some cases under

vacuum at room temperature (172).

H(fod) forms anhydrous and monoaquo tris complexes with

trivalent rare earth ions. The monohydrated complexes and

anhydrous are interconvertible by a totally reversible process

with no hydrolysis (174). This is believed to be possible

because the water molecule may be hydrogen bonded to an

electronegative site on the ligand, rather than coordinated

to the metal ion (174), due to steric crowding in the complex.

Fifteen tris rare earth H(fod) complexes have been

successfully chromatographed (174). They were found to be

more volatile than any other known compounds of the lanthanide

elements. The synthetic procedure, which involves precipita-

tion of the chelate from aqueous solution, leads to the same

difficulties as the similar extraction procedure described

earlier. The product, a monohydrate, is typically dehydrated

at gas chromatographic elution temperatures and separated in

an anhydrous form (174). Purification by sublimation must be

done carefully, since the chelates decompose at temperatures

slightly above 200 oC (172).

Quantitative extraction of Co(fod)3 from aqueous solutions

at sub-ppm levels has been described (193). The aqueous

sample solution is mixed with portions of NaOH solution and

I1202 solution. After the reaction is complete, the resulting








chelate-benzene solution is decanted and the excess H(fod)

is back-extracted to prevent electron capture detector.

poisoning. The efficiency of extraction was found to be

highly dependent on the order in which the reagents were added,

while variations in pH from 0 to 6.4 had little effect. The

relative insensitivity to pH allows the choice of a pH at

which the extraction efficiencies of potential interferents

are at a minimum. This method has been used to determine Co

in vitamin B12 in liver extracts at levels as low as 4 x 10-11

g (193).

One disadvantage of the fod chelates for solvent extrac-

tion, as compared to tfa chelates, is the longer equilibration

time required. For example, a minimum of 4 hr is necessary

for reaching equilibrium when extracting Zn and Co (175).

This can sometimes be avoided by using direct reaction tech-

niques.

Sievers et al. (192) report an extremely simple direct

chelation procedure for forming fod chelates of Al, Be, Cr,

Fe, Ga, Hf, In, La, Mn, Ti, Sc, V, Zn and Zr. A solid sample

containing the metals to be analyzed is sealed in a glass

capillary with the neat ligand. Heat is then applied briefly

to speed up the reaction before injection into the gas

chromatograph.

A similar direct reaction procedure has been used in

the analysis of Co as the fod chelate (193). Nonaqueous

samples were placed in a small vacuum sublimation apparatus

along with a small amount of HNO3 and the ligand. After








reacting at 104 OC for 4 hr, the excess ligand was removed

and tris-Co(fod)3 was recovered by sublimation. When the

same procedure was carried out at 127 OC, bis-Co(fod)2*2H20

was formed. The reaction temperature is thus critical if

one is to avoid the thermally unstable hydrate.

Instrumentation

Injection.-Metal chelates can be injected into the gas

chromatograph either as solids or in solution. One type of

solid sampler successfully used by Sievers et al. is a bayonet

type which crushes sealed glass melting point capillaries con-

taining the sample in the injection port (192). The solid

sample injector consists of a hollow stainless steel tube with

a plunger which fits inside. To operate the device the plunger

is removed, the sealed capillary is placed in the tube and the

plunger is inserted all the way, breaking the capillary by

forcing it against the beveled end of the device and releasing

the sample in the carrier gas stream. This type of sampler was

used with the sealed tube microanalysis method described by

Sievers (192). Besides applications such as this where the

capillary is also a reaction vessel, the solid sampling tech-

nique is useful where a large amount of chelate is to be

injected, or where the presence of a solvent is to be avoided.

Liquid injection by microsyringe is used more frequently.

It is more convenient than solid samplers, which require that

the carrier gas flow be interrupted before each injection

and that the glass capillary fragments be removed from the

injection port periodically.

Another advantage of liquid injection is that, since the









chelate is already dispersed in the solution before solvent

volatilization, instantaneous vaporization can be achieved

at-lower injection port temperatures than would be required

if the sample were introduced as a crystalline solid (187).

In addition, the last step in many chelation reactions being

solvent extraction, the organic extract lends itself to liquid

injection. No capillary sealing, with the accompanying risk

of sample loss or decomposition, is required. Repetitive

sampling can be made much more quickly and conveniently when

a liquid sample and syringe are used.

Detectors.-The three most popular gas chromatographic

detectors for metal chelates are the thermal conductivity

(TC) detector, the electron capture (EC) detector, and the

flame ionization detector (FID). The TC detector is the

least sensitive of the three, with a detection range for

metals extending from 10-4 to 10-8 g. It is easy to operate

and responds to all compounds. It must be used where non-

destructive detection is required due to effluent sampling for

further analysis.

The EC detector is selective and extremely sensitive to

compounds with high electron affinities such as fluorine

substituted metal chelates. It is relatively insensitive to

hydrocarbon solvents, permitting detection of picogram quanti-

ties of a halogenated compound without interference from the

tail of the solvent peak. Some reported lower limits of

detection with the EC detector are 3.1 x 10-15 moles for Cr,

1.2 x 10"-3 moles for Al, and 2.0 x 10-1 moles for Cu (194);









10-'3 g for Be (181), 10- 4 g for Cr (67, 68), and 10-1 g

for Co and Al (176), all as tfa chelates.

Intermediate between the EC and TC detectors in sensi--

tivity and ease of operation, the FID is an excellent choice

for many metal chelate determinations. Lower limits of

detection with the FID have been reported as 3.7 x 10-1'

moles for Cr(tfa)3, 1.7 x 10- 0 moles for Al(tfa)3, and 1.6

x 10-9 moles for Cu(tfa)2. The linear response ranges from

10-8 to 10-4 g for Cr(tfa)3 compared to 10-11 to 10-9 g for

the EC detector (194). Its operational stability is better

than that of the EC detector, and it is not poisoned by

large amounts of chelate or ligand. In the ppb range the

greater sensitivity of the EC detector is required, but above

5 ppm the FID is the method of choice (194).

Serum application.-The application of metal chelate-gas

chromatographic analysis to serum has been limited. Savory,

Mushak, and Sunderman (68), and Savory et al. (67) have de-

termined Cr concentrations in serum in the vicinity of 40 ng/ml.

They report good recovery of Cr added to serum samples, and

good reproducibility. Chelation and solvent extraction of Cr

in untreated serum yielded no Cr(tfa)3. This was also true of

serum which had been partially wet ashed. In contrast to this,

Hansen et al. (69) determined Cr in plasma at the 50 ng/ml

level using a direct chelation-sealed tube procedure with no

attempt to break down the protein bound Cr complexes by ashing.

Their work, which included live animal experiments to evaluate

chelation efficiency, describes a simple, direct process with

a lower limit of detection around 5 ng/ml of serum.









Taylor et al. (112, 113) have determined Be in bio-

logical fluids using a sealed tube reaction similar to Hansen's

Cr procedure. They, too, found H(tfa) to be an effective

direct chelating agent, forming quantitative chelate yields

in the presence of the naturally occurring completing agents

present in serum, thus avoiding time consuming ashing and

extraction steps. They added Na2EDTA to untreated blood to

prevent clotting and to sequester interfering metals such as

Fe in the blood sample. This enables Be concentrations as

low as 0.02 ug/ml to be determined.

H(hfa) has not been used successfully in serum trace

metal analysis. Hansen et al. (69) in an analysis of Cr

in plasma attempted a direct chelate formation reaction using

H(hfa) dissolved in hexane. This solution was placed with

untreated plasma in a sealed tube and heated. A maximum of

50% of the metal was chelated under optimum conditions. A

white layer at the plasma-hexane interface was assumed to

be the solid dihydrate of H(hfa).

Hansen et al. (69) used a modified direct chelation

procedure to determine Cr in serum using H(fod). Only 50%

of the Cr was chelated, however. H(fod) is a useful com-

plement for H(tfa) in that it forms volatile chelates with

metals with which H(tfa) does not. The lower volatility of

the fod chelates, however, makes H(tfa) the ligand of choice

where possible.

Neutron Activation

Neutron activation analysis (NAA) has been used frequently









as an analytical tool for trace element analysis of serum

samples. The general applicability of the technique at the

trace level is illustrated by the lower limits of detection

listed in Table 9 (116). These figures are based on the

measurement of 40 disintegrations per second in a sample

irradiated at 5 x 1013 n cm-2 s -1 for an optimum time period.

NAA has the advantages of multielement capability, low detec-

tion limits, wide dynamic range, and invariant sensitivity

with the nature, size and concentration of the sample. The

sensitivity is determined by fixed nuclear parameters such

as half-life, cross section and natural abundance, which are

constant under all conditions.

In complex samples, the small photopeak from a trace

constituent is often partly or completely obscured by y-rays

from other radionuclides. If the half-lives of the offending

radioisotope and the radioisotope of interest differ suf-

ficiently, partial photopeak resolution can be obtained by

carefully choosing the duration of neutron irradiation and

decay. This technique is used in nondestructive, or instru-

mental NAA. If this approach cannot be used, or if the photo-

peaks cannot be resolved by a high-resolution detection and

counting system, a chemical separation step must be included.

Instrumental neutron activation analysis.-If the

difference in half-life between the radionuclide of interest

and an interferent can be exploited, separation is not

necessary. Short half-life isotopes are measured using irradia-

tion times too short to allow interferent activity buildup,









Table 9. NAA Lower Limits of Detection

E ppm E ppm E ppm


0.0002

0.0014

0.0016

0.004

6

0.00004

0.01

0.01

0.18

0.006

0.0008

0.02

0.000006

0.001

0.000001

0.008

0.00001

0.0012

0.002

0.006

0.0004

0.006

0.0004


0.000005

0.00008

0.000001

0.1

0.0004

0.1

0.00008

0.00004

0.01

0.002

0.008

0.006

0.04

0.00012

0.002

0.002

0.08

0.0018

0.04

0.0012

0.0004

0.000001

0.004


0.0008

0.0008

0.00006

0.00006

0.1

0.000006

0.001

0.1

0.016

0.0012

0.004

0.002

0.004

0.00008

0.016

0.014

0.012

0.0018

0.0004

0.000016

0.002

0.018

0.0012

0.1









while analyses of longer half-life radionuclides incorporate

a delay after irradiation during which a short half-life

interferent decays to an insignificant level. Long and short

half-life isotopes can sometimes be counted together by

using a long irradiation, a decay period, and a short reir-

radiation followed immediately by counting.

The most pervasive interference in NAA of biological

materials is due to 15-hr 24Na. The parent isotope, 2Na,

has a large natural abundance, a high thermal neutron cross

section, and is present at concentrations from 103 to 105

times greater than the trace elements of interest. These

factors combined with the relatively high energy of the emitted

y-rays prevent the measurement of small amounts of radio-

nuclides in many biological matrices.

A high resolution y-ray spectrometer is essential for

performing multielement instrumental activation analysis. Co-

axial Ge(Li) detector resolution is 10 times better than that

of Nal(Tl) detectors. Planar Ge(Li) detectors have still an-

other order of magnitude of improved resolution, although they

suffer from a lack of sensitivity due to their smaller active

volume. The higher sensitivity of the coaxial detector com-

pensates for its lower resolution, making it the detector

of choice for activation analysis. Although NaI(T1) detectors

are 10-20 times more sensitive than Ge(Li) detectors, this

does not compensate for their extremely poor resolution.

A high resolution multichannel pulse height analyzer

with several thousand channels is needed to take advantage









of the resolution of the Ge(Li) detector when a large energy

range spectrum is being recorded. Where possible, direct

computer interfacing is also desirable in order to facilitate

data handling and reduction.

In spite of interference problems, successful instru-

mental determinations in biological matrices have been

reported. Nadkarni and Morrison (195) described a nondestruc-

tive NAA procedure for trace element analysis capable of

yielding data for 36 elements in certain types of biological

matrices. Irradiation periods ranging from 1 min to 80 hr,

with decay periods ranging from 1 min to 4 weeks were used.

Due to the extremely low concentrations present, however, the

number of elements which can be determined in blood by this

method is not as large (195). Blood analysis with short

irradiation periods (196, 197), and long irradiation periods

coupled with Ge(Li) anticoincidence counting (198) and an

elaborate instrumental setup including a filtered reactor

neutron spectrum and triple coicidence counting with a

56Mn/2Na activity ratio of 700 (197) have been reported.

Yule (199) has published a list of instrumental detection

limits in various matrices. Table 10 lists his estimates

for blood. Comparison with Table 1 shows that most elements

whose serum concentrations are known are present at levels

too low to be detected by NAA without first removing the sodium.

Comparison with Table 9 shows the decrease in detection limit

achieved in blood samples by chemical separation.

Destructive activation analvsis.-Advances in the resolu-

tion of detectors and multichannel analyzer resolution have










Table 10. Limits of Detection in Blood by Nondestructive NAA

E ppm E ppm E ppm


Ag 0.6 Ho 0.4 Re 4
Al I 4 Rh 4
As 4 In 0.1 Ru 2
Au 0.002 Ir 0.006 S 6000
Ba 12 K 1000 Sb 0.08
Br Kr 21 Sc 0.04
Ca 200 La 0.05 Se 4
Cd 500 Lu 0.006 Si 500
Ce 2 Mg 800 Sm 0.04
Cl Mn 0.9 Sn 300
Co Mo 2 Sr 40
Cr 3 N Ta 0.1
Cs 0.4 Na Tb 0.2
Cu 40 Nb 1000 Te 3
Dy 0.06 Nd 6 Ti 200
Er 20 Ne 500 Tl 300,000
Eu 1 Ni 8000 Tm 2
F 30 Os 30 V 0.8
Fe 300 P 1000 W 2
Ga 300 Pb 200,000 Xe 8
Gd 60 Pd 3 Y 10
Ge 50 Pr 8 Yb 0.2
Hf 0.2 Pt 3 Zn 2000
Hg 20 Rb 7 Zr 20,000









greatly reduced the degree of radiochemical purity required

from radiochemical separations. In many cases, a group

separation is all that is required. Nevertheless, high

separation factors are still needed if a very large excess

of interferent is present, as is the case with sodium in

serum.

Meinke (200) has compared pre- and postirradiation sep-

arations. The choice depends on the nature of the sample

and the separation technique involved. Postirradiation tech-

niques permit carrier use and require no blank correction,

while preirradiation separation is necessary whenever the

removed interferent would have resulted in too high a radia-

tion level in the sample.

Ashing.-Most chemical separation procedures require

that all atoms of each species, including the carrier, be in

ionic form, in aqueous solution, and in the same chemical

state. In matrices such as serum, where the trace elements

are combined in complex molecules, complete breakdown of the

sample is required prior to separation. Ashing is thus

considered necessary in most cases.

Trace element losses during wet ashing and high temper-

ature dry ashing have been thoroughly explored (201). The

700-800 oC temperature required for dry ashing in a furnace

results in the loss of volatile metals and metal compounds,

such as chlorides. Wet ashing, while using more moderate

temperatures, introduces impurities via the reagents. The

temperatures reached in a low temperature dry asher of the









rf-excited oxygen plasma type can be kept as low as 120 C.

No reagents are required. The technique is slow, but if

large sample surface areas are exposed ashing can be completed

in a few hours.

Precipitation.-One of the oldest classical techniques,

precipitation is frequently used with NAA. The comprehensive

group separation schemes of Robertson (202), and the analysis

biological samples by Bowen (203), Heydorn and Damsgaard (204),

and Rakovic (205) illustrate its usefulness. Special appli-

cations of precipitation to sodium separation from biological

samples include coprecipitation of sodium on KC104 (206-208)

and precipitation of NaCl in n-butanol saturated with HC1 (209).

In general, precipitation is not selective due to co-

precipitation and absorption. Large quantities of carrier

are required, thus ruling out preirradiation separation and

complicating the use of precipitation with other techniques.

Distillation.-One of the best techniques for obtaining

high purity separated fractions is distillation. Elements

can be distilled as halides, oxides, hydrides, and sometimes

even in the elemental state (210). It has been used in NAA

as a preliminary step in complex schemes (211, 212), for

various biological samples (213, 214), and for trace analysis

in blood (38, 57). Compounds for which distillation is

applicable are also amenable to more powerful separation

techniques such as gas chromatography. Distillation, then,

often does not fully exploit the existing potential for

efficient separation.










Electrochemical methods.-Controlled potential electro-

deposition is a somewhat selective separation technique use-

ful for concentrating elements from large volumes of dilute

solutions. It has been used for preirradiation (156, 215)

and postirradiation (10, 216) separations in urine (216) and

serum (10) analyses. However, since large volumes of serum

cannot be drawn from an individual, the potentially extensive

preconcentration of this technique cannot be achieved. The

selectivity is useful for sodium separation if high decon-

tamination factors can be obtained, but such group separations

for counting with high-resolution detectors are better done

using another technique.

Electrophoresis has been used for NAA separations (217-

220). Copper in serum has been determined this way (220).

Since metals in serum can be separated as their naturally

occurring complexes, this technique may prove to be of great

value in future serum trace element determinations. The

restrictions of microgram sample size and reduced selectivity

associated with electrophoretic focusing do not seriously

affect serum determinations (221).

Dialysis.-During the dialysis of a serum sample all

unbound ions, including all the Na are removed. Protein

bound concentrations rather than total concentration are thus

obtained. Kanabrocki et al. have determined nondialyzable

manganese and copper levels in saliva (222), urine (223),

cerebrospinal fluid (224), and serum (14, 15). Cotzias et al.

(61) have shown that serum proteins will combine with manganese









in the dialysis solution and extract contaminants from the

dialysis bag, thereby yielding results which are too high.

The labile nature of the serum complexes, which can result

in figures being low as well as high, requires that dialysis

results be viewed with caution. Olivares (225) combined NAA

and dialysis to determine the serum concentration of the metal-

containing complex selenocystathionine. In cases such as

this, where it is the complex itself which is of interest,

dialysis is quite useful.

Gel filtration.-Serum trace elements can be separated

as their naturally occurring protein complexes using gel

filtration (226). NAA determinations thus obtained include

protein-bound I (227), and As, Ag, Au, Cr and Sb (228). The

values reported are semiquantitative and do not allow for

complex dissociation due to dilution with eluent.

Szilard-Chalmers effects.-Radiochemical separations using

Szilard-Chalmers reactions take advantage of metal-ligand

bond cleavage in chelates during (n,y) reactions for isotope

enrichment (229, 230). When applied to biological samples,

enrichment factors as large as 10 for elements in dried blood

and liver have been obtained (231, 232).

Separation techniques such as electrophoresis, dialysis,

gel filtration, and paper and thin layer chromatography require

that the complicated protein complexes in which the trace

elements are bound remain intact for postirradiation separa-

tion. Annealing to repair the extensive damage from Szilard-

Chalmers reactions must accompany such separations. Little








work on annealing protein complexes has been published,

although Szilard-Chalmers effects have been noted (233).

The annealing of 6-diketonates has been discussed at length

(234-239). Heat and gamma-radiation do some repair in the

reactor during irradiation, but the greatest metal retention

is obtained by heating after irradiation. Plots of retention

versus annealing time at constant temperature reach plateaus,

the heights of which are temperature dependent (237). At

higher temperatures retention quickly falls off due to decom-

position (237). Szilard-Chalmers reactions, then, are both

useful for some types of separation and a hindrance which

must be compensated for in other separation procedures.

Solvent extraction.-Highly selective extraction procedures

have been reported for use in radiochemistry and other appli-

cations. The systems available are quite numerous. Books

are available which discuss fundamental principles as well

as specific methods (240, 241). Several areas which have

received recent attention may be useful for trace element

analysis in serum. Solvent extraction of metal chelates such

as the B-diketonates discussed earlier, 8-mercaptoquinoline

(22), and thenoyltrifluoroacetone (18, 242-245) is one such

area. Another is the use of liquid ion exchangers such as

di-(2-ethylhexyl)orthophosphoric acid (246). Liquid ion

exchangers combine the speed of solvent extraction with the

selectivity of ion exchange. Direct elemental extraction

without complexation has not received much attention, although

it has been reported (247).









NAA with solvent extraction has been used with biological

samples including serum, with reported sodium decontamination

factors of 104 (248). One major obstacle is that extraction

conditions which allow extraction of a large number of elements

often do not result in adequate sodium removal. For this

reason, where both multielement analysis and high decontamina-

tion factors are needed, solvent extraction is often found

as one of a series of separation steps.

Inorganic separations.-Rapid, selective separations can

be accomplished by passing sample mixtures over columns of

various inorganic materials. Several possible reaction

mechanisms exist: isotope exchange, redox reactions, ion

exchange, and mixed crystal formation by recrystallization.

In reviews of this topic (249, 250), Girardi reports both

large decontamination factors and lower sample capacities

than for organic ion exchangers. A survey of the NAA applic-

ability of over 3,000 adsorption experiments on various ionic

precipitates lists hydrated metal oxides among the more useful

materials (251, 252). Hydrated MnO2 retains numerous ions

completely (253), while hydrated tin oxide, Sn02, is mainly

useful for 32P retention. Hydrated antimony pentoxide (HAP)

in highly acidic systems has been shown to remove only Na

and Ta from a mixture of 60 elements (254). HAP and SnO2 have

been used together for Cr determination in serum by NAA (64).

The high selectivity and efficient sodium decontamination

have been matched by isotopic exchange with a column of NaC1

crystals using an organic eluent (255). The technique is









rapid and selective, with sodium decontamination on the order

of 108 and neglible retention of most other elements. Magnesium

removal from biological matrices with a decontamination factor
3
of 10 using MgNH4PO'46H20 (96) illustrates extension of the

principle where suitable eluents and solid forms of the

element involved are available.

Amalgam exchange with liquid mercury has been used by

Kim and Silverman to determine mercury in various matrices

(256, 257). Mercury has been determined in whole blood using

this technique (258, 259). Amalgam exchange has been used to

remove radioactive Bi, Cd, In, Pb, Sr, Ti, Zn and Cu from

aqueous solutions (260, 261). Equilibrium is attained quickly,

and excellent decontamination can be achieved at trace levels,

making this technique suitable for trace element analysis in

serum. Unfortunately, one can work best with only one element

at a time. Moreover, the technique is applicable to only a

limited number of elements.

Ion exchange.-The use of ion exchange resins in NAA is

well established, with many routine procedures for both selec-

tive individual separations and sequential group separations

having been reported.

Ion exchange material can be divided into four groups:

polymeric resins, celluloses such as carboxymethyl cellulose,

inorganic exchangers such as zirconium phosphate, and liquid

exchangers such as dinonylnaphtalene sulfonic acid (262). Of

these, the resins are the most widely used cation exchange

material, with advantages including high capacity, hydrophilic









gel structure with easily reached functional groups, high

.resistance to degradation by acids and alkalies, and

relatively unchanging ion exchange properties (263). There

are many different kinds and brands of resin available. For

many separations the kind of resin used is not important.

Those with the same functional group, often -SO31, will behave

similarly. One useful property of sulfonic acid exchangers

on a polystyrene base is that these resins are much more

resistant to radiation damage than the anion exchange resins

usually employed (262).

Chelating ion exchangers differ from ordinary ion

exchangers in three respects. The first is higher selectivity

due to dependence on the chelating group rather than ion size,

charge, or other physical properties. The second difference

is that the bond strengths are greater, 15-20 kcal/mole, as

compared to 2-3 kcal/mole for electrostatic bonds in ordinary

resins. The third difference is kinetic: The chelating resin

exchange processes are slower, controlled by either particle

diffusion or second order chemical reactions (264). The

strong bond between the metals and the chelating resins

permits their extraction from very dilute solutions of high

ionic strength (264), such as seawater (265) and dissolved

geological samples (266). Copper has been separated from

biological matrices in this way (267). This property also

allows the extraction of completed metal ions from a complex-

ing medium. Metal ion removal from protein complexes such

as those in serum has not been reported.








There are many excellent publications describing ion

exchange theory and technique. Massart's monograph (262)

deals specifically with radiochemical cation separations and

discusses both the selectivity and kinetics of ion exchange

separations. Resin characteristics such as capacity (263),

particle diameter (268), cross linkage, porosity (269) and the

effect of temperature on HETP (270) have also been treated.

Special techniques dealing with matrix separation in

trace NAA have been developed. Precipitation ion exchange,

which uses the insolubility of matrix metal chlorides such as

NaCI in the 12M eluent,is one such method (271). It works

well for matrices containing large amounts of Na, Ag, Ba, K

and Sr. Another technique for matrix activity removal uses

isotope exchange on a resin in a stable form of the cation.

Breakthrough of the radioactive matrix is delayed while the

radioactive trace elements are rapidly eluted (272).

Ion exchange resins lend themselves to automated multi-

element separations because the eluate from one column can

already be in a form suitable for introduction into the next

column.. Systems using many columns in complex, branched

separation schemes have been described (273-282). Ricq (273)

has explored the use of superimposed columns, while Moiseev

et al. have described a system for determining 27 elements

at trace levels in a silicon matrix (274). Similar systems

designed for biological samples have been developed by Comar

and LePoec (275). Samsahl has described automated multiple

column ion exchange and solvent extraction group separation

systems for biological samples allowing determination of up









to 40 elements (276-278). Speed is emphasized (277-279), as

little as 5 min being required for the processing of four

samples (278). Of the numerous other automatic and semi-

automatic group separation schemes developed, that of Morrison

et al. (280, 281) is perhaps the most extensive and analytically

useful. It enabled measurement of 41 trace elements in the

Apollo 11 lunar samples by NAA after separation into six

groups using ion exchange, solvent extraction, and HAP

extraction. This degree of radiochemical purity is sufficient

with high-resolution counting equipment (282).

NAA with ion exchange separations has been used for the

analysis of many different types of biological samples (225,

265, 283-287). Determination of various elements (5, 38, 57,

83), including I (76, 284-290) using iodine-specific resins

(289) and a sequential group separation scheme for 12 elements

(290) have been reported. Ion exchange membranes have also

been used in biological analyses (291). The characteristics

of ion exchange resins, especially chelating resins, strongly

suggest their application to serum trace element analysis

by NAA.

Paper chromatography.-Various chromatographic techniques

for postirradiation separation in NAA have been reported, but

separation before irradiation without diluting or disturbing

the separated fractions can only be accomplished with paper

as the support. The use of paper chromatography in NAA has

been reviewed by Budzynski (292). Studies involving separ-

ration and activation of metal ions (292-296), complex mole-

cules of biochemical interest (74, 297-302), and possible









interferents in the chromatographic paper (303) have been

reported. The number of elements involved in reported

separations varies, with 7 (293, 294), 26 (295), and

isolation from mixtures of 40 or more (296) being representa-

tive. The techniques used range from ordinary procedures

(77) to the use of columns consisting of stacked ion exchange

paper disks which provide sodium decontamination factors of

104 (304).

Molecular separations in serum analysis offer an additional

dimension for trace element metabolism study. Biological

molecules or chelates of trace elements which lack atoms with

desirable activation characteristics can be determined using

derivative activation chromatography (297-300). Serum mole-

cules such as phospholipids (292) and iodine-containing thyroid

hormones (74, 226), and protein complexes of activating metals

yield quantitative and qualitative data for both the element

activated and the protein complex (292). This is possible

because the unique preirradiation separation capability elimin-

ates adverse effects from Szilard-Chalmers decomposition.

For determinations above the trace level, interference

from impurities in the paper is not significant (74, 303).

The only major contaminants found are Na, Cl, and Mn. These

can be effectively reduced by washing the paper. However,

due to the tubular structure of the paper fibers,all interferents

cannot be completely removed, thus jeopardizing the accuracy

of trace determinations (292). Other drawbacks include

deterioration of the paper in the reactor and lack of resolu-

tion of complex mixtures due to inhomogeneity of the paper.









The advantages of thin layer chromatography over paper

chromatography include greater speed and more selectivity in

the stationary phase. Separation of metal ions has been

reported (305, 306), but the application of thin layer

chromatography to NAA has not been explored, probably due to

the lack of feasibility of preirradiaiton separation.

Gas chromatography.-Gas chromatographic separations are

rapid, highly efficient, and can be carried out without the

use of carriers. In addition, these separations can be opti-

mized with respect to time, resolution, or sample size. This

allows the application of the technique to short-lived isotopes,

multicomponent mixtures and trace constituents. These separa-

tions are very useful in trace analysis since they in effect

result in infinite resolution in the y-ray spectrum. This

allows the lowest limits of detection to be reached using

NaI(Tl) detectors.

Gas chromatography has been used fairly extensively in

radiation chemistry (307), but the application of the technique

to the separation of photon-emitting radionuclides in neutron

activation analysis has been reported infrequently. The

application of gas chromatography to NAA was first reported

by Cram and Brownlee (308-310). They demonstrated the use-

fulness of gas chromatography for the separation of short-

lived radioisotopes in high speed postirradiation separations,

some as fast as 5 s, allowing measurement of 10.7 s 2 F.

The use of metal chelates extends the usefulness of gas

chromatography to encompass many trace elements. As stated









in the Gas Chromatography section, B-diketonates offer the

greatest potential, allowing more than three-quarters of the

elements in the periodic table to be separated and eluted

easily, quickly, and with high resolution. The inherent

speed and selectivity of gas chromatography combined with

the desirable physical properties and ease of synthesis of

the metal chelates make the technique an excellent prospect

for NAA separations.

The use of a high-resolution y-ray spectrometer with a

gas chromatograph constitutes the ultimate in specific

detection for NAA, while the use of efficient NaI(T1)

scintillation detectors allows the lowest limits of detection

to be reached.














CHAPTER II

EXPERIMENTAL

Physical Facilities and Equipment


Nuclear Reactors

UFTR.-The University of Florida Training Reactor (UFTR)

is an Argonaut type, graphite reflected and water moderated,

with 100 kW thermal maximum power. It was used in all the

short irradiation, short decay experiments because of its

close proximity to the separation and counting equipment.

Flux data for the port used in this work appear in Table 11.

GTRR.-The Georgia Tech Research Reactor (GTRR) is a CP-5

type, heterogeneous, heavy water moderated and cooled, fueled

with highly enriched plates of aluminum-uranium alloy. Higher

thermal neutron flux combined with longer daily operation hours

made the GTRR more useful for long irradiations. Flux data

for the ports used in this work appear in Table 11.

Separation Equipment

Low-temperature dry asher.-Ashing was carried out using

a model LTA-600 asher, Tracerlab-Richmond, Richmond, California.

Ashing times of 4-12 hr at an rf power level of 200 W were

used.

Gas chromatograph.-A Varian Model 1200-1 gas chromato-

graph with a hydrogen flame ionization detector, a linear












Table 11. Irradiation Facilities

Georgia Tech
Research Reactor


University of Florida
Training Reactor


Power Level

Cd Ratio

Gamma Flux, R/hr

Thermal Neutron
Flux, n cm-2 s-'


1 MW

2000:1

1 x 106 R/hr

GTRR V-21
2 x 1013

GTRR V-37
7 x 101

GTRR V-43
5 x 101


100 kW

1:1

2.4 x 106 R/hr

Center Vertical Port
1.8 x 1012









temperature programmer, and a 1/4-in injection port was used

in these studies. The system was all glass and Teflon to

eliminate reactive surfaces on which the chelates could

decompose. The column tubing was nominal 1/8-in o.d.

(0.095-in i.d., 0.016-in wall thickness) Teflon (Alpha type,

TFT-250/11). The injection port liner was a 6-mm (approxi-

mately 0.25 in) o.d. by 14.5-cm Pyrex tube with the downstream

end constricted to give a gas tight seal with the 1/8-in o.d.

Teflon column forced into the constriction. A smaller 7.5-cm

by 3-mm o.d. Pyrex liner fit inside the larger tube. One end

of the liner was tapered to fit snugly inside the Teflon

column 5 mm from the packing material. It was this liner

which was removed for y-ray counting.

The instrument was also equipped for injecting solid

samples. For this type of injection the injection port liner

was replaced by a solid sampling device. It consisted of a

stainless steel tube and mating plunger which broke sealed

2-mm by 2-cm capillaries containing the sample inside the

injection port in a flowing He stream (192).

No Swagelok fittings were used. Instead, the Teflon

tubing was forced inside pieces of tapered Pyrex tubing to

give a tight fit. The Pyrex was tapered from both ends,

allowing the two Teflon pieces to touch each other, thus

giving zero dead volume. Liquid injections were made with

a 10-pl Hamilton syringe.

For chromatographic peak area measurements, the electro-

meter output was connected to an Infotronics Model CRS 100










Digital Readout System digital gas chromatographic peak

integrator with Victor printer.

Vapor-phase separation apparatus.-The first vapor-phase

separation apparatus was simply the gas chromatograph with

its column replaced by an empty 6-in piece of 1/8-in o.d.

Teflon tubing. A hole in the oven wall allowed a 10-cm long,

5-mm o.d. Pyrex tube to extend from .the room temperature

atmosphere of the laboratory into the oven, where it was

attached to the empty Teflon tube. Sufficiently volatile

injected samples could thus pass through the system and

condense in the Pyrex tube. A 100-pl syringe with a Teflon-

tipped plunger (Precision Sampling Corp., Baton Rouge,

Louisiana) was used for injection.

The improved vapor-phase separation apparatus did not

use the injection port for sample introduction. Instead, the

6-in Teflon tube in the first apparatus was replaced by a

horizontal shallow "U-shaped" Pyrex tube, 8 cm long, 1 cm o.d.,

with a depression at the bottom to hold the chelate solution

(Figure 2). Both the inlet and the outlet arms were horizontal.

The He flowed through the tube carrying vapor from the mixture

into the cooler Pyrex tube for trapping.

Counting Geometry

Three different types of counting geometry were constructed

and their usefulness in counting chromatographically separated

radioactive metal chelates evaluated. Except for the Pyrex

traps, the effluent reached the counting geometry configura-

tions via a 1/8-in o.d. Teflon tube enclosed in a 1/4-in o.d.
































Figure 2. Vapor-phase Separation Apparatus










copper tube wrapped with heating wire and maintained at the

injection port temperature. The tube passed through the oven

wall and into the lead cave beside the instrument. This

tube, the gas chromatograph, the lead cave and the improved

reversible counting geometry are shown in Figure 3.

Stainless steel spiral.-The stainless steel spiral is a

flow-through type counting geometry. The radioactive'sample

is not trapped; it flows past the detector without stopping.

The counting time in flow-through geometries can be increased

by forming the tube into a spiral. The length of tube in

front of the detector is increased, and the sample is in front

of the detector for a longer period of time. If even more

counting time is required, the carrier flow can be stopped

and the entire separation and elution process temporarily

interrupted. Under most circumstances, this has no deleterious

effects on the separation of components still on the column.

The spiral was constructed from a 57-in long piece of

0.125-in o.d., 0.005-in wall thickness 316 stainless steel

tubing (Superior Tube Company, Norristown, Pennsylvania). It

was wound in a flat spiral with an outside diameter of 2.9 in

and an inside diameter of 1.1 in. The spiral was wrapped as

a unit with heating wire and asbestos tape. A thermocouple

was placed inside at the center of the spiral.

Charcoal traps.-The charcoal traps were constructed

from 10-cm long, 5-mm o.d. Pyrex tube tapered at one end to

fit tightly inside the Teflon tube. A 1-in section of the

trap was packed with 30-50 mesh activated charcoal. The


























Figure 3. Gas Chromatograph and Lead Cave, with Column, Heated Transfer Line,
Reversible Counting Geometry, NaI(Tl) Detector and Exit Line.




















7-


I
I

K^.


.< j


|I


^-.
a-
mzts










inlet to the trap was located inside the column oven. The

rest of the tube extended through a hole in the oven wall.

The outlet and most of the charcoal were at room temperature.

A 5-in piece of thin-walled stainless steel tubing packed with

a 1-in bed of 40-50 mesh activated charcoal was also used.

Reversible counting geometry.-A simple reversible counting

geometry and a more refined version were constructed. The

simple geometry consisted of a 6-in long, 0.25-in o.d. piece

of Teflon tubing packed with 2.5 in of 20% SE-30 coated on

acid washed, DMCS treated Chromosorb P. This tube was enclosed

in a larger aluminum tube wrapped with heating wire. Cool

air could be passed around the Teflon tube inside the aluminum

tube to cool the packing prior to trapping. Purge He could

be introduced at the upstream end to empty the geometry while

it was being heated by heating wire.

The improved geometry, Figure 4, had a better design

than the geometry described above. A 1-in long segment in

the center of a 13-in long, 0.25-in o.d. Teflon tube was

packed with 15% SE-30 on 80-100 mesh High Performance

Chromosorb W (Hewlett-Packard). The packed region was located

at the midpoint of a 1-in radius 180 bend. The center of

curvature of the bend coincided with the axes of the two

cylindrical Nal scintillation detectors. This design allowed

a radioactive sample to give the same detector response at

every point along the bend.








































CO ------


SABSORBANT
TRAP
c-4-



WT



HASTELLOY SLEEVE

1/4" TEFLON TUE




Figure 4. Improved Reversible Counting Geometry









The Teflon tube and packing material were enclosed in a

0.332-in o.d., 0.0125-in wall seamless Hastelloy tube

(Superior Tube Company, Norristown, Pennsylvania). The

Hastelloy had special brass fittings on each end which enclosed

1/4-to 1/8-in unions made of Pyrex. They joined the 1/4-in

Teflon geometry tube and the 1/8-in Teflon inlet and outlet

tubes. The brass fittings joined the Hastelloy and the

1/4-in copper tubes which enclosed the 1/8-in Teflon tubes.

The Teflon was enclosed in metal to ensure uniform heating of

the system to prevent chelate condensation or decomposition.

The metal tubes were wrapped with asbestos-insulated #24 Chromel

heating wire. The entire system was wrapped with asbestos

tape to provide additional insulation and thermal stability.

Carbon dioxide with a pressure of 20 psi could be forced

through a 0.025-in orifice into the space between the Hastelloy

and the 1/4-in Teflon tube at the 1-in packed region to cool

the packing. This created a large thermal gradient along the

packing material. The inlet was operated at 150 OC, while

the outlet was operated at 60 oC. The chelate is trapped by

the 90 OC gradient in the packing. The CO2 was vented through

the outlet arm of the geometry. A special sleeve in the inlet

arm prevented C02 from being forced back and cooling the

inlet. The inlet and outlet heat were controlled independently,

allowing inlet heat for trapping, no heat for y-ray counting,

and both inlet and outlet heat for purging.

The He flow network between the two-stage pressure regu-

lator and the injection port was modified to allow rapid









shutoff of the carrier gas supply to the column for stop

flow experiments. This was done with a normally open

electric valve (Skinner type V15DA1125V09). A normally

closed electric valve (Skinner type V52DA1125V18) and a

precision metering valve (Nupro, Inc.) were installed to

allow bypassing the column with carrier gas to purge the

counting geometry and its connecting tubing with a carefully

regulated He flow.

Counting and Data Processing Equipment

High-resolution y-ray spectrometer.-A 50-cc active

volume, wrap-around coaxial design, high resolution lithium-

drifted germanium detector (Nuclear Diodes, Prairie View,

Illinois) was used. The manufacturers specifications were

as follows: resolution 2.3 keV at 1.33 MeV; peak-to-compton

ratio 23:1; efficiency relative to 3 x 3 sodium iodide

detector 12.5%.

An Ortec Model 456 high voltage power supply (Ortec, Oak

Ridge, Tennessee) was used to bias the detector at 2500 V.

The output from a Nuclear Diodes Model 103 preamplifer and

an Ortec Model 451 spectroscopy amplifier was digitized by

a Wilkinson type analog-to-digital converter, Northern

Scientific Model NS 629 (northern Scientific, Middleton,

Wisconsin). The ADC was capable of 8192 channel resolution

and operation at a 50 MHz clock rate.

The ADC was connected through an interface to a PDP-8/L

minicomputer (Digital Equipment Corp., Maynard, Massachusetts)

with 8K of available core storage. The 12 bit words, when









double precision storage was used in the upper 4K of core,

allowed the use of 2048 channels, each with a 16,777K count

capacity.

Input-output devices used with the computer system

included a Model ASR-33 teletype (Teletype Corp., Skokie,

Illinois), an ITT Model 1935D 15-in display Oscillscope on

which spectra were displayed, a Hewlett-Packard Model 7127A

strip chart recorder with a multiple range input module

(Hewlett-Packard, Pasadena, California) with which the spectra

were plotted, and a Tri-Data Model 4096 magnetic tape cartridge

unit with which all bulk storage, both spectra and software,

was done.

The interface and software for the spectrometer system

were designed and assembled by David B. Cottrell. A more

complete description of the system can be found elsewhere(311).

NaI(Tl) scintillation spectrometer.-Two detectors were

used. Each was a 3-in by 3-in thallium-activated sodium

iodide crystal with mating multiplier phototube (Harshaw

Integral Line Detector, Nuclear Chicago Model 10-9, Nuclear

Chicago Corp., Des Plaines, Illinois) connected to a scintilla-

tion preamplifier (Nuclear Chicago 10-17) and spectroscopy

amplifier (Nuclear Chicago Model 27001). The single channel

analyzer (Nuclear Chicago Model 27352) output was counted

with a preset count six decade electronic scaler (Nuclear

Chicago Model 27104) for gross y-ray measurements.

The y-ray spectra were recorded using a 400-channel

Nuclear Chicago RIDL 34-27 Series Scientific Analyzer System









with a Model 30-35 Four Input Mixer/General Purpose Amplifier,

a Model 23-4 Analog-to-Digital Converter with Buffer Register,

and a Model 52-57 Deluxe Display Control. Also included in

the system were a display oscilloscope (Nuclear Chicago Model

52-56), a Teletype (ASR-33, Nuclear Chicago Model 44-29),

and a five decade preset timer (Nuclear Chicago Model 27101).

Plots of the spectra were made with a Plotamatic 850 X-Y

Plotter (Data Equipment Corp., Santa Ana, CA).

Lead shielding.-All of the 400-channel NaI(Tl) scintil-

lation spectra were measured inside a lead cave with 40-in x

12-in x 14-in inside dimensions and 4-in thick walls on the

top and four sides. The bottom was 2 in thick. The NaI(Tl)

detectors were mounted horizontally with the metal chelate

effluent passing between them.

The high-resolution Ge(Li) detector, having lower sensi-

tivity, required less shielding than the NaI(Tl) detectors.

It was housed in a cave with 12-in x 16-in x 16-in inside

dimensions and 2-in walls on four sides and the top. The

bottom was unshielded.


Procedures

Ashing

Serum samples came from a human serum pool drawn and

processed in the normal course of activities at the Clinical

Laboratories at the J. Hillis Miller Health Center, University

of Florida. All glassware with which the serum came in contact

before irradiation was cleaned by washing with soap and water,










soaking in Nochromix (a metal-free substitute for dichromate

in sulfuric acid, Godax Laboratories, New York, New York) for

12 hr, soaking in concentrated HN03 for 12 hr and then leaching

in demineralized water for 12 hr. The glassware was rinsed

in demineralized water after each step. The serum was freeze-

dried in sample boats in the ashing chambers. This was

accomplished by freezing the serum with dry ice outside the

chamber and reducing the pressure to 1 Torr immediately after

placing the frozen serum in the chamber. Ashing periods from

4 to 12 hr at an rf power level of 200 W were used.

Chelation from Solid Ash

Three serum samples were processed using different

procedures for separation and irradiation. The first sample

was subjected to a simple separation before a long thermal

neutron irradiation. The second sample was processed in a

similar way, except that the separation scheme was more

complex. The third sample was irradiated for a short time

prior to separation.

Irradiation.-The solid residue and supernatant liquid

from the first serum sample were irradiated together in V-21

of the GTRR for 49.5 hr. The solid residue, supernatant

liquid, and blank solution from the second serum sample were

irradiated together in V-21 of the GTRR for 42.5 hr. The

ash from the third sample was irradiated in the UFTR for 30 min.

Gamma-ray counting.-Counting was done with the high-

resolution spectrometry system. The organic layer from the

first sample decayed 26 days before counting. The solid









residue decayed 20 days before counting. The organic layer

and blank from the second sample decayed four days before

counting. The solid residue decayed two weeks before counting.

The third sample decayed one hour before counting.

Chelate Extraction from Ash Solution

After dissolving the ashed irradiated serum, carrier

was added and the chelates of the trace elements in the

aqueous layer were extracted into a benzene solution of the

ligands H(fod) and H(tfa), leaving the sodium in the aqueous

layer. Purification of the metal chelates in the vapor

phase provided still further isolation from sodium. One

simple procedure and three other procedures using a complex

carrier solution were used.

Carrier preparation.-The carrier used in the simple

procedure consisted of 6 mg of Cs(NO3)3, 6 mg of La203,

4 mg of CuCl2, and 2 mg of Sm20a. These salts were added to

the ash as solids. The makeup of the carrier solution used

in the multielement procedures for the other three samples is

shown in Table 12. The weight of each compound in the complex

mixture was chosen to yield 0.01 mmole of element per milliliter

of final carrier solution. The metals and metal salts were

dissolved, mixed, and the mixture was diluted to 25 ml.

Irradiation.-Four 6.5-g serum samples were ashed and

irradiated separately for 30 min in the UFTR.

Chelation.-Each of the four samples was processed

immediately after it was removed from the reactor. The

simple carrier procedure used in the first extraction










Table 12. Complex Carrier Solution


As Hf.

Au Hg

Ba In

Cd La

Co Mn

Cr Ni

Cs Pd

Cu Sc

Dy Se

Eu Sr

Ga Zn










experiment consisted of adding 0.5 ml of a benzene solution

containing 50 mg/ml each of the ligands H(fod) and H(tfa)

to the irradiated ash, stirring, and allowing the mixture

to react for 10 min. The supernatant liguid was decanted.

Next, the ash was dissolved in 0.5 ml of 16M IINO and 0.2 ml

of deionized water, and the solid carrier mixture was added

to the dissolved ash solution. Some solid residue remained

after the addition. The pH was adjusted to 6 with 2M NHOH.

A 0.75-mi portion of the benzene-ligand solution was then

added and the mixture was shaken for 30 sec. After allowing

10 min for the mixture to separate, the organic layer was

removed. The aqueous portion was washed with 0.2 ml of

benzene and all three benzene portions were combined for

separation and counting.

A similar procedure with a complex multielement carrier

solution was used for the next three ashed serum samples.

After irradiation, the sample A ash was dissolved in 0.5 ml

of 2M HNO3, 1 ml of carrier solution was added, and the pH

was adjusted to 6 with 2M NH4OH. One milliliter of the

benzene-ligand solution was added, the mixture was shaken,

and then it was allowed to equilibrate 5 min. After

another shaking and 5 min equilibration the pale blue benzene

layer was removed. The pH was then adjusted to 4, the extrac-

tion was repeated, and the two organic solutions were combined

for counting.

Sample B was processed in nearly the same way. The

organic solution was added before the 1-ml carrier addition










and subsequent pH adjustment to 6 with 2M KOH. After all the

components had been added, the mixture was centrifuged and

the pale yellow organic layer was removed.

The radioactive ash from serum sample C was dissolved in

1 ml of 2M HNOand 1 ml of the carrier solution was added.

The pH was adjusted to 6 with 2M KOH and the metal hydroxide

precipitate was removed by centrifugation. The supernatant

was extracted with 1 ml benzene-ligand solution. The aqueous

layer and the precipitate were then recombined and the pH

was adjusted to 4 with 2M HNO,. The precipitate remained

30% undissolved at this point. The solution was extracted

with 1 ml of the benzene-ligand solution and the two extracts

were combined for counting. The remaining solid was purified

by reprecipitation prior to counting.

Vapor-phase separation.-The extract from the first

experiment was injected into the first vapor-phase separation

apparatus in 30-pl segments 1 min apart. After 0.75ml had been

injected, the oven and injection port temperatures were

increased to 200 C over a 15-min interval. The flow was

then stopped and the charcoal trap was removed for counting.

The next three samples, A, B, and C, were processed

using the improved vapor-phase separation apparatus. The

apparatus was operated for sample A at 100 C for 7 min,

150 C for 5 min, 175 C for 5 min, and 200C for 10 min.

The He flow rate was 20 ml/min. Samples B and C were processed

at 100 C for 3 min, 140 C for 3 min, 170 C for 3 min, and









200 OC for 10 min. The He flow rate was 75 ml/min. The

purpose of the multiple temperature vaporization was to allow

the lower vapor pressure chelates to vaporize without risking

thermal decomposition of the higher vapor pressure chelates,

which leave the oven before the temperature gets too high.

Gamma-ray counting.-Counting was done with the high-

resolution spectrometry system. The radiation from the ash

was measured for 1 min prior to solvent extraction. A 4-MeV

full scale y-ray spectrum was recorded. After extraction, and

before vaporization, a 2-MeV full scale y-ray spectrum of

the benzene solution containing the chelates was recorded.

After separation, several spectra of the charcoal traps con-

taining the radioactive metal chelates were recorded.

Annealing Studies

The extent of Szilard-Chalmers degradation in both

treated and untreated thermal neutron-irradiated benzene

solutions of some metal chelates was determined. The solu-

tions were injected into the gas chromatograph and the y-ray

spectra of the material trapped at the outlet were compared

with the spectra of the material retained by the various

parts of the instrument with which the sample came in contact.

Chelates with no treatment.-Three metal chelate solutions

were prepared. The first contained 32.5 mg of Cf(tfa)3, the

second contained 7.0 mg of Cr(hfa)3, and the third contained

30.0 mg of Cr(fod)3. Each chelate was dissolved in 1 ml of

benzene (Nanograde, Mallinckrodt Chemical Works) which had

been dried over molecular sieve (13X Linde). The three

solutions were irradiated in V-43 of the GTRR for 14.5 hr.









The gas chromatographic columns and conditions used for

retention measurement are listed as No. 1, Table 13. They

were conditioned overnight at 170 C and treated with 10 pi

of DMCS (Dimethyldichlorosilane, PCR, Inc., Gainesville, FL).

The injection port liner was included in the silanization.

The solutions were injected in 5 pA portions 1 min apart.

A total of 100 pl each of Cr(tfa)3 and Cr(fod)3, and 150 ui

of Cr(hfa)3 was injected. After each of the three series of

injections, the carrier gas and oven heat were left on for

30 min before dismantling the system. All counting time

periods were 10 hr except for column measurements which were

15 and 20 hr.

Chelates with treatment.-Three treatment procedures were

evaluated for rechelating the metal ions which had been se-

parated from their ligands by recoil from the complex after

neutron capture. The procedures were irradiation with 10% ex-

cess ligand, irradiation with 10% excess ligand followed by

annealing (heating), and irradiation with 100% excess ligand

followed by annealing. These evaluations were all done with

Cr(tfa)3.

Two solutions, one containing 30.2 mg (Cr(tfa)s and 2.7 mg

of H(tfa) (10% excess) in 1 ml of benzene, and one containing

30.4 mg of Cr(tfa)3 and 27 mg of H(tfa) (100% excess) in 1 ml

of benzene, were irradiated in V-43 of the GTRR for 14 hr.

After irradiation, 20 injections, each 5-pl of the 10%

excess ligand sample, were made into the gas chromatograph.

After dismantling the instrument and measuring the radio










activity of its various components, the remainder of the

sample was resealed, annealed 9 hr at 100 OC, and the experi-

ment repeated. The same injection and counting procedure

was followed with the 100% excess ligand sample af er it had

been annealed 9 hr at 100 C beginning immediately after

irradiation. All chromatographic conditions were the same

as those in the previous section describing chelates without

annealing.

Gas Chromatographic Studies

The extent to which various radioactive metal chelates

of l,l,l-trifluoro-2,4-pentanedione [H(tfa)], 1,1,1,5,5,5-

hexafluoro-2,4-pentanedione [H(hfa)], and 1,1,1,2,2,3,3-

heptafluoro-7,7-dimethyl-4,6-octanedione [H(fod)] were

quantitatively eluted from a gas chromatograph was studied.

Each tfa and fod metal chelate was synthesized from radio-

active metal and dissolved in benzene. The hfa chelates

were purchased. They were irradiated and injected as solids

without any chemical processing or purification. The chelate

solution (or solid sample) was injected into the gas chromato-

graph and the chelate eluted and trapped. After trapping the

chelate, the system was dismantled, and the packing was

placed in a 60- by 150-mm Pyrex petri dish. Separate radiation

measurements of the metal chelate remaining in the solid

sampler, injection port liner, column tubing, column packing,

and charcoal trap were made. Where solid samples had been

used, the quartz capillary fragments in the injection port









were counted and the injection port was rinsed with acetone

and the solution was.counted.

Materials.-The H(tfa) was redistilled and the 106-107 C

fraction retained. The H(fod) was also redistilled. Both

were obtained from PCR, Inc., Gainesville, Florida. The

Cu(hfa)2, Cr(hfa)3, and Mn(hfa)2 were obtained from Pierce

Chemical, Rockford, Illinois.

The radioactive isotopes used were produced in the UFTR

from commonly available reagents,with the exception of the

following, which were purchased from New England Nuclear,

Boston, Massachusetts: 5sCr in 1M HC1, 0.02 mg/ml, 3.7 mCi/ml;

carrier-free 7Be in 0.5M HC1, 4 mCi/ml; and "Fe in 1M HC1,

0.165 mg/ml, 2.42 mCi/ml. The benzene (Nanograde, Mallickrodt

Chemical Works) was dried over molecular sieve (13 X Linde).

The system was silanized using dimethyldichlorosilane (DMCS,

PCR, Inc.).

Chelate synthesis

Cr(tfa)3.-The radioactive metal chelate was prepared by

dissolving 40 mg of CrCl3'6H20 in 0.6 ml of H20. Then 0.9 ml

of 51Cr solution, 300 mg of urea, and 86 mg of H(tfa) were

added. The mixture was refluxed for 7 hr at 100-115 C. The

solid product was washed with water and air dried on filter

paper. It was then dissolved in ether and sublimed at 150 C.

The improved vapor-phase separation apparatus was used for

purifying the chelate. Dark purple Cr(tfa)3 crystals weighing

47 mg were recovered and redissolved in 1.5 ml of dry Nano-

grade benzene for chromatographic injection.









Cr(fod)3.-The radioactive metal chelate was prepared

by dissolving 52.5 mg of Cr(NO3)3*9H20 and 0.3 ml of CrC13

solution containing 51Cr in 11 ml of absolute ethanol and

removing the H20 by boiling until only 0.5 ml liquid remained.

Next 134.2 mg of H(fod) in 1 ml of absolute ethanol was

added and the mixture was heated for 1 hr at 80 C. The

solution was transferred to the vapor-phase separation

apparatus, the ethanol was evaporated, and the chelate was

sublimed at 140 OC. The 6.5-mg purple crystalline product

was dissolved in 0.2 ml of dry Nanograde benzene for chromato-

graphic injection.

Be(tfa)2.-The radioactive metal chelate was prepared

using a procedure similar to that of Berg and Truemper (312).

A 41-mg portion of BeSO4"4 H20 and 90.0 mg of NaAc were

dissolved in 1.5 ml of H20. To this 0.25 ml of 7Be solution

and H(tfa) dissolved in 1 ml absolute ethanol were added.

The Be(tfa)2 was extracted using dry Nanograde benzene.

The product was purified by vacuum sublimation at 65 oC.

The pure radioactive chelate weighed 34.5 mg. It was dis-

solved in 1 ml of dry Nanograde benzene for chromatographic

injection.

Fe(fod)3.-The radioactive metal chelate was prepared

by dissolving 1 mg of FeC13 in ethanol and adding 59Fe

solution and H(fod). The deep red chelate which formed was

extracted into 2.5 ml of hexane. The hexane solution was

placed in the improved vapor-phase separation apparatus, the

hexane was evaporated, and the chelate was sublimed at 140 OC.










The 6-mg deep red product was dissolved in 0.4 ml of dry

Nanograde benzene for chromatographic injection. The lower

melting pale red portion was discarded.

Cu(tfa)2.-The radioactive metal chelate was prepared

using a procedure similar to that of Berg and Truemper (312).

A 140.6 mg portion of CuC12 and 141.5-mg of NaAc were dis-

solved in 10 ml of H20. Next 305.5 mg of H(tfa) in 1 ml of

ethanol was added. The insoluble product was extracted

with dry Nanograde benzene and purified by vacuum sublimation

at 115 C. A 65.5-mg portion of the pure Cu(tfa)z was added

to a mixture of 2 ml of dry Nanograde benzene and 65 mg of

H(tfa). The solution was sealed in an ampoule and irradiated

for 7.5 hr in the UFTR. The radiation damage to the chelate

was annealed at 115 OC for 7 hr before opening the sealed

irradiation ampoule. Next the chelate was purified by vacuum

sublimation at 115 OC. A 54.4-mg yield of radioactive

Cu(tfa)2 was obtained. The pure chelate was dissolved in

1 ml of dry Nanograde benzene for chromatographic injection.

Gd(fod)3.-The radioactive metal chelate was prepared by

dissolving 89.3 mg of Gd(fod)3 87.5 mg of H(fod) in 3 ml of

dry Nanograde benzene. One milliliter of this solution was

irradiated in a sealed ampoule in port V-43 of the GTRR for

12.33 hr. The radiation damage to the chelate was annealed

for 9.5 hr at 100 OC before opening the sealed irradiation

ampoule. Vacuum sublimation at 150 OC yielded a yellow oil

weighing 15 mg. This pure radioactive chelate was dissolved

in 1 ml dry :anograde benzene for chromatographic injection.









Lu(fod) 3.-The radioactive metal chelate was prepared by

sealing 43.5 mg of Lu(fod)3 and 42.1 mg of H(fod) in 1.5 ml

of dry Nanograde benzene in an ampoule and irradiating in

port V-43 of the GTRR for 14 hr. The radiation damage to

the chelate was annealed at 100 OC for 10 hr before opening

the sealed irradiation ampoule. A 33.5-mg yield was obtained

from vacuum sublimation at 115 C. The purified chelate

was dissolved in 0.5 ml of dry Nanograde benzene for chromato-

graphic injection.

Gas chromatography

Cr(tfa)3.-The column, No. 1, Table 13, was conditioned

overnight at 175 OC, treated with 10 pl of DMCS, and con-

ditioned two more hours. Ten 5-pl injections of the radio-

active Cr(tfa)3 solution were made 1 min apart. The carrier

flow and oven heat were kept on for 30 min after the last

injection before dismantling the system for counting. The

chelate solution was injected into three identical columns.

Cr(fod)3.-The column, No. 1, Table 13, was conditioned

overnight at 175 OC, treated with 10 pl of DMCS, and con-

ditioned two more hours. Ten 5-pl injections of the radio-

active Cr(fod)3 solution were made 1 min apart. The carrier

flow and oven heat were kept on for 30 min after the last

injection before the system was dismantled for counting.

Be(tfa)2.-The gas chromatographic conditions were taken

from Ross and Sievers (181). After conditioning column No. 2,

Table 13, for 30 min at 80 OC, twenty 5-il injections of

the radioactive Be(fta)2 solution were made 1 min apart.









Table 13. Chromatographic Columns and Conditions

Column Teflon Tube Liquid Solid He Flow Temperature, OC Chelates
No. Dimensions Phase Support Rate, I.P. Col. Det.
ml/min


1 4 ft, 1/8 in



2 4 ft, 1/8 in


3 7.5 ft, 1/8 in


4 2 ft, 1/8 in


5 6 in, 1/8 in


6 2 ft, 1/8 in



7 3 ft, 1/8 in



8 5 ft, 1/8 in



9 1.5 ft, 1/8 in


5% QF-1 80/100 mesh
AW, DMCS
Chrom. W

5% SE-52 60/80 mesh
Gas Chrom Z

10% SE-30 60/80 mesh
Gas Chrom Z

5% SE-30 100/200 mesh
Chrom. W

10% SE-30 80/100 mesh
Chrom. W


20% High
Vacuum
Grease


100/120 mesh
AW, DMCS
Chrom. P


15% QF-1 80/100 mesh
AW, DMCS
Chrom. W


20% High
Vacuum
Grease

20% SE-30


100/120 mesh
AW, DMCS
Chrom. P

100/120 mesh
AW, DMCS
Chrom. P


60 160 135 (150)
120 100
170 150 (170)


50 170 80


30 205 170


60 170 100


30 230 170

75 170 150



20 165 100


Cr(tfa)3
Cr(hfa)3
Cr(fod)


Be(tfa)2


Fe(fod) 3


Cu(tfa)2

Lu(fod)
Gd(fod) 3

Cu(hfa)2

Cr(hfa)3

Mn(hfa)2


30 170 150 (170) Cu(hfa)2



30 150 150 (150) Cu(hfa)2




Full Text

PAGE 1

ACTIVATION ANALYSIS OF TRACE ELEMENTS IN SERUM By FREDERICK TURNER VARCOE , JR. A THESIS PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1974

PAGE 2

TO NILA

PAGE 3

ACKNOWLEDGEMENTS The author would like to thank his research director, Dr. Stuart Cram, for suggesting the research problem, for valuable advice, suggestions, and support, and for a wealth of practical experience. Thanks also go to the other members of the supervisory committee. Dr. Roger Bates, Dr. James Winefordner, Dr. Robert Hanrahan, and Dr. John James for their confidence and help. The project could not have been completed without them. Thanks also go to the machine shop personnel, in particular to Mr. Arthur Grant, Mr. Chester Eastman, and Mr. Frank Ottinger for their expert fabrication of many pieces of equipment used in this work. The glassblowing skills of Mr. Truman Robbins, Jr., are also greatly appreciated. The reactor operators and other personnel, in particular Mr. Henry Gogun at the University of Florida and Dr. Milton McLain, Jr. , at Georgia Tech deserve heartfelt thanks for providing the irradiations necessary for these experiments. The author also appreciates the contributions made by the other members of his research group. Dr. David Cottrell deserves special recognition for the design and assembly of the computer interface and software with which all the high-resolution gammaray spectra v/ere recorded. Dr. Jack Leitner's expert help v/ith the operation of the system and his encouragement and advice are also appreciated. The author is deeply grateful to his parents for their support, encouragement, and sage advice throughout his education. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii ABSTRACT IX CHAPTER I Introduction. Trace Analysis 1 Trace Elements in Serum 1 Elements Identified 2 Function 2 Disease Relationships Established 7 Heart disease 7 Cancer 8 Other diseases , 9 Serum Trace Element Analysis 10 Flame Spectroscopy 11 Atomic absorption 13 Atomic emission 14 Atomic fluorescence 18 Arc-Spark Emission Spectroscopy 18 Colorimetry and Fluor imetry 19 Colorimetry 19 Fluorimetry 21 X-Ray Techniques 21 X-Ray fluorescence 21 Electron probe and laser probe X-ray microanalysis 23 Mass Spectrometry 23 Spark source mass spectrometry 23 Isotope dilution mass spectrometry 24 Electrochemical Methods 25 Gas Chromatography 2 6 Metal chlorides 27 Organometallic compounds 28 Metal chelates 28 H(tfa) 31 H (hfa) 38 H(fod) 39 Instrumentation 4 3 Injection 43 Doctors 44 Serum application 45 Neutron Activation 46 Instrumental neutron activation analysis 47 Destructive activation analysis 50 Ashing 52 IV

PAGE 5

Table of Contents Continued, Page Precipitation. 53 Distillation 53 Electrochemical methods 54 Dialysis 54 Gel filtration 55 Szilard-Chalmers effects 55 Solvent extraction 56 Inorganic separations 57 Ion exchange 58 Paper chromatography 61 Gas chromatography 63 CHAPTER II Experimental 65 Physical Facilities and Equipment 65 Nuclear Reactors 65 UFTR 65 GTRR 65 Separation Equipment 65 Low-temperature dry asher 65 Gas chroraatograph 65 Vapor-phase separation apparatus 68 Counting Geometry 68 Stainless steel spiral 70 Charcoal traps 70 Reversible counting geometry 73 Counting and Data Processing Equipment 76 High-resolution Y~i^sy spectrometer 76 Nal (Tl) scintillation spectrometer 77 Lead shielding 78 Procedures 78 Ashing 78 Chelation from Solid Ash 79 Irradiation 79 Gamma-ray counting 79 Chelate Extraction from Ash Solution 80 Carrier preparation 80 Irradiation 80 Chelation 80 Vapor-phase separation 83 Gamma-ray counting 84 Annealing Studies 84 Chelates with no treatment 84 Chelates with treatment 85 Gas Chromatographic Studies 86 Materials 87 Chelate synthesis 87 Cr (tfa) 3 8 7 Cr (fod) 3

PAGE 6

Taole of Contents Continued. . . Page Be (tfa) 2 88 Fe (fod) 3 88 Cu(tfa)2 89 Gd (fod) 3 89 Lu(fod) 3 90 Gas chromatography 90 Cr (tfa) 3 90 Cr (fod) 3 90 Be (tfa) 2 90 Fe (fod) 3 93 Cu(tfa)2 93 Gd (fod) 3 93 Lu (fod) 3 94 Cu (hfa) 2 and Cr (hf a) 3 94 Mn (hfa) 2 94 Chelate residue exchange 94 Silanization 95 Counting geometry operation 96 Stainless steel spiral 96 Charcoal traps 96 Simple reversible geometry 96 Improved reversible geometry 97 Ion Exhange 97 Postirradiation separation 98 Sodium separation optimization 98 Trace element determination 98 Preirradiation separation 98 Short irradiation 98 Resin blank 99 Solution blank 99 Serum solution 99 Irradiation and counting 100 Long irradiation . ." 100 CHAPTER III Results and Discussion 101 Objectives 101 Chelation Studies 102 Chelation Procedure 109 Chelation from solid ash 110 Advantages 110 Procedures 110 Preirradiation separation Ill Postirradiation separation 121 Extraction from solution 129 Advantages 129 Procedures 129 Carrier Preparation 133 Additional Separation 139 Interf erent Removal 142

PAGE 7

Table of Contents Continued... Page Annealing Studies 145 Untreated chelates ' 145 Treated chelates 147 Gas Chromatography 148 Incomplete Elution. 148 Residue measurement 148 Injection procedures 151 Counting Geometry 15 3 Stainless steel spiral 158 Charcoal trap 163 Reversible counting geometry 163 Ion Exchange 170 Postirradiation Separation 170 Sodium separation optimization. 170 Serum analysis , 173 Preirradiation Separation 173 Short irradiation 173 Long irradiation , 174 Conclusions 179 Sequence of Operations 179 Preirradiation separation 179 Postirradiation separation 179 Separation Techniques 181 Chelation 181 Chelate separation 181 Ion exchange 183 Future Work 183 REFERENCES 185 BIOGRAPHICAL SKETCH , 203

PAGE 8

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ACTIVATION ANALYSIS OF TRACE ELEMENTS IN SERUM By FREDERICK TURNER VARCOE , JR. June, 1974 Chairman: Roger G. Bates Major Department: Chemistry This research deals with radiochromatographic separations in neutron activation analysis of trace elements in serum. Understanding the metabolic roles of trace elements as well as their toxicological effects in biological systems has been limited by the sensitivity and selectivity of existing analytical techniques. The lack of human serum concentration data for many trace elements prevents optimum clinical utilization of this extremely powerful diagnostic tool. Neutron activation analysis has extremely low limits of detection for a large number of elements. However, interference from those elements present in human serum at relatively high levels must be eliminated before even a highresolution solid state detector can be effective. The utilization of high sensitivity sodium iodide detectors requires still more separation. Vlll

PAGE 9

Chromatographic separations of metal chelates were developed for both preand postirradiation separations. Quantitative elution from the gas chromatographic column for fluorine substituted beta-diketonate complexes of Cr, Cu, Mn, Fe, Be, Lu, and Gd was studied; recoveries between 52 and 92% were found. Extensive studies of decomposition and adsorption losses in the system are reported. Construction and evaluation of various sampling systems, transfer lines, and counting geometries are included. A completely reversible counting geometry for radioactive metal chelates using wide-range temperature and carrier gas control to trap and subsequently release eluate fractions is described. Also included is an apparatus for carrying out chelate separations from solutions whose large volumes preclude gas chromatographic injection. The successful use of a chelating ion exchange resin in human serum neutron activation analysis is reported. Operating conditions which optimize sodium removal with respect to sample loss are described.

PAGE 10

CHAPTER I INTRODUCTION. TRACE ANALYSIS Trace Elements in Serum Most of the elements in the geochemical environment can be detected in animal and plant tissue if a sufficiently sensitive technique is available (1) . Elements are present in bulk and trace concentrations. The trace elements, usually considered to be those present at concentrations below 100 ppm, can be divided into two groups: those v/hich serve an essential biochemical function without which subnormal functioning or illness results, and those which are nonessential. Abnormally low concentrations of an essential element can cause a deficiency disease, while the lack of a nonessential element has no ill effects. Excess of either essential or nonessential elements can be toxic, however. As our knowledge of the biochemistry of trace metals grows, more elements may come to be considered essential. It is already known that many of those not now considered essential have very important roles in metabolism. The concentrations of essential trace elements within the organism are maintained within very narrow limits, which is one reason why so much attention has been given to the relationships of these elements to normal and abnormal metabolic functions (2). Normal concentration ranges for nonessential

PAGE 11

elements also exist, but are much broader than those of essential elements. The trace elements must be present in the diet. Their only sources are the earth's crust and the ocean. Life cannot exist without them (3) . Elem.ents Identified Those elements considered essential for healthy human bio chemical functioning (1,4) are listed in Table 1, along with serum concentrations which have been found in normal human subjects. Also included are elements considered nonessential environmental contaminants, often having toxic effects. Reports dealing with elements whose serum concentrations are relatively high, or which for other reasons can be determined using less sophisticated analytical techniques, have existed for some years. Among these are Cu (5-48), Zn (19-59), Mn (14-27, 56, 59-62), Cr (13, 41-44, 57-71), I (47, 48, 58, 7279) and Fe (11, 24-27, 40-42, 56, 58, 80-83). Recently new techniques have allowed identification of as many as 14 elements simultaneously in blood and serum samples (26, 40, 53), while V (84), Ba (85), Co (86), Sr (85, 87), U (88, 89), Ca (90-95), Mg (95, 96), Hg (97-99), As (99), Pb (100-102), Cd (101), Se (103-105), Au (106-108), Te (109), Li (110), W (111), Be (70, 112, 113), Br (114) and F (115), have been determined individually. Several review articles have been written on trace element analysis in biological systems (116, 117, 118) . Function The importance of trace elements lies in their roles in

PAGE 12

Table 1. Essential and Nonessential Elements in Human Biochemistry Element Serum Concentration Reference yg/ml Essential Bulk Ca 2.8 (116) CI 1,000 (116) K 1,700 (116) Mg 13 (116) Na 3,250 (116) P 0.50 (116) S 5,400 (27) Essential Trace Co 0.00018 (117) Cr 0.02-0.05 (69) Cu 1.16 (117) F 0.028 (117) Fe 1,14 (117) I 0.0029-0.11 (117) Mn 0.0083 (117) Mo 0.4 (117) Ni 0.0042 (117) Se 1.1 (117) Si 7.0 (27) Sn 0.033 (117) V 1.0 . (117) Zn 0.98 (117)

PAGE 13

Table 1. continued... Element Serum Concentration Reference yg/ml Nonessential Ag 0.28 (116) Al 5 (27) As 0.19 (117) Au 0.03 (116) B . 2 (27) Ba 0.059-0.066 (116) Be <0.004 (117) Br 830 (116) Cd <0.1 (117) Cs 0.05 (27) Hg 0.0027-4.0 (116) La 18 (116) Li 0.031 (117) Pb 0.046 (117) Rb 14 (27) Sb 0.054 (117) Sc 0.04 (116) Sr 0.057 (117) Th 0.4 (27) U 0.048-0.072 (116)

PAGE 14

biological functioning. They take part as active components of enzyme systems, in transport mechanisms, in tissue synthesis, and in bone formation (2). The concentrations of trace elements in marine organisms are fixed by the concentrations of trace elements present in the surrounding water. In terrestrial animals, highly developed homeostatic processes are required to maintain trace element concentrations at fixed levels in the face of widely varying amounts in the environment. When, for one reason or another, an essential trace element's concentration lies outside the normal range, a serious metabolic upset must also be present. Trace element concentration variations have been implicated in diseases ever since it became known that these elements were an integral part of metabolism. It has become obvious that knowledge of the function of trace elements in metabolism should help the diagnosis and treatment of disease (2). Most diseases of any consequence are accompanied by changes in the concentrations of one or more trace elements in some tissue or body fluid. However, it must be kept in mind that only a few diseases have abnormal metal metabolism as a primary underlying cause, even though correcting the imbalance will sometimes alleviate the symptoms, e.g. , oral Cr dosage will improve abnormal glucose tolerance, although abnormally low serum Cr probably does not cause diabetes (119) Even though many changes in the normal tissue concentration of a trace element are nonspecific indications of an abnormal

PAGE 15

condition, the analysis of body fluids and tissues for trace elements is becoming an important clinical test in humans. Blood is both the easiest to obtain for analysis and the most pervasive of the body's tissues. It travels everywhere in the body, providing an exchangeable reservoir of elements and compounds for all the other tissues. The component of blood which is most useful for trace element studies is the serum which remains after the blood cells and clotted fibrin have been removed by centrifugation. Only a small fraction of the trace elements in serum are present as free ions. Most are in the form of complexes. They may be loosely bound to a ligand such as albumin, or tightly bound as in the iron-porphyrin compounds (2, 120). This provides a high degree of selectivity in the combinations available for metal complex formation (2, 120), and thus gives the body the means to select and accumulate the required metals from circulating body fluids. These complexes assist homeostatic mechanisms in maintaining constant trace element concentrations by forming or dissociating when necessary, Also, heavy metals can be maintained in the system at nontoxic levels while still providing the mineral reservoir needed for normal metabolism. The critical balance between metal ion and metal complex is easily upset by any of several factors which affect the stability constants of the complexes. Today scientists need to learn more about the biochemical phenomena that depend on these metals and more about the stability constants and specificity of the biochemical ligands. The serum concentrations of trace elements have the broadest

PAGE 16

range of diagnostic applicability and therefore are of the most interest at this time from the standpoint of improving diesease diagnosis and etiological knowledge, and understand ing trace metal metabolism in the human body. Disease Relationships Established Heart disease . -Heart disease is responsible for nearly half of the deaths in the United States each year. For this reason, it has received much study. Recently the possibility that trace elements play a part in cardiovascular diseases has become evident. Studies exploring this possibility show how trace element concentration information can play an important part in understanding and diagnosing a disease. Injured heart tissue from patients who die from heart attacks (acute myocardial infarction) show abnormally low concentrations of Co, Cs , K, Mo, P, Rb and Zn and elevated concentrations of Br, Ca , Ce , La, Na, Sb and S (121). Serum levels in patients with myocardial infarction are also abnormal. Increases in serum Cu (11, 14, 15), Mn (14, 15), Ni, Mo and B (122) have been found, as v/ell as a decrease in serum Zn (2, 123). Serum Cu determination has been suggested as a diagnostic test for patients with myocardial infarcation (11) . Trace element intake in drinking water has been linked with death from myocardial infarction. A highly significant negative correlation has been found between water hardness expressed as Ca ion concentration and the death rate in the local population due to degenerative cardiovascular disease in several countries, including the U.S. (124-126). Calcium

PAGE 17

may only be a benign indicator of the presence or absence of another trace element v/hich takes an active part in degenerative cardiovascular disease. For example, Ca is thought to inhibit the transfer of divalent metals across intestinal membranes (93) . Perhaps this inhibition controls the presence of a trace element which takes an active part in the disease. Serum cholesterol has been linked to cardiovascular disease. The presence of trace quantities of V and Fe decreases cholesterol synthesis, while Cr and Mn have the opposite effect (127). Administration of CaNa2EDTA has been shown to reduce serum cholesterol and to increase urinary output of certain trace elements, especially Zn (123) . Cancer . -Trace metal studies have also been made in connection with cancer, another major cause of death in the U.S. Brune found no variation in the serum concentrations of 12 elements in leukemia patients (26) , while other v/orkers have found abnormal Fe (80) and elevated Zn (52) concentrations. High serum Cu concentrations are present with breast (12) and prostate (15) cancer, chronic and acute leukemias (2) , and Hodgkin's disease (2). Hrgovcic as cited by McCall (2), gives an example of how determination of trace element levels can aid in evaluating the treatment of disease. In cases where malignancies are being treated, changes in the serum Cu concentration can appear long before there are any clinically apparent changes. Studies attempting to correlate the death rate from cancer with various types of trace clement environments in

PAGE 18

different geographical areas have been made. Berg, in his study of carcinogenic trace metals in water supplies, states that although the literature contains some speculation, almost no convincing data exist with respect to these correlations (128). Pories, hov/ever, states that the large differences in death rate from cancer in different parts of the U.S. suggest that environmental factors may be the principal and preventable causes of cancer (129) . He also states that ' elements can inhibit or foster tumor grov;th. Selenium and platinium interfere with tumor growth; iodine prevents thyroid neoplasia. Zinc metabolism may be the key for limiting tumor growth by manipulating necessary protein synthesis (129) . Other diseases . -The relationships between serum trace elements and other diseases have also received attention. Protein bound iodine determination in serum has been used for diagnosing abnormal thyroid activity for some time (72-79). Abnormal serum Cu and Zn concentrations have been linked with rheumatoid arthritis (31). In sera from arthritic patients Cu, Ba, Cs, Sn and Mo concentrations were elevated, Zn , Fe and Pb concentrations were depressed, and Al , Ni , Sr, Cr and Cd concentrations were normal (40). Hepatitis and liver cirrhosis are accompanied by elevated serLUTi Zn and Iln (15, 49, 52). Serum Cr is depressed and Mn is elevated in individuals suffering from diabetes mellitus (15, 119). Oral Cr supplementation will improve such patient's glucose tolerance (15). Zinc is required by more enzyme systems in the human body than any other metal (2) . In general, decreases in

PAGE 19

10 serum Zn are accompanied by increases in serum Mn, although this is not true in every case. Their antagonism has been noted in lymphoma, pneumonia, and other diseases (2) . Low Cu levels have been measured in patients , with Wilson's disease (9), chronic and acute leukemias, aplastic anemia, cirrhosis, viral and microbial infections and nephritis (2). Elevated serum Zn accompanies psoriasis (30) . Abnormal trace element concentrations have even been correlated with mental illness. Serum Zn and Cu concentrations have been measured in schizophrenics. Both are needed for histaminase enzymic action, variation of which has been associated with schizophrenia. The Cu concentration was found to be abnormally high, but returned to normal following effective therapy (33) . Zinc and copper are antagonists in the human body. Oral Zn treatment has been used successfully with such patients, apparently forcing the Cu concentration down. Penicillamine, a chelating agent which removes trace metals from the body, has also been used for effective treatment (33) . Data from trace element studies using new analytical techniques can provide the basic information necessary for a better understanding of the roles of trace elements in diseased and healthy metabolic acitivity. These techniques can then be used to obtain the necessary information for individual diagnosis. Serum Trace Element Analysis Research on the metabolic roles and toxicological effects of trace elements in biological systems is highly dependent

PAGE 20

11 on the development of new analytical techniques which are reproducible, rapid, sensitive, specific, and suitable for automation. The lack of such methods has thus far impaired development of the vast potential of serum analysis for diagnosis and treatment in the medical sciences. Many instrumental methods for elemental analysis in serum have been reported, but the n-umber well suited to trace analysis is small. The techniques reported have been limited in scope by several considerations. Matrix interference is severe. The Na concentration is 2-3 mg/ral , lO"* times greater than even the higher trace element concentrations. With five or six other elements present at macro levels (Table 1) , determining concentrations for the large number of trace elements present at ppm to sub-ppm levels requires advanced techniques, especially if highly desirable simultaneous multi-element analyses are to be achieved. The lower limits of detection presently available have undoubtedly limited the scope of previous determinations. The more popular trace element techniques, their advantages, shortcomings, and potential for serum analysis are discussed below. Flame Spectroscopy Flame spectroscopy is the most sensitive of the techniques requiring only moderately expensive equipment. More than 60 elements, including all the metals, can be determined this way. Interference is less likely than when using many other trace analytical techniques. Chemical interferences, those v;hich prevent the atoms from reaching the free, dissoci-

PAGE 21

12 ated, un-ionized state necessary for measurement in the flame, and spectral interferences due to bands emitted by the OH and CN radicals can present problems. 'Spectral interferences from other trace elements in the sample are seldom encountered. However, the total salt concentration in the liquid being analyzed should be kept below 1%, which can require sample dilution and adversely affect trace determinations (130) . Tenfold dilution of serum samples has been found necessary by some workers to avoid clogging the burner (32) . A solvent extraction step prior to analysis can eliminate the need for dilution, although the risk of contamination is greatly increased. VJhatever separations are required prior to determination are usually simple (131) . Probably the main disadvantage flame spectroscopic techniques have is that except in the case of atomic emission, only one or two elements can be determined at a time. Research is currently under way to remedy this, but until these systems become fully operational either another technique must be used or one must make a separate series of measurements for each element. Atomic absorption (AA) , atomic emission (AE) , and atomic fluorescence (AF) are complementary spectroscopic techniques. When the limit of detection must be optimized, AA and AF are generally preferred for elements with resonance lines belov; 300 nm, since there is usually insufficient energy available in a properly adjusted flame to excite these higher energy transitions. For elements v/ith resonance lines above 400 nm,

PAGE 22

13 there is adequate flame energy available and an external radiation source is not required. Elements whose resonance lines lie between 300 and 400 nm can usually be determined equally well by any of the three techniques (130) . Atomic absorption . -Those elements for which AA has a lower limit of detection than AE by at least a factor of 5 are listed in Table 2. The classification is based on comparison, as flame techniques, without regard for the fact that in some cases v;here sample size is limited, and in other cases where the nonflame technique is unusually sensitive, a lower limit of detection may be obtained with nonflame AA. Table 3 lists those elements for which AA and AE have limits of detection which are comparable. Together these lists contain 51 elements. When Table 1 is compared with Tables 2 and 3, it can be seen that of those essential trace elements which can be determined spectroscopically all have AA detection limits equal to or lower than the corresponding AE limit. V7ith this in mind, it is not surprising that AA has been the most popular spectroscopic technique for trace element analysis of serum. Of course, classification as essential requires an adequate method for detecting the presence of the element in question, and as other techniques are applied to trace element analysis of biological samples the essentiality of other elements may be discovered at levels too low for analysis by AA. A few representative determinations of trace elements in serum by A?, include Ca (90, 94); Fe, Cu and Zn (32, 34,

PAGE 23

14 50); Au (108); Li (110); and Pb (77). Two of these references, Au (108) and Li (110) , describe metal determinations in sera of patients currently undergoing therapy with these metals and therefore involve abnormally high concentrations. Techniques which require only microliter samples are being used more frequently for serum, analysis. The Delves cup technique has been applied to serum chelate extracts of Pb and Cd (102) , and the carbon rod atomizer has been used to determine Mg, Fe , Cu, Pb and Zn concentrations in serum (35). These techniques are especially valuable v\?here multielement determinations are required, since until simultaneous multielement spectroscopic determinations are improved, a separate portion of the sample will have to be used for each element determined. Atomic emission . -Those elements for which AE has a lower limit of detection than AA by at least a factor of 5 are listed in Table 4. None of the essential trace elements appears on this list, although several of the bulk essential elements do. Tables 3 and 4 combined list 43 elements, including five essential trace elements, for which the AE limit of detection is at least as low as the AA limit of detection. AE , then, can be a valuable flame spectroscopic tool in serum trace metal studies, both because it is best for several elements presently knov/n to take an active part in metabolism, and because it can be used for studying elements for which AA is not adequate. Besides this, it may also provide a basic technique for multielement analysis. For example, eight elements have been determined in blood using plasma torch excitation of an emission spectrum (138).

PAGE 24

15 Table 2. Elements for Which AA Has a Detection Limit Lower than AE by at Least a Factor of 5 Element As Au Be Bi Cd Co Fe Ge Hf Hg Ir Mg Ni Pb Pt Rh Sb Se Si Sn Te Ti Zn Zr Lower Limit of Detection Flame AA, yg/ml Nonflame AA,

PAGE 25

16 Table 3. Elements for which AA and AE have Comparable Detection limits Lower Limit of Detection, yg/ml Element Technique Ag Ba Cr Cu Dy Er Gd Ho K Mn .Mo Nd Pd Pr Rb Re Ru Sc Ta Tb Tl Tm U V w y Yb 0.005

PAGE 26

17 Table 4. Elrments for VJhich AE has a DG*-.ection Limit Lov;er than I\A by at Least a Fa.ctor o€ 5 Al Ca Ce Cs Eu Ga In La Li Lu Ka Kb Os Sin Sr Th Element AE Detection Limit, ug/ral 0.005

PAGE 27

18 Atomic fluorescence . -AF is the newest of the atomic spectroscopic methods. According to Winefordner et al_. (139) , for a population of atoms in an atom reservoir with low background the sensitivity of AF measurements is inherently greater than that of AA because of benefits gained from signal amplification and increased source intensity. This greater sensitivity has been applied to serum analysis by a few workers, such as Kolihova and Sychra (7) , who determined Cu. As source stability and intensity problems are overcome, AF will become more popular for trace analysis. Arc-Spark Emission Spectroscopy Emission spectroscopy is capable of determining up to 70 elements with absolute lov/er limits of detection down to 10 ng (140) . For serum, sample preparation consists only of dilution if the Stallwood Jet or similar apparatus is used, or dry ashing if a conventional graphite cup electrode is used. No separation steps are required. Matrix interference is not a problem, especially if the cyanogen bands have been eliminated by using an arc stand with an inert atmosphere. Besides these advantages, simultaneous multielement determination is possible, an important consideration when only a small amount of sample is available for a survey-type determination. Drawbacks include the difficulty of obtaining quantitative data of even limited accuracy. Photographic emulsions require troublesome calibration, and tedious and time consuming development. Direct readers eliminate this, but allow observation of only a limited number of v/avelengths . These

PAGE 28

19 must be chosen before analysis, thereby limiting the technique even further. Direct readers also add significantly to the already large expense associated with emission spectroscopy. Monacelli et al . (46) , and Paixo and Yoe (45) determined Cr, Cu, Mg , Ni and Zn in plasma using emission spectroscopy. Nedermieier and Griggs have used a direct reading emission spectrometer to determine the concentrations of 14 trace elements in serum (40) . Colorimetry and Fluorimetry These two techniques can be very simple, sensitive, and require only inexpensive apparatus under certain ideal conditions. Unfortunately, since hardly any of the organic reagents used with these techniques are specific for only one element, interference is a major problem. Complex separation schemes can be necessary, especially for complex samples such as serum, thereby removing any benefits from the simplicity of the technique. Colorimetry . -Trace quantities of elements in serum determined colorimetrically include I (78) , Br (112) , Fe (81, 82), and Cu (6). In their review of trace element analysis in clinical chemistry, Schroeder and Nason (117) include some lower limits of detection for several elements using colorimetry. These are shown in Table 5. Where the limit of detection is specified in yg rather than yg/g, the figure indicates the smallest amount detectable by difference in two solutions, both v/ith known am.ounts added (117),

PAGE 29

20 Table 5. Colorimetric Lower Limits of Detection Sn V Nonessential As Pb Element Limit, (yg/g) Essential Cr 0.006 Mn Mo Ni 0.05 0.01 yg 1.0 yg 5.0 yg 0.5 yg 0.5 yg B 0.5 Ge 0.1 Nb 0.04 0.05 yg Ti , 0.25 yg Zr 0.8

PAGE 30

21 Fluorime try . -Since the lower limit of detection of any spectroscopic technique is more a function of the signal-tonoise ratio than signal strength alone, the limits of detection for inorganic fluorimetry are usually much lower than those for inorganic absorption spectrophotometry (141) . The limits of detection in interference-free media for several elements as noted by Winefordner et al . are listed in Table 6. These limits show that fluorimetry can be comparable to other sensitive methods of trace analysis if interference is not a problem. Interference in complex samples such as serum is greater in fluorimetry than in AE, AA or AF , however. X-Ray Techniques X-Ray fluorescence . -Elements with atomic numbers above 15 can be effectively determined at ppm levels by X-ray fluorescence. Recent reductions in the size of X-ray sources, lower detector background, and improved resolution and counting rate capabilities of detectors and associated electronics have greatly increased the capabilities of the technique (142) . Semiconductor detectors capable of surveying an entire spectrum of trace elements present at levels below 1 ppm have been used in serum analysis (143, 144) . Analyses of petroleum (145) and air filter particulates (146, 147) have demonstrated absolute detection limits down to 10 ng/cm^ for small particles. These levels cannot be reached with serum without first concentrating the dissolved solids to a greater degree than is provided by drying alone. Ion exchange membranes (145) and chelating ion exchange resins (148) provide

PAGE 31

22 Table 6. Fluorimetric Lower Limits of Detection (141) Element Limit, yg/ml Element Limit, yg/ml Ag

PAGE 32

23 both preconcentration and thin geometry. Cobalt determination at 0.001 ppm (145) and Cr(VI) and/or Cr(III) determination at microgram levels (147) have been carried out in this manner. Quantitative analysis in a matrix as complex as serum requires complicated mathematical manipulation of the data to compensate for the X-ray enhancement and absorption effects of other elements in the sample. These interelement effects are compensated for by solving sets of simultaneous equations containing measured constants for each element present (149). Losses of sensitivity also occur if higher spectrometer resolution is required for analysis of complex mixtures. This becomes even more significant if the high resolution of less sensitive dispersive systems is required. Electron probe and laser probe X-ray microanalysis . -This technique has been applied successfully to the analysis of small regions in hetreogeneous samples. Its relative lower limits of detection are not outstanding, but absolute limits as low as 10"^ pg have been reported (145) . Since the mass of a serum sample after dry ashing and sodium removal is very low, this technique may be successfully applied. Mass Spectrometry Spark source mass spectrometry . -For trace element qualitative and semiquantitative analysis, this technique is unsurpassed in cases where both lov/ limits of detection and a wide range of elements analyzed are required. Interferences are not severe, thus allov/ing lov/er limits of detection to range

PAGE 33

24 from 0.1 to 0.001 pprn for most elements in any matrix (150). The small sample size requirement makes this technique especially useful for serum analysis, where the mass of the "ash to be analyzed is quite low. The most serious drawback of this technique is its poor quantitative accuracy. The precision is limited to about 25% (151) . The discharge is unstable, thus preventing spectrum scanning and requiring photographic detection. Ion yield is influenced by sample preparation, thus making comparisons between natural samples and synthetic standards unreliable. Lack of good quantitative data combined v;ith the expense associated with purchasing and maintaining the instrument are reasons why little trace elonent determination in serum has been done. One of the few publications using this technique is by Wolstenholme (27), who determined 25 elements in plasma. The reported concentrations deviate from accepted values by factors as large as 100, thus illustrating the difficulty of accurate quantitative analysis. Isotope dilution mass spectrometry . -Much of the uncertainty in qualtitative mass spectrometry mentioned above can be eliminated by using isotope dilution. No reference standards are required, and the technique is sensitive and specific. Isotope dilution allov/s precise quantitative determination because only isotope ratios from lines recorded under identical matrix, exposure, development, and emulsion conditions are measured. All the uncertainty due to lack of reproducibility is eliminated.

PAGE 34

25 Early application of the isotope dilution technique with electron bombardment sources was limited by element volatility. The technique has been extended to less volatile metals by Frew et al . (152) , who determined Cr as a 3-diketonate. Isotope dilution can also be used with an rf spark source (153). This involves additional sample preparation, as does a chelation step. The care required in the tedious steps, as well as equipment expense keep this technique from being popular for serum analysis, although some serum analyses have been reported (154) . Electrochemical Methods With the exception of potentiometry , electrochemical methods have not been widely used for trace element determinations in biological media. For example, in a recent Analytical Chemistry reviev; article which dealt with biological trace analysis only one article of the 49 reviewed dealt with a nonpotentiometric electrochemical determination (155) . One reason for this is the susceptibility of electrochemical techniques to interference. Anodic stripping voltammetry, for example, can be applied to extremely low concentrations and is still sufficiently selective to allow separation from species present in a thousandfold or greater excess (156) . Even so, multielement serum trace analysis cannot be carried out for elements with similar unmasked electrodeposition potentials. Because only a 1.5-V anodic dissolution "window" exists between the cathodic deposition potentials and the

PAGE 35

26 anodic water-salt breakdown potential limit, one cannot obtain separate and distinct dissolution peaks for the large number of different ions present in serum (156) . Potentiometry with ion selective electrodes has been applied to serum analysis. Serum calciiim, a bulk constituent, has been determined (91, 92) as has serum fluorine (115). As sufficiently sensitive ion selective electrodes become available for more elements, more such determinations can be made. Czaban and Rechnitz (157) have constructed ion-selective electrodes with tip sizes in the 100-150 ym range. Microelectrodes such as these are capable of making in situ measurements in biological systems. Even though concentration measurements as low as 10"^ M can be made (157) , in situ serum trace element determination may not be possible due to the strong metal binding properties of the serum proteins. The protein complex formation constants must be known and the extremely low free ion concentration must be measured before a meaningful concentration figure for a given element can be obtained. Coulometry has received limited use in biological trace element analyses. Selenium in urine (158) as well as chromium in serum (71) has been determined. As little as 0.2 yg of chromium was detected. Gas Chromatography Gas-liquid chromatography is one of the most widely used methods of qualitiative and quantitative organic analysis, as well as an extremely powerful tool for separation prior to

PAGE 36

27 analysis by other techniques. The chemical similarity of the compounds being chroma tographed poses no problem, since separation depends only on very small differences in the volatility and solubility of the compounds in the liquid partitioning phase. Instrumentation can be inexpensive, yet provide analyses which would be difficult, if not impossible using another method. Extraordinary sensitivity, selectivity, ease of operation, and speed of analysis are some of the characteristics of gas chromatographic separations. The extremely high resolution available with long columns is unsurpassed by nonchromatographic separation methods. Optical isomers, geometrical isomers, even isotopes have been separated by gas chromatography . For several years after its initial development this powerful analytical tool was used exclusively for organic compounds. This was due to the lack of suitable volatile, thermally stable metal compounds. In 1959, Wachi (159) and Preiser (160) reported gas chromatographic metal chloride separations, and Duswalt (161) reported separation of metal acetylacetonates. From this point metal analysis by gas chromatography has grown to be the well established technique it is today. Metal chlorides . -Some metals can be separated by gas chromatography in the pure state. Sokolov et al . (162) have separated cadmium and zinc vapor at 850 °C. Such separations are difficult, requiring severe conditions. The use of volatile metal chlorides allows separation at much lower tempera-

PAGE 37

28 tures. The chlorides of Ni, Sb, Sn, Ta and Ti are sufficiently volatile for elution. Separations of Sn and Ti, Sb and Ti, and the gas chromatographic behavior of ten other metal chlorides have been reported (163). Sie, Bleumer and Rijnders (164-166) have studied the gas chromatography of 12 elements, all of which were found to elute satisfactorily. The boiling points of the chlorides studied are listed in Table 7. All but two boil below 200 °C and are eluted at 125 °C after reasonable periods of time (164). Metal halides can require moderately high temperatures for elution. Their reactivity, especially their hygroscopicity , presents problems in sample handling and choosing adequate column packing materials. Orqanometallic compounds . -Most classes of volatile metal compounds are organometallic in nature. Among these are metal alkoxides, carbonyls, alkyls, hydrides and TT-bonded metal complexes such as cyclopentadienyls (167) . Metal alkoxides, such as Al, Ge, Si and Ti isopropoxides , as well as Hf, Ti and Zr tertiary-amyloxides , have been eluted v/ith little decomposition at moderate column temperatures (168) . Most of the rest have proved unsatisfactory for various reasons. Metal carbonyls, hydrides and alkyls rarely form in quantitative or easily reproducible yields. In addition, solvolysis during formation and in the gas chromatographic liquid phase is often encountered with all of the classes mentioned. Metal chelates . -Of all the various classes of metal compounds, metal chelates are the most useful for metal analysis by gas chromatography. This is due to the ease with

PAGE 38

29 Table 7. Boiling Points of Various Chlorides Element C Si P Ti V Cr Ga Ge As Sn Sb Substance CCI4 SiCl^ PCI 3 POCI3 SCI2 S2CI2 SOCI2 SO2CI2 TiClit veil, VOCI3 Cr02Cl2 Ga2Cl6 GeClit ASCI3 SnCl^ SbCls SbCls Boiling point, °C 76.7 57.0 76 105.3 59 138 75.7 69.2 135.8 152 127 117 200 83.1 130 113 221

PAGE 39

30 which they form in quantitative yields and their general lack of hydrolysis. There are several ligands whose chelates have been studied using gas chromatography. The 8-hydroxyquinolate complexes of AI(III), Cd(II), Co(II), Cr(III), Cu(II), Fe(III), Ga{III), In(III), Mg(II), Mn(II), Ni(II), Pb(II) and Zn(II) will sublime at reduced pressure at temperatures between 250 and 500 °C leaving no significant residue (169). However, their chromatography is complicated by a lack of suitable high temperature packings. Benzoyltrifluoroacetone [H(tba)] and thenoyl-trif luoroacetone [H(tta)] have been used as ligands with Al, Cr, Cu, Fe and Ga in successful gas chromatographic separations (170). Pivaloyltrifluoroacetone [H(pta)] chelates of Y, Sc , and nine rare earth elements have also been successfully eluted (171) . Metal chelates of the 3-diketones are the best choice for gas chromatography of metals thus far studied. They form with quantitative yields, several metals in a mixture being chelated simultaneously and reproducibly with ease. In most cases solvolysis and hydrolysis are not significant. Besides this, fluorinated derivatives yield chelates whose vapor pressures are low enough to permit gas chromatography without appreciable decomposition, while taking advantage of the extreme sensitivity of the electron capture detector for electronegative elements. iMost fluorinated 3-diketonates will rapidly sublime at reduced pressures at temperatures from ambient up to 100 °C.

PAGE 40

31 Rare earth chelates of several 3-diketone ligands have been successfully eluted. These include 2 ,2 ,6 ,6-tetramethyl3,5-heptanedione [H(thd)] (172), 1 ,1 , 1-trif luoro-5 , 5-dimethyl_ 2,4-hexanedione [H(tapm)] (173) and 1 ,1,1,2,2 ,3 ,3-heptafluoro7,7-dimethyl-4 ,6-octanedione [H(fod)] (174). All three formed anhydrous, stable chelates suitable for gas chromatography, but the fod chelates appear to be the most useful. The ability of several 3-diketones to form efficiently extracted, completely eluted chelates of Zn and Co has been studied. The ligands compared were H(fod), 1 , 1 , 1-trif luoro2,4-pentanedione commonly called trif luoroacetylacetone [H(tfa)], 1,1,5,5,5-hexaf luoro-2,4-pentanedione commonly called hexa-1 f luoroacetylacetone [H(hfa)], 1 , 1 , 1 ,2 , 2 , 6 , 6 , 6-octaf luoro-3 ,5hexanedione [H(ofhd)], and 1 , 1 ,1 ,2 ,2 , 3 , 3 ,7 ,7 ,7-decaf luoro4,6-heptanedione [H(dfhd)]. The order of extraction efficiency from aqueous solution into benzene was H (f od) >H (dfhd) >H (of hd) > H(tfa) >H(hfa) . Even though H(fod) extracts more efficiently than the rest, the fact that H(dfhd) forms chelates much more rapidly (10 min versus 4 hr for the Zn complex) makes it more desirable for some applications (175) . Three of the most successful of the fluorinated g-diketone ligands are H(tfa), H(hfa), and H(fod). Their anions are shown in Figure 1. Their chemistry is discussed below. H(tfa) .-Trif luoroacetylacetone is easy to work with, forming many stable chelates. H(tfa) boils at 107 °C and is soluble in most organic solvents. It is slightly soluble in water. The ease with which tfa chelates are formed is typical

PAGE 41

32 •^O c o 0) a •4— c CD Q. I cvT ljl O Ll.-A_Llo 3 ^ 11 a a; JO. I in LO in Or.6 iL CD ^— a c o "O CD cz o c o CL %. o o 3 LlX I rO :l •b li. CD •— a o CD c a •— o o I ^0) £ o 3 cx CD x: ro CM ro o-o-o ^ OrrO li o-Ll. Ll_-QiJL ll c o +J Q) U to r-\ >^ +J
PAGE 42

33 of the B-diketones. Solvent extraction is applicable to most samples and yields a chelate solution which can often be injected into the gas chromatograph without further treatment. If sample dissolution is required, either nitric or perchloric acid should be used. This avoids adding anions which can compete with the chelating agent for the metal ion. Next the pH of the solution must be adjusted. For some extractions there is a v;ide range of pH values within which the optimum extraction efficiency is obtained. Genty et al . (176) found 100% extraction of Al(tfa)3 between pH 5 and 9. Scribner et al . (177) found essentially 100% extraction of Cu(tfa)2 between pH 4 and 8, and of Fe (tf a) 3 between pH 4 . 5 and 6.1. In a later paper Scribner et al . (178) report 99.9% extraction for Be(tfa)2 between pH 5 and 7. A sharp rise in extraction efficiency around pH 3 and a high plateau continuing to at least pH 7 is typical of tfa chelates of trivalent metal ions. The divalent ions of Mg , Mn , Ni and Zn are all poorly extracted as tfa's in the pH 3 to 7 region due to hydrate formation. In some cases the pH adjustment is critical. This is particularly true where variations in pH are to be used to carry out the separation of groups of metals prior to gas chromatographic analysis. Some metals form insoluble hydroxides which can interfere with chelate formation. Cu can be extracted efficiently from solutions cloudy with Cu (Oil) 2 , but Fe does not extract

PAGE 43

34 cleanly when initially present as an hydroxide (177) . One way to circumvent this problem when dealing with trace concentrations is to keep the pH so low that the solubility product constant for the hydroxide is not reached. If this cannot be done within the pH range of efficient extraction, weak auxiliary complexing agents such as acetate ion can be used. Some neutralization of the acidic H(tfa) is often required. Acetate salts can thus bring the pH up to a suitable level while preventing hydroxide formation. One drawback is that if a large variety of ions is being extracted, some may be sequestered by the acetate ion and not extracted at all. Sequestration has been shown to occur with ammonia at pH 9 and higher, as well as with EDTA at pH 5 (177) . Since there is competition for the basic sites on the ligand anion, complex formation is favored at higher pH. The pH also affects ligand solubility. The larger the concentration of ligand in the aqueous phase, the farther the chelate formation reaction will proceed. This consideration is not critical with H(tfa), since its distribution between the organic phase and the aqueous phase can favor chelation at pH values as low as -0.5, with D , = 2.9 (177) o/w The organic solvent chosen for the extraction should have a high chelate solubility to avoid chelate precipitation in the aqueous phase. If an electron capture detector is to be used for gas chromatographic detection, a hydrocarbon such as benzene for which detector response is slight can be used to advantage. Chelation and extraction generally

PAGE 44

35 occur immediately after the ligand-organic solvent mixture is introduced. Excess ligand and an additional extraction step are usually helpful in ensuring complete extraction. The solid chelate can be obtained by evaporating the solvent. Alternatively, an aliquot of the solution can be injected directly into the gas chromatograph. While solvent extraction is the oldest and best established technique for forming metal chelates for analysis by gas chromatography, faster and less complex methods are being developed. Ross and Sievers (179) report a direct chelation procedure used for the analysis of trace chromium in steel. A metal particle to be analyzed is placed in a glass tube with a small amount of H(tfa) and a drop of HNO 3 as a catalyst. Next the sealed tube is heated for 30 min with a microwave generator. The use of microwave energy stimulates reaction at the metal-solution interface without overheating the major part of the unreacted ligand and risking decomposition. The solution is then dissolved in benzene and the excess ligand, along with Fe(tfa)3, is rem.oved. Sievers et al . have studied the suitability of various metals for chelation and gas chromatographic analysis (180). Their results are shov/n in Table 8. Class I is made up of chelates chromatographed with no evidence of decomposition in the injection port or column. Sharp, well defined peaks were obtained. Class II compounds, although eluting with sharp peaks, left residue in the injection port, indicating partial decomposition.

PAGE 45

36 Table 8. Metals Suitable for Gas Chromatography Class I Class II Al Hf Be Mn Cr Zn Cu Zr Fe Ga In Rh Sc

PAGE 46

37 Metal analysis by gas chromatography has fewer interference problems than most other techniques because the analysis process involves separation. Interference in the analysis of various tfa complexes has been studied. Ross and Sievers (181) found no interference with ppm Be analyses from a mixture of seven metal cations, most of which form stable chelates with H(tfa). A mixture of five anions also failed to interfere. It has also been shown that quantitative extraction of some metals, including Be, can occur in the presence of EDTA (178). If interference at high concentrations should pose a problem, the interferents could probably be effectively masked by EDTA. Genty et al . (176) found some metals interfered to a small extent in the analysis of microgram quantities of Al in uranium. The interferents were present at concentrations 100 times greater than the Al , yet only errors below 25% were encountered. Savory et al . (67) tested various ions for interference in the analysis of Cr as Cr(tfa) 3. The ions tested were As, Pb, Hg, Li, Mn and Sr at 50 yg/100 ml; K at 40 mg/100 ml ; Ba, Cu, Ni, and Zn at 100 ug/100 ml; Ca at 10 mg/100 ml; and (H2P0i»)~at 5 mg/100 ml. The presence of these ions had no effect on the Cr determination. The usefulness of H(tfa) has been shown in many analyses. A few representative analyses include Al , Fe and Cu determined simultaneously in alloys (177, 182); Fe and Al (183) and Al , Ga and In (184) determined using the solvent extraction

PAGE 47

38 procedure; microgram quantities of Al , Ga, In and Be determined with an overall relative mean error of 2% (185); and Be, Al , and Cr (186) , as well as Al in uranium, determined at levels down to 10"^^ g (176) . H(hfa) .-Hexafluoroacetylacetone boils betv/een 63 and 70 °C and is soluble in most organic solvents. Since H (hf a) is more highly fluorinated than H(tfa), its chelates have higher vapor pressures. This allows the hfa chelates to elute at lower column temperatures, thereby allowing more thermal instability. Cr(hfa)3, for example, can be eluted rapidly at column temperatures as low as 30 °C. This temperature would have to be more than 100 °C higher to elute the nonfluorinated chelate under similar conditions (187) . Unlike H (tf a) , H(hfa) reacts with water to form a tetrahydroxy compound, 1 , 1 , 1 , 5 , 5 , 5-hexaf luoro-2 , 2 , 4 , 4-tetrahydroxypentane. This makes synthesis of hfa metal chelates from aqueous solutions of the metals much more complicated than the tfa synthesis and extraction described above. Although some aqueous hfa syntheses have been reported, they are probably best avoided. One alternative is conversion to the metal chloride followed by direct reaction with the ligand to yield the chelate and HCl . This is more time consuming and awkward than the extraction procedure. Hill and Cesser (186) successfully performed quantitative analysis of Be, Al and Cr as hfa chelates. Cr has been determined as the tfa chelate in the 10~^ 10~® g range (188),

PAGE 48

39 while Cr and Al have been determined at the 10"^^ g level (189). Moshier and Sievers list Al , Be, Cr , Rh and Ti as having been successfully eluted as hfa complexes without any trace of decomposition (190). The hfa complexes of Ni , Ta and Ti have been successfully chromatographed, but only with great care to prevent reaction with moisture in the air. It has been shown that Cu(hfa)2 will deposit elemental Cu on the walls of a glass tube through which its vapor is passed at 275 °C. In a hydrogen atmosphere this deposition takes place at 250 °C. Injection port temperatures should be chosen with this in mind. On the other hand, Cr(hfa)3 is unusually stable, resisting decomposition at temperatures up to 375 °C. It can easily be steam distilled, and distills unchanged from boiling aqua regia or concentrated sulfuric acid (191) . H (fod) . -Heptaf luorodimethyloctanedione boils at 33 °C (2.7 Torr) and is soluble in most organic solvents. Its complexation reactions are similar to those of the other 3-diketones, but it can provide additional latitude in chelation-gas chromatographic analyses. Where volatile chelates containing a given metal are formed by several ligands, the H(fod) and H(tfa) chelates of Cr(III) and Al(III) being typical examples, the vapor pressures of the H(fod) chelates are slightly lower than those of the H(tfa) chelates. However, the primary utility of H(fod) lies in its capability of forming chelates with metals with which no other ligands can form useful complexes. Anhydrous metal complexes which are coordination

PAGE 49

40 saturated with the ligand have been the most satisfactory for this type of analysis. Hydrate formation causes poor solvent extraction efficiencies into organic solvents (175) . For example tfa chelates of divalent metal ions such as Mn , Mg, Zn and Ni, which form hydrates to satisfy coordination requirements, are extracted only to the extent of a few percent (177) . In addition, the presence of water has been suggested as the cause of the thermal instability encountered in many hydrated chelates (172) . Hydrolysis at elevated temperatures is thought to occur. H(fod) is a much bulkier molecule than H(tfa) or H(hfa). It forms sterically crowded chelates. This crowding results in a decrease in the tendency for a complex to form a hydrate, even though in some cases coordination is incomplete. Anhydrous Co (II) and Zn(II) bis chelates of H(fod) have been formed, extracted, and chromatographed successfully, even though the coordination requirement of six has not been met (172). This illustrates the usefulness of H(fod) chelates for the analysis of divalent metal ions. In another study, the bis fod chelates of Ni(II), Pa(II), Cu(II), and Be(II) were also quantitatively formed and eluted without apparent decomposition (192). The tris chelates of the B-diketones generally have the most desirable physical properties for gas chromatography. Rare earth ions tend to accept extra donor groups, easily forming complexes with coordination numbers greater than six. Most tris bidentate rare earth complexes therefore are

PAGE 50

41 hydrates (172) . These hydrated chelates have been reported to be thermally unstable (174) . Hydrolysis yielding the hydroxobis diketonate and neutral ligand is suspected to occur, both at elevated temperatures and in some cases under vacuum at room temperature (172) . H(fod) forms anhydrous and monoaquo tris complexes with trivalent rare earth ions. The monohydrated complexes and anhydrous are interconvertible by a totally reversible process with no hydrolysis (174) . This is believed to be possible because the water molecule may be hydrogen bonded to an electronegative site on the ligand, rather than coordinated to the metal ion (174), due to steric crowding in the complex. Fifteen tris rare earth H(fod) complexes have been successfully chromatographed (174) . They were found to be more volatile than any other known compounds of the lanthanide elements. The synthetic procedure, which involves precipitation of the chelate from aqueous solution, leads to the same difficulties as the similar extraction procedure described earlier. The product, a monohydrate, is typically dehydrated at gas chromatographic elution temperatures and separated in an anhydrous form (174) . Purification by sublimation must be done carefully, since the chelates decompose at temperatures slightly above 200 °C (172). Quantitative extraction of Co(fod)3 from aqueous solutions at sub-ppm levels has been described (193) . The aqueous sample solution is mixed vzith portions of NaOK solution and H2O2 solution. After the reaction is complete, tlxe resulting

PAGE 51

42 chelate-benzene solution is decanted and the excess H(fod) is back-extracted to prevent electron capture detector, poisoning. The efficiency of extraction was found to be highly dependent on the order in which the reagents were added, while variations in pH from to 6.4 had little effect. The relative insensitivity to pH allows the choice of a pH at which the extraction efficiencies of potential interferents are at a minimum. This method has been used to determine Co in vitamin B^^ i'^ liver extracts at levels as low as 4 x 10"^^ g (193) . One disadvantage of the fod chelates for solvent extraction, as compared to tfa chelates, is the longer equilibration time required. For example, a minimum of 4 hr is necessary for reaching equilibrium when extracting Zn and Co (175) . This can sometimes be avoided by using direct reaction techniques. Sievers et al . (192) report an extremely simple direct chelation procedure for forming fod chelates of Al , Be, Cr , Fe, Ga, Hf, In, La, Mn, Ti , Sc, V, Zn and Zr. A solid sample containing the metals to be analyzed is sealed in a glass capillary with the neat ligand. Heat is then applied briefly to speed up the reaction before injection into the gas chroma tograph . A similar direct reaction procedure has been used in the analysis of Co as the fod chelate (193) . Nonaqueous samples were placed in a small vacuum sublimation apparatus along with a small amount of FiNOj and the ligand. After

PAGE 52

43 reacting at 104 °C for 4 hr, the excess ligand was removed and tris-Co (fod) 3 was recovered by sublimation. When the same procedure was carried out at 127 °C, bis-Co (fod) 2 * 2H2O was formed. The reaction temperature is thus critical if one is to avoid the thermally unstable hydrate. Instrumentation Injection . -Metal chelates can be injected into the gas chromatograph either as solids or in solution. One type of solid sampler successfully used by Sievers et al . is a bayonet type which crushes sealed glass melting point capillaries containing the sample in the injection port (192) . The solid sample injector consists of a hollow stainless steel tube with a plunger which fits inside. To operate the device the plunger is removed, the sealed capillary is placed in the tube and the plunger is inserted all the way, breaking the capillary by forcing it against the beveled end of the device and releasing the sample in the carrier gas stream. This type of sampler was used with the sealed tube microanalysis method described by Sievers (192). Besides applications such as this where the capillary is also a reaction vessel, the solid sampling technique is useful where a large amount of chelate is to be injected, or where the presence of a solvent is to be avoided. Liquid injection by microsyringe is used more frequently. It is more convenient than solid samplers, which require that the carrier gas flow be interrupted before each injection and that the glass capillary fragments be removed from the injection port periodically. Another advantage of liquid injection is that, since the

PAGE 53

44 chelate is already dispersed in the solution before solvent volatilization, instantaneous vaporization can be achieved at lov;er injection port temperatures than would be required if the sample were introduced as a crystalline solid (187) . In addition, the last step in many chelation reactions being solvent extraction, the organic extract lends itself to liquid injection. No capillary sealing, v/ith the accompanying risk of sample loss or decomposition, is required. Repetitive sampling can be made much more quickly and conveniently when a liquid sample and syringe are used. Detectors . -The three most popular gas chromatographic detectors for metal chelates are the thermal conductivity (TC) detector, the electron capture (EC) detector, and the flame ionization detector (FID) . The TC detector is the least sensitive of the three, with a detection range for metals extending from 10"'* to 10"^ g. It is easy to operate and responds to all compounds. It must be used where nondestructive detection is required due to effluent sampling for further .analysis. The EC detector is selective and extremely sensitive to compounds with high electron affinities such as fluorine substituted metal chelates. It is relatively insensitive to hydrocarbon solvents, permitting detection of picogram quantities of a halogenated compound without interference from the tail of the solvent peak. Some reported lower limits of detection with the EC detector are 3.1 x 10"'^ moles for Cr, 1.2 x 10"^^ moles for Al , and 2.0 x 10~^ ^ moles for Cu (194);

PAGE 54

45 10"^^ g for Be (181), lO"^"* g for Cr (67, 68), and 10"^^ g for Co and Al (176) , all as tfa chelates. Intermediate between the EC and TC detectors in sensi• tivity and ease of operation, the FID is an excellent choice for many metal chelate determinations. Lower limits of detection with the FID have been reported as 3.7 x 10"^ ° moles for Cr (tfa) 3, 1.7 x 10"^ ° moles for Al(tfa) 3, and 1.6 X 10~^ moles for Cu(tfa)2. The linear response ranges from 10"^ to 10""* g for Cr(tfa)3 compared to 10"^^ to 10~^ g for the EC detector (194). Its operational stability is better than that of the EC detector, and it is not poisoned by large amounts of chelate or ligand. In the ppb range the greater sensitivity of the EC detector is required, but above 5 ppm the FID is the method of choice (194). Serum application . -The application of metal chelate-gas chromatographic analysis to serum has been limited. Savory, Mushak, and Sunderman (68) , and Savory et al . (67) have determined Cr concentrations in serum in the vicinity of 40 ng/ml. They report good recovery of Cr added to serum samples, and good reproducibility. Chelation and solvent extraction of Cr in untreated serum yielded no Cr(tfa)3. This was also true of serum which had been partially wet ashed. In contrast to this, Hansen et al . (69) determined Cr in plasma at the 50 ng/ml level using a direct chelation-sealed tube procedure with no attempt to break down the protein bound Cr complexes by ashing. Their work, which included live animal experiments to evaluate chelation efficiency, describes a simple, direct process with a lower limit of detection around 5 ng/ml of serum.

PAGE 55

46 Taylor et al . (112, 113) have determined Be in biological fluids using a sealed tube reaction similar to Hansen's Cr procedure. They, too, found H(tfa) to be an effective direct chelating agent, forming quantitative chelate yields in the presence of the naturally occurring complexing agents present in serum, thus avoiding time consuming ashing and extraction steps. They added NaaEDTA to untreated blood to prevent clotting and to sequester interfering metals such as Fe in the blood sample. This enables Be concentrations as low as 0.02 yg/ml to be determined. H(hfa) has not been used successfully in serum trace metal analysis. Hansen et al . (69) in an analysis of Cr in plasma attempted a direct chelate formation reaction using H(hfa) dissolved in hexane. This solution was placed with untreated plasma in a sealed tube and heated. A maximum of 50% of the metal was chelated under optimum conditions. A white layer at the plasma-hexane interface was assumed to be the solid dihydrate of H(hfa). Hansen et al . (69) used a modified direct chelation procedure to determine Cr in serum using H(fod). Only 50% of the Cr was chelated, however. H(fod) is a useful complement for H(tfa) in that it forms volatile chelates with metals with which H(tfa) does not. The lower volatility of the fod chelates, however, makes H(tfa) the ligand of choice where possible. Neutron Activation Neutron activation analysis (NAA) has been used frequently

PAGE 56

47 as an analytical tool for trace element analysis of serum samples. The general applicability of the technique at the trace level is illustrated by the lower limits of detection listed in Table 9 (116). These figures are based on the measurement of 40 disintegrations per second in a sample irradiated at 5 x lO'^ n cm~^ s "^ for an optimum time period. NAA has the advantages of multielement capability, low detection limits, wide dynamic range, and invariant sensitivity with the nature, size and concentration of the sample. The sensitivity is determined by fixed nuclear parameters such as half-life, cross section and natural abundance, which are constant under all conditions. In complex samples, the small photopeak from a trace constituent is often partly or completely obscured by y-rays from other radionuclides. If the half-lives of the offending radioisotope and the radioisotope of interest differ sufficiently, partial photopeak resolution can be obtained by carefully choosing the duration of neutron irradiation and decay. This technique is used in nondestructive, or instrumental NAA. If this approach cannot be used, or if the photopeaks cannot be resolved by a high-resolution detection and counting system, a chemical separation step must be included. Instrumental neutron activation analysis . -If the difference in half-life between the radionuclide of interest and an interferent can be exploited, separation is not necessary. Short half-life isotopes are measured using irradiation times too short to allow interferent activity buildup.

PAGE 57

48 Table 9. NAA Lower Limits of Detection E

PAGE 58

49 while analyses of longer half-life radionuclides incorporate a delay after irradiation during which a short half-life interferent decays to an insignificant level. Long and short half-life isotopes can sometimes be counted together by using a long irradiation, a decay period, and a short reirradiation followed immediately by counting. The most pervasive interference in NAA of biological materials is due to 15-hr ^'^Na. The parent isotope, ^^Na, has a large natural abundance, a high thermal neutron cross section, and is present at concentrations from 10^ to 10^ times greater than the trace elements of interest. These factors combined with the relatively high energy of the emitted Y-rays prevent the measurement of small amounts of radionuclides in many biological matrices. A high resolution y-ray spectrometer is essential for performing multielement instrumental activation analysis. Coaxial Ge(Li) detector resolution is 10 times better than that of Nal(Tl) detectors. Planar Ge(Li) detectors have still another order of magnitude of improved resolution, although they suffer from a lack of sensitivity due to their smaller active volume. The higher sensitivity of the coaxial detector compensates for its lower resolution, making it the detector of choice for activation analysis. Although Nal(Tl) detectors are 10-20 times more sensitive than Ge(Li) detectors, this does not compensate for their extremely poor resolution. A high resolution multichannel pulse height analyzer with several thousand channels is needed to take advantage

PAGE 59

50 of the resolution of the Ge(Li) detector when a large energy range spectrum is being recorded. Where possible, direct computer interfacing is also desirable in order to facilitate data handling and reduction. In spite of interference problems, successful instrumental determinations in biological matrices have been reported. Nadkarni and iMorrison (195) described a nondestructive NAA procedure for trace element analysis capable of yielding data for 36 elements in certain types of biological matrices. Irradiation periods ranging from 1 min to 8 hr, with decay periods ranging from 1 min to 4 weeks were used. Due to the extremely low concentrations present, however, the number of elements which can be determined in blood by this method is not as large (195) . Blood analysis with short irradiation periods (196, 197), and long irradiation periods coupled with Ge(Li) anticoincidence counting (198) and an elaborate instrumental setup including a filtered reactor neutron spectrum and triple coicidence counting with a ^^Mn/2'*Na activity ratio of 700 (197) have been reported. Yule (199) has published a list of instrumental detection limits in various matrices. Table 10 lists his estimates for blood. Comparison with Table 1 shows that most elements whose serum concentrations are known are present at levels too low to be detected by NAA without first removing the sodium, Comparison with Table 9 shows the decrease in detection limit achieved in blood samples by chemical separation. Destructive activation analysis: . --Zydvances in the resolution of detectors and multichannel analyzer resolution have

PAGE 60

51 Table 10. Limits of Detection in Blood by Nondestructive NAA E

PAGE 61

52 greatly reduced the degree of radiochemical purity required from radiochemical separations. In many cases, a group separation is all that is required. Nevertheless, high separation factors are still needed if a very large excess of interferent is present, as is the case with sodium in serum. Meinke (200) has compared preand postirradiation separations. The choice depends on the nature of the sample and the separation technique involved. Postirradiation techniques permit carrier use and require no blank correction, while preirradiation separation is necessary whenever the removed interferent would have resulted in too high a radiation level in the sample. Ashing . -Most chemical separation procedures require that all atoms of each species, including the carrier, be in ionic form, in aqueous solution, and in the same chemical state. In matrices such as serum, where the trace elements are combined in complex molecules, complete breakdown of the sample is required prior to separation. Ashing is thus considered necessary in most cases. Trace element losses during wet ashing and high temperature dry ashing have been thoroughly explored (201) . The 700-800 °C temperature required for dry ashing in a furnace results in the loss of volatile metals and metal compounds, such as chlorides. Wet ashing, while using more moderate temperatures, introduces impurities via the reagents. The temperatures reached in a low temperature dry asher of the

PAGE 62

53 rf-excited oxygen plasma type can be kept as low as 120 ''C. No reagents are required. The technique is slow, but if large sample surface areas are exposed ashing can be completed in a few hours. Precipitation . -One of the oldest classical techniques, precipitation is frequently used with NAA. The comprehensive group separation schemes of Robertson (202), and the analysis biological samples by Bowen (203), Heydorn and Damsgaard (204), and Rakovic (205) illustrate its usefulness. Special applications of precipitation to sodium separation from biological samples include coprecipitation of sodium on KClOi, (206-208) and precipitation of NaCl in n-butanol saturated with HCl (209). In general, precipitation is not selective due to coprecipitation and absorption. Large quantities of carrier are required, thus ruling out preirradiation separation and complicating the use of precipitation with other techniques. Distillation . -One of the best techniques for obtaining high purity separated fractions is distillation. Elements can be distilled as halides, oxides, hydrides, and sometimes even in the elemental state (210) . It has been used in NAA as a preliminary step in complex schemes (211, 212), for various biological samples (213, 214), and for trace analysis in blood (38, 57). Compounds for which distillation is applicable are also amenable to more powerful separation techniques such as gas chromatography. Distillation, then, often does not fully exploit the existing potential for efficient separation.

PAGE 63

54 Electrochemical methods . -Controlled potential electrodeposition is a somewhat selective separation technique useful for concentrating elements from large volumes of dilute solutions. It has been used for preirradiation (156, 215) and postirradiation (10, 216) separations in urine (216) and serum (10) analyses. However, since large volumes of serum cannot be drawn from an individual, the potentially extensive preconcentration of this technique cannot be achieved. The selectivity is useful for sodium separation if high decontamination factors can be obtained, but such group separations for counting with high-resolution detectors are better done using another technique. Electrophoresis has been used for NAA separations (217220) . Copper in serum has been determined this way (220) . Since metals in serum can be separated as their naturally occurring complexes, this technique may prove to be of great value in future serum trace element determinations. The restrictions of microgram sample size and reduced selectivity associated with electrophoretic focusing do not seriously affect serum determinations (221) . Dialysis . -During the dialysis of a serum sample all unbound ions, including all the Na , are removed. Protein bound concentrations rather than total concentration are thus obtained. Kanabrocki et al . have determined nondialyzable manganese and copper levels in saliva (222), urine (223), cerebrospinal fluid (224), and serum (14, 15). Cotzias et al . (61) have shown that serum proteins will combine with manganese

PAGE 64

55 in the dialysis solution and extract contaminants from the dialysis bag, thereby yielding results which are too high. The labile nature of the serum complexes, which can result in figures being low as well as high, requires that dialysis results be viewed with caution. Olivares (225) combined NAA and dialysis to determine the serum concentration of the metalcontaining complex selenocystathionine. In cases such as this, where it is the complex itself which is of interest, dialysis is quite useful. Gel filtration . -Serum trace elements can be separated as their naturally occurring protein complexes using gel filtration (226). NAA determinations thus obtained include protein-bound I (227), and As, Ag , Au, Cr and Sb (228). The values reported are semiquantitative and do not allow for complex dissociation due to dilution with eluent. Szilard-Chalmers effects . -Radiochemical separations using Szilard-Chalmers reactions take advantage of metal-ligand bond cleavage in chelates during (n,Y) reactions for isotope enrichment (229, 230). When applied to biological samples, enrichment factors as large as 10 for elements in dried blood and liver have been obtained (231, 232). Separation techniques such as electrophoresis, dialysis, gel filtration, and paper and thin layer chromatography require that the complicated protein complexes in which the trace elements are bound remain intact for postirradiation separation. Annealing to repair the extensive damage from SzilardChalmers reactions must accompany such separations. Little

PAGE 65

56 work on annealing protein complexes has been published, although Szilard-Chalmers effects have been noted (233) . The annealing of 6-diketonates has been discussed at length (234-239). Heat and gamma-radiation do some repair in the reactor during irradiation, but the greatest metal retention is obtained by heating after irradiation. Plots of retention versus annealing time at constant temperature reach plateaus, the heights of which are temperature dependent (237) . At higher temperatures retention quickly falls off due to decomposition (237). Szilard-Chalmers reactions, then, are both useful for some types of separation and a hindrance which must be compensated for in other separation procedures. Solvent extraction . -Highly selective extraction procedures have been reported for use in radiochemistry and other applications. The systems available are quite numerous. Books are available which discuss fundamental principles as well as specific methods (240, 241). Several areas which have received recent attention may be useful for trace element analysis in serum. Solvent extraction of metal chelates such as the B-diketonates discussed earlier, 8-mercaptoquinoline (22), and thenoyltrif luoroacetone (18, 242-245) is one such area. Another is the use of liquid ion exchangers such as di(2-ethylhexyl)orthophosphoric acid (246). Liquid ion exchangers combine the speed of solvent extraction with the selectivity of ion exchange. Direct elemental extraction without complexation has not received much attention, although it has been reported (247).

PAGE 66

57 NAA with solvent extraction has been used with biological samples including serum, with reported sodium decontamination factors of lO"* (248) . One major obstacle is that extraction conditions which allow extraction of a large number of elements often do not result in adequate sodium removal. For this reason, where both multielement analysis and high decontamination factors are needed, solvent extraction is often found as one of a series of separation steps. Inorganic separations . -Rapid, selective separations can be accomplished by passing sample mixtures over columns of various inorganic materials. Several possible reaction mechanisms exist: isotope exchange, redox reactions, ion exchange, and mixed crystal formation by recrystallization. In reviews of this topic (249, 250), Girardi reports both large decontamination factors and lower sample capacities than for organic ion exchangers. A survey of the NAA applicability of over 3,000 adsorption experiments on various ionic precipitates lists hydrated metal oxides among the more useful materials (251, 252). Hydrated Mn02 retains numerous ions completely (253), while hydrated tin oxide, SnOz , is mainly useful for ^^P retention. Hydrated antimony pentoxide (HAP) in highly acidic systems has been shown to remove only Na and Ta from a mixture of 60 elements (254). HAP and Sn02 have been used together for Cr determination in serum by NAA (64). The high selectivity and efficient sodium decontamination have been matched by isotopic exchange with a column of NaCl crystals using an organic eluent (255). The technique is

PAGE 67

58 rapid and selective, with sodium decontamination on the order of 10 and neglible retention of most other elements. Magnesium removal from biological matrices with a decontamination factor 3 of 10 using MgNH^POi, • 6H2O (96) illustrates extension of the principle where suitable eluents and solid forms of the element involved are available. Amalgam exchange with liquid mercury has been used by Kim and Silverman to determine mercury in various matrices (256, 257). Mercury has been determined in whole blood using this technique (258, 259). Amalgam exchange has been used to remove radioactive Bi, Cd, In, Pb, Sr, Ti , Zn and Cu from aqueous solutions (260, 261). Equilibrium is attained quickly, and excellent decontamination can be achieved at trace levels, making this technique suitable for trace element analysis in serum. Unfortunately, one can work best with only one element at a time. Moreover, the technique is applicable to only a limited number of elements. Ion exchange . -The use of ion exchange resins in NAA is well established, with many routine procedures for both selective individual separations and sequential group separations having been reported. Ion exchange material can be divided into four groups: polymeric resins, celluloses such as carboxymethyl cellulose, inorganic exchangers such as zirconium phosphate, and liquid exchangers such as dinonylnaphtalene sulfonic acid (262) . Of these, the resins are the most widely used cation exchange material, with advantages including high capacity, hydrophilic

PAGE 68

59 gel structure v/ith easily reached functional groups, high resistance to degradation by acids and alkalies, and relatively unchanging ion exchange properties (263) . There are many different kinds and brands of resin available. For many separations the kind of resin used is not important. Those with the same functional group, often -SO3H, will behave similarly. One useful property of sulfonic acid exchangers on a polystyrene base is that these resins are much more resistant to radiation damage than the anion exchange resins usually employed (262). Chelating ion exchangers differ from ordinary ion exchangers in three respects. The first is higher selectivity due to dependence on the chelating group rather than ion size, charge, or other physical properties. The second difference is that the bond strengths are greater, 15-20 kcal/mole, as compared to 2-3 kcal/mole for electrostatic bonds in ordinary resins. The third difference is kinetic. The chelating resin exchange processes are slower, controlled by either particle diffusion or second order chemical reactions (264) . The strong bond between the metals and the chelating resins permits their extraction from very dilute solutions of high ionic strength (264) , such as seawater (265) and dissolved geological samples (266). Copper has been separated from biological matrices in this way (267) . This property also allov/s the extraction of complexed metal ions from a complexing medium. Metal ion removal from protein complexes such as those in serum has not been reported.

PAGE 69

60 There are many excellent publications describing ion exchange theory and technique. Massart's monograph (262) deals specifically with radiochemical cation separations and discusses both the selectivity and kinetics of ion exchange separations. Resin characteristics such as capacity (263) , particle diameter (268), cross linkage, porosity (269) and the effect of temperature on HETP (270) have also been treated. Special techniques dealing with matrix separation in trace NAA have been developed. Precipitation ion exchange, which uses the insolubility of m.atrix metal chlorides such as NaCl in the 12M eluent, is one such method (271) . It works well for matrices containing large amounts of Na, Ag , Ba , K and Sr. Another technique for matrix activity removal uses isotope exchange on a resin in a stable form of the cation. Breakthrough of the radioactive matrix is delayed while the radioactive trace elements are rapidly eluted (272) . Ion exchange resins lend themselves to automated multielement separations because the eluate from one column can already be in a form suitable for introduction into the next column. Systems using many columns in complex, branched separation schemes have been described (273-282). Ricq (273) has explored the use of superimposed columns, while Moiseev et_al. have described a system for determining 27 elements at trace levels in a silicon matrix (274). Similar systems designed for biological samples have been developed by Comar and LePoec (275) . Samsahl has described automated multiple column ion exchange and solvent extraction group separation systems for biological samples allowing determination of up

PAGE 70

61 to 40 elements (276-278). Speed is emphasized (277-279), as little as 5 min being required for the processing of four samples (278) . Of the numerous other automatic and semiautomatic group separation schemes developed, that of Morrison et al . (280, 281) is perhaps the most extensive and analyticallyuseful. It enabled measurement of 41 trace elements in the Apollo 11 lunar samples by IIAA after separation into six groups using ion exchange, solvent extraction, and HAP extraction. This degree of radiochemical purity is sufficient with high-resolution counting equipment (282). NAA with ion exchange separations has been used for the analysis of many different types of biological samples (225, 265, 283-287). Determination of various elements (5, 38, 57, 83), including I (76, 284-290) using iodine-specific resins (289) and a sequential group separation scheme for 12 elements (290) have been reported. Ion exchange membranes have also been used in biological analyses (291) . The characteristics of ion exchange resins, especially chelating resins, strongly suggest their application to serum trace element analysis by NAA. Paper chromatography . -Various chromatographic techniques for postirradiation separation in NAA have been reported, but separation before irradiation without diluting or disturbing the separated fractions can only be accomplished with paper as the support. The use of paper chromatography in NAA has been reviewed by Budzynski (292) . Studies involving separration and activation of metal ions (292-296) , complex molecules of biochemical interest (74, 297-302), and possible

PAGE 71

62 interferents in the chromatographic paper (303) have been reported. The number of elements involved in reported separations varies, with 7 (293, 294), 26 (295), and isolation from mixtures of 40 or more (296) being representative. The techniques used range from ordinary procedures (77) to the use of columns consisting of stacked ion exchange paper disks which provide sodium decontamination factors of 10** (304). Molecular separations in serum analysis offer an additional dimension for trace element metabolism study. Biological molecules or chelates of trace elements which lack atoms with desirable activation characteristics can be determined using derivative activation chromatography (297-300) . Serum molecules such as phospholipids (292) and iodine-containing thyroid hormones (74, 226), and protein complexes of activating metals yield quantitative and qualitative data for both the element activated and the protein complex (292) . This is possible because the unique preirradiation separation capability eliminates adverse effects from Szilard-Chalmers decomposition. For determinations above the trace level, interference from impurities in the paper is not significant (74, 303). The only major contaminants found are Na, Cl , and Mn. These can be effectively reduced by washing the paper. However, due to the tubular structure of the paper fibers, all interferents cannot be completely removed, thus jeopardizing the accuracy of trace determinations (292) . Other drawbacks include deterioration of the paper in the reactor and lack of resolution of complex mixtures due to inhomogeneity of the paper.

PAGE 72

63 The advantages of thin layer chromatography over paper chromatography include greater speed and more selectivity in the stationary phase. Separation of metal ions has been reported (305, 306) , but the application of thin layer chromatography to HAA has not been explored, probably due to the lack of feasibility of preirradiaiton separation. Gas chromatography . -Gas chromatographic separations are rapid, highly efficient, and can be carried out without the use of carriers. In addition, these separations can be optimized with respect to time, resolution, or sample size. This allows the application of the technique to short-lived isotopes, multicomponent mixtures and trace constituents. These separations are very useful in trace analysis since they in effect result in infinite resolution in the y -i^^Y spectrum. This allows the lowest limits of detection to be reached using Nal(Tl) detectors. Gas chromatography has been used fairly extensively in radiation chemistry (307) , but the application of the technique to the separation of photon-emitting radionuclides in neutron activation analysis has been reported infrequently. The application of gas chromatography to NAA v;as first reported by Cram and Brownlee (308-310) . They demonstrated the usefulness of gas chromatography for the separation of shortlived radioisotopes in high speed postirradiation separations, some as fast as 5 s, allowing measurement of 10.7 s ^ °F. The use of metal chelates extends the usefulness of gas chromatography to encompass many trace elements. As stated

PAGE 73

64 in the Gas Chromatography section, 6-diketonates offer the greatest potential, allowing more than three-quarters of the elements in the periodic table to be separated and eluted easily, quickly, and with high resolution. The inherent speed and selectivity of gas chromatography combined with the desirable physical properties and ease of synthesis of the metal chelates make the technique an excellent prospect for NAA separations. The use of a high-resolution y^^^Y spectrometer with a gas chromatograph constitutes the ultimate in specific detection for NAA, while the use of efficient Nal(Tl) scintillation detectors allows the lowest limits of detection to be reached.

PAGE 74

CHAPTER II EXPERIFiENTAL Physical Facilities and Equipment Nuclear Reactors UFTR.-The University of Florida Training Reactor (UFTR) is an Argonaut type, graphite reflected and water moderated, with 100 kW thermal maximum power. It v;as used in all the short irradiation, short decay experiments because of its close proximity to the separation and counting equipment. Flux data for the port used in this work appear in Table 11. GTRR.-The Georgia Tech Research Reactor (GTRR) is a CP-5 type, heterogeneous, heavy water moderated and cooled, fueled with highly enriched plates of aluminum.-uranium alloy. Higher thermal neutron flux combined with longer daily operation hours made the GTRR m.ore useful for long irradiations. Flux data for the ports used in this work appear in Table 11. Separation Equipment Lowtemperature dry asher . -Ashing was carried out using a model LTA-600 asher, Tracerlab-Richmond, Richmond, California. Ashing times of 4-12 hr at an rf pov/er level of 200 W were used. Gas chromatograph .-A Varian Model 1200-1 gas chromatograph with a hydrogen flame ionization detector, a linear 65

PAGE 75

66 Table ll. Irradiation Facilities Georgia Tech • Research Reactor University of Florida Training Reactor Power Level Cd Ratio Gamma Flux, R/hr Thermal Neutron Flux, n cm~^ s~ ^ 1 MW 2000:1 1 X 10^ R/hr GTRR V-21 2 X 10^ ^ GTRR V-37 7 X 10^^ GTRR V-4 3 5 X 10^ ^ 100 kW 1:1 2.4 X 10^ R/hr Center Vertical Port 1.8 X 10^2

PAGE 76

67 temperature progranuner , and a 1/4-in injection port was used in these studies. The system was all glass and Teflon to eliminate reactive surfaces on which the chelates could decompose. The column tubing was nominal 1/8-in o.d. {0.095-in i.d., 0.016-in wall thickness) Teflon (Alpha type, TFT-250/11) . The injection port liner was a 6-mm (approximately 0.25 in) o.d. by 14.5-cm Pyrex tube with the downstream end constricted to give a gas tight seal with the 1/8-in o.d. Teflon column forced into the constriction. A smaller 7.5-cra by 3-mm o.d. Pyrex liner fit inside the larger tube. One end of the liner v/as tapered to fit snugly inside the Teflon column 5 mm from the packing material. It was this liner which was removed for y^ay counting. The instrument was also equipped for injecting solid samples. For this type of injection the injection port liner was replaced by a solid sampling device. It consisted of a stainless steel tube and mating plunger which broke sealed 2-mra by 2-cm capillaries containing the sample inside the injection port in a flowing He stream (192) . No Swagelok fittings were used. Instead, the Teflon tubing was forced inside pieces of tapered Pyrex tubing to give a tight fit. The Pyrex was tapered from both ends, allowing the two Teflon pieces to touch each other, thus giving zero dead volume. Liquid injections were made with a 10-yl Hamilton syringe. For chromatographic peak area measurements, the electrometer output v;as connected to an Infotronics Model CRS 100

PAGE 77

68 Digital Readout System digital gas chromatographic peak integrator with Victor printer. Vapor-phase separation apparatus . -The first vapor-phase separation apparatus was simply the gas chromatograph with its column replaced by an empty 6-in piece of 1/8-in o.d. Teflon tubing. A hole in the oven wall allowed a 10-cm long, 5-mm o.d. Pyrex tube to extend from the room temperature atmosphere of the laboratory into the oven, where it was attached to the empty Teflon tube. Sufficiently volatile injected samples could thus pass through the system and condense in the Pyrex tube. A 100-yl syringe with a Teflontipped plunger (Precision Sampling Corp., Baton Rouge, Louisiana) v;as used for injection. The improved vapor-phase separation apparatus did not use the injection port for sample introduction. Instead, the 6-in Teflon tube in the first apparatus was replaced by a horizontal shallow "U-shaped" Pyrex tube, 8 cm long, 1 cm o.d., with a depression at the bottom to hold the chelate solution (Figure 2). Both the inlet and the outlet arms were horizontal. The He flowed through the tube carrying vapor from the mixture into the cooler Pyrex tube for trapping. Counting Geometry Three different types of counting geometry were constructed and their usefulness in counting chromatographically separated radioactive metal chelates evaluated. Except for the Pyrex traps, the effluent reached the counting geometry configurations via a 1/8-in o.d. Teflon tube enclosed in a 1/4-in o.d.

PAGE 78

69 CO 3 +J nJ u Id a, < G O •H P (0 u Pi to 0) (0 (0 x; I u o 0, nj > 0) •H

PAGE 79

70 copper tube wrapped v/ith heating wire and maintained at the injection port temperature. The tube passed through the oven wall and into the lead cave beside the instrument. This tube, the gas chromatograph, the lead cave and the improved reversible counting geometry are shov/n in Figure 3. Stainless steel spiral . -The stainless steel spiral is a flowthrough type counting geometry. The radioactive' sample is not trapped; it flows past the detector without stopping. The counting time in flow-through geometries can be increased by forming the tube into a spiral. The length of tube in front of the detector is increased, and the sample is in front of the detector for a longer period of time. If even more counting time is required, the carrier flow can be stopped and the entire separation and elution process temporarily interrupted. Under most circumstances, this has no deleterious effects on the separation of components still on the column. The spiral was constructed from a 57-in long piece of 0.125-in o.d. , 0.005-in wall thickness 316 stainless steel tubing (Superior Tube Company, Norristown, Pennsylvania) . It was wound in a flat spiral v/ith an outside diameter of 2.9 in and an inside diameter of 1.1 in. The spiral was wrapped as a unit with heating wire and asbestos tape. A thermocouple was placed inside at the center of the spiral. Charcoal traps . -The charcoal traps were constructed from 10-cm long, 5-mm o.d. Pyrex tube tapered at one end to fit tightly inside the Teflon tube. A 1-in section of the trap was packed v;ith 30-50 mesh activated charcoal. The

PAGE 80

4J H -H Eh ^ (0 > (0 ^ U >i u Q) e :] O 0) c to tn C X!-H a-P «J G U 3 Cn O O U P (d 0) en ox> ^ ra u u itJ 0) 0) M 3 tn J

PAGE 81

72

PAGE 82

73 inlet to the trap was located inside the column oven. The rest of the tube extended through a hole in the oven wall. The outlet and most of the charcoal were at room temperature. A 5-in piece of thin-walled stainless steel tubing packed with a 1-in bed of 40-50 mesh activated charcoal was also used. Reversible counting geometry . -A simple reversible counting geometry and a more refined version were constructed. The simple geometry consisted of a 6-in long, 0.25-in o.d. piece of Teflon tubing packed with 2.5 in of 20% SE-30 coated on acid washed, DMCS treated Chromosorb P. This tube was enclosed in a larger aluminum tube wrapped with heating wire. Cool air could be passed around the Teflon tube inside the aluminum tube to cool the packing prior to trapping. Purge He could be introduced at the upstream end to empty the geometry while it was being heated by heating wire. The improved geometry. Figure 4, had a better design than the geometry described above. A 1-in long segment in the center of a 13-in long, 0.25-in o.d. Teflon tube was packed with 15% SE-30 on 80-100 mesh High Performance Chromosorb W (Hewlett-Packard) . The packed region was located at the midpoint of a 1-in radius 180° bend. The center of curvature of the bend coincided with the axes of the two cylindrical Nal scintillation detectors. This design allov/ed a radioactive sample to give the same detector response at every point along the bend.

PAGE 83

74 >i u -P Q) s o Q) O c H +J c o u H w u 0) > o (U > o a e 0) >-( tn -H fa

PAGE 84

75 The Teflon tube and packing material were enclosed in a 0.332-in o.d., 0.0125-in wall seamless Hastelloy tube (Superior Tube Company, Norristown, Pennsylvania). The Hastelloy had special brass fittings on each end which enclosed 1/4-to 1/8-in unions made of Pyrex. They joined the 1/4-in Teflon geometry tube and the 1/8-in Teflon inlet and outlet tubes. The brass fittings joined the Hastelloy and the 1/4-in copper tubes which enclosed the 1/8-in Teflon tubes. The Teflon was enclosed in metal to ensure uniform heating of the system to prevent chelate condensation or decomposition. The metal tubes were wrapped with asbestos-insulated #24 Chromel heating wire. The entire system was wrapped with asbestos tape to provide additional insulation and thermal stability. Carbon dioxide with a pressure of 2 psi could be forced through a 0.025-in orifice into the space between the Hastelloy and the 1/4-in Teflon tube at the 1-in packed region to cool the packing. This created a large thermal gradient along the packing material. The inlet was operated at 150 °C, while the outlet was operated at 60 °C. The chelate is trapped by the 90 °C gradient in the packing. The CO2 was vented through the outlet arm of the geometry. A special sleeve in the inlet arm prevented CO2 from being forced back and cooling the inlet. The inlet and outlet heat were controlled independently, allowing inlet heat for trapping, no heat for Y~^a.y counting, and both inlet and outlet heat for purging. The He flow network between the two-stage pressure regulator and the injection port v;as modified to allow rapid

PAGE 85

76 shutoff of the carrier gas supply to the column for stop flow experiments. This was done with a normally open electric valve (Skinner type V15DAil25V09) . A normally closed electric valve (Skinner type V52DA1125V18) and a precision metering valve (Nupro, Inc.) were installed to allow bypassing the column with carrier gas to purge the counting geometry and its connecting tubing with a carefully regulated He flov;. Counting and Data Processing Equipment High-resolution y-ray spectrometer . -A 50-cc active volume, v;rap-around coaxial design, high resolution lithiumdrifted germanium detector (Nuclear Diodes, Prairie View, Illinois) was used. The manufacturers specifications were as follows: resolution 2.3 keV at 1.33 MeV; peak-to-compton ratio 23:1; efficiency relative to 3 x 3 sodium iodide detector 12.5%. An Ortec Model 456 high voltage power supply (Ortec, Oak Ridge, Tennessee) was used to bias the detector at 2500 V. The output from a Nuclear Diodes Model 103 preamplifer and an Ortec Model 451 spectroscopy amplifier was digitized by a Wilkinson type analog-to-digital converter. Northern Scientific Model NS 629 (Northern Scientific, Middleton, Wisconsin) . The ADC was capable of 8192 channel resolution and operation at a 50 MHz clock rate. The ADC was connected through an interface to a PDP-8/L minicomputer (Digital Equipment Corp., Maynard, Massachusetts) with 8K of available core storage. The 12 bit vrords, when

PAGE 86

77 double precision storage was used in the upper 4K of core, allowed the use of 2048 channels, each with a 16,777K count capacity. Input-output devices used with the computer system included a Model ASR-33 teletype (Teletype Corp., Skokie, Illinois) , an ITT Model 1935D 15-in display Oscillscope on which spectra were displayed, a Hewlett-Packard Model 7127A strip chart recorder with a multiple range input module (Hewlett-Packard, Pasadena, California) with which the spectra were plotted, and a Tri-Data Model 4096 magnetic tape cartridge unit with which all bulk storage, both spectra and software, was done. The interface and software for the spectrometer system were designed and assembled by David B. Cottrell. A more complete description of the system can be found elsewhere (311) . Nal(Tl) scintillation spectrometer . -Two detectors were used. Each was a 3-in by 3-in thallium-activated sodium iodide crystal with mating multiplier phototube (Harshaw Integral Line Detector, Nuclear Chicago Model 10-9, Nuclear Chicago Corp., Des Plaines, Illinois) connected to a scintillation preamplifier (Nuclear Chicago 10-17) and spectroscopy amplifier (Nuclear Chicago Model 27001) . The single channel analyzer (Nuclear Chicago Model 27352) output was counted with a preset count six decade electronic scaler (Nuclear Chicago Model 27104) for gross y -i^ay measurements. They -ray spectra were recorded using a 400-channel Nuclear Chicago RIDL 34-27 Series Scientific Analyzer System

PAGE 87

78 with a Model 30-35 Four Input Mixer/General Purpose Amplifier, a Model 23-4 Analog-to-Digital Converter with Buffer Register, and a Model 52-57 Deluxe Display Control. Also included in the system were a display oscilloscope (Nuclear Chicago Model 52-56), a Teletype (ASR-33, Nuclear Chicago Model 44-29), and a five decade preset timer (Nuclear Chicago Model 27101) . Plots of the spectra were made with a Plotamatic 850 X-Y Plotter (Data Equipment Corp., Santa Ana, CA) . Lead shielding . -All of the 400-channel Nal(Tl) scintillation spectra were measured inside a lead cave with 40-in x 12-in X 14-in inside dimensions and 4-in thick walls on the top and four sides. The bottom was 2 in thick. The Nal (Tl) detectors were mounted horizontally with the metal chelate effluent passing between them. The high-resolution Ge(Li) detector, having lower sensitivity, required less shielding than the Nal(Tl) detectors. It was housed in a cave with 12-in x 16-in x 16-in inside dimensions and 2-in walls on four sides and the top. The bottom was unshielded. Procedures Ashing Serum samples came from a human serum pool drawn and processed in the normal course of activities at the Clinical Laboratories at the J. Hillis Miller Health Center, University of Florida. All glassware with which the serum came in contact before irradiation was cleaned by washing with soap and v/ater.

PAGE 88

79 soaking in Nochromix (a metal-free substitute for dichromate in sulfuric acid, Godax Laboratories, New York, New York) for 12 hr, soaking in concentrated KNO3 for 12 hr and then leaching in demineralized water for 12 hr. The glassware was rinsed in demineralized water after each step. The seriam was freezedried in sample boats in the ashing chambers. This was accomplished by freezing the serum v;ith dry ice outside the chamber and reducing the pressure to 1 Torr immediately after placing the frozen serum in the chamber. Ashing periods from 4 to 12 hr at an rf power level of 200 W were used. Chelation from Solid Ash Three serum samples were processed using different procedures for separation and irradiation. The first sample was subjected to a simple separation before a long thermal neutron irradiation. The second sample was processed in a similar way, except that the separation scheme was more complex. The third sample was irradiated for a short time prior to separation. Irradiation . -The solid residue and supernatant liquid from the first serum sample were irradiated together in V-21 of the GTRR for 49.5 hr. The solid residue, supernatant liquid, and blank solution from the second serum sample were irradiated together in V-21 of the GTRR for 42.5 hr. The ash from the third sample was irradiated in the UFTR for 30 min. Gamma-ray counting . -Counting v/as done with the highresolution spectrometry system. The organic layer from the first sample decayed 26 days before counting. The solid

PAGE 89

80 residue decayed 2 days before counting. The organic layer and blank from the second sample decayed four days before counting. The solid residue decayed two weeks before counting. The third sample decayed one hour before counting. Chelate Extraction from Ash Solution After dissolving the ashed irradiated serum, carrier v;as added and the chelates of the trace elements in the aqueous layer were extracted into a benzene solution of the ligands H(fod) and H(tfa) , leaving the sodium in the aqueous layer. Purification of the metal chelates in the vapor phase provided still further isolation from sodium. One simple procedure and three other procedures using a complex carrier solution were used. Carrier preparation . -The carrier used in the simple procedure consisted of 6 mg of Cs (1103)3, 6 mg of LazOs, 4 rag of CUCI2, and 2 mg of SmzOs. These salts were added to the ash as solids. The makeup of the carrier solution used in the multielement procedures for the other three samples is shown in Table 12. The weight of each compound in the complex mixture was chosen to yield 0.01 mmole of element per milliliter of final carrier solution. The metals and metal salts were dissolved, mixed, and the mixture was diluted to 2 5 ml. Irradiation . -Four 6.5-g serum samples were ashed and irradiated separately for 30 min in the UFTR. Chelation . -Each of the four samples was processed immediately after it was removed from the reactor. The simple carrier procedure used in the first extraction

PAGE 90

81 Table 12. Complex Carrier Solution As Hf. Au Hg Ba In Cd La Co Mn Cr Ni Cs Pd Cu Sc Dy Se Eu Sr Ga Zn

PAGE 91

82 experiment consisted of adding 0.5 ml of a benzene solution containing 50 rag/ml each of the ligands H(fod) and H(tfa) to the irradiated ash, stirring, and allowing the mixture to react for 10 min. The supernatant liguid was decanted. Next, the ash was dissolved in 0.5 ml of 16M HNO 3 and . 2 ml of deionized water, and the solid carrier mixture was added to the dissolved ash solution. Some solid residue remained after the addition. The pH was adjusted to 6 with 2M NH4OH. A 0.75-ml portion of the benzene-ligand solution was then added and the mixture was shaken for 30 sec. After allowing 10 min for the mixture to separate, the organic layer was removed. The aqueous portion was washed with 0.2 ml of benzene and all three benzene portions were combined for separation and counting. A similar procedure with a complex multielement carrier solution was used for the next three ashed serum samples. After irradiation, the sample A ash was dissolved in 0.5 ml of 2M HNO 3, 1 ml of carrier solution was added, and the pH was adjusted to 6 with 2M NH^OH. One m.illiliter of the benzene-ligand solution was added, the mixture was shaken, and then it was allowed to equilibrate 5 min. After another shaking and 5 min equilibration the pale blue benzene layer was removed. The pH was then adjusted to 4, the extraction was repeated, and the two organic solutions were combined for counting. Sample B was processed in nearly the same way. The organic solution was added before the 1-ml carrier addition

PAGE 92

83 and subsequent pH adjustment to 6 with 2M KOH. After all the components had been added, the mixture was centrifuged and the pale yellow organic layer v/as removed. The radioactive ash from serum sample C was dissolved in 1 ml of 2M HNO3 and 1 ml of the carrier solution was added. The pH was adjusted to 6 with 2M KOH and the metal hydroxide precipitate was removed by centrif ugation. The supernatant was extracted with 1 ml benzene-ligand solution. The aqueous layer and the precipitate were then recombined and the pH was adjusted to 4 with 2M HNO3. The precipitate remained 30% undissolved at this point. The solution v;as extracted with 1 ml of the benzene-ligand solution and the two extracts were combined for counting. The remaining solid v/as purified by reprecipitation prior to counting. Vapor-phase separation . -The extract from the first experiment v/as injected into the first vapor-phase separation apparatus in 30-^1 segments 1 min apart. After 0.75 ml had been injected, the oven and injection port temperatures were increased to 200 °C over a 15_min interval. The flow was then stopped and the charcoal trap was removed for counting. The next three samples. A, B, and C, were processed using the improved vapor-phase separation apparatus. The apparatus was operated for sample A at 100 °C for 7 min, 150 °C for 5 min, 175 °C for 5 min, and 200°C for 10 min. The He flow rate v/as 20 ml/min. Samples B and C were processed at 100 °C for 3 min, 140 °C for 3 min, 170 °C for 3 min, and

PAGE 93

84 200 °C for 10 min. The He flow rate was 75 ml/rain. The purpose of the multiple temperature vaporization was to allow the lower vapor pressure chelates to vaporize without risking thermal decomposition of the higher vapor pressure chelates, which leave the oven before the temperature gets too high. Gamma-ray counting . -Counting was done with the highresolution spectrometry system. The radiation from the ash was measured for 1 min prior to solvent extraction. A 4-MeV full scale Y-ray spectrum was recorded. After extraction, and before vaporization, a 2-MeV full scale y-ray spectrum of the benzene solution containing the chelates was recorded. After separation, several spectra of the charcoal traps containing the radioactive metal chelates were recorded. Annealing Studies The extent of Szilard-Chalraers degradation in both treated and untreated thermal neutron-irradiated benzene solutions of some metal chelates was determined. The solutions were injected into the gas chromatograph and the y-ray spectra of the material trapped at the outlet were compared with the spectra of the material retained by the various parts of the instrument with v^hich the sample came in contact. Chelates with no treatment . -Three metal chelate solutions were prepared. The first contained 32.5 mg of Cf(tfa)3, the second contained 7.0 mg of Cr(hfa)3, and the third contained 30.0 mg of Cr(fod)3. Each chelate was dissolved in 1 ml of benzene (Nanograde, Mallinckrodt Chemical Works) which had been dried over molecular sieve (13x Linde) . The three solutions were irradiated in V-43 of the GTRR for 14.5 hr.

PAGE 94

85 The gas chromatographic columns and conditions used for retention measurement are listed as No. 1, Table 13. They were conditioned overnight at 170 °'C and treated with 10 yl of DMCS (Dimethyldichlorosilane, PCR, Inc., Gainesville, FL) . The injection port liner was included in the silanization. The solutions were injected in 5 yl portions 1 min apart. A total of 100 yl each of Cr(tfa)3 and Cr(fod)3, and 150 yl of Cr(hfa)3 was injected. After each of the three series of injections, the carrier gas and oven heat were left on for 30 min before dismantling the system. All counting time periods were 10 hr except for column measurements which were 15 and 20 hr. Chelates with treatment . -Three treatment procedures were evaluated for rechelating the metal ions which had been separated from their ligands by recoil from the complex after neutron capture. The procedures were irradiation with 10% excess ligand, irradiation with 10% excess ligand followed by annealing (heating) , and irradiation with 100% excess ligand followed by annealing. These evaluations were all done with Cr (tfa) 3. Two solutions, one containing 30.2 mg (Cr(tfa)3 and 2.7 mg of H(tfa) (10% excess) in 1 ml of benzene, and one containing 30.4 mg of Cr(tfa)3 and 27 mg of H(tfa) (100% excess) in 1 ml of benzene, were irradiated in V-43 of the GTRR for 14 hr . After irradiation, 20 injections, each 5yl of the 10% excess ligand sample, were made into the gas chromatograph . After dismantling the instrument and measuring the radio

PAGE 95

86 activity of its various components, the remainder of the sample was resealed, annealed 9 hr at 100 °C, and the experiment repeated. The same injection and counting procedure was followed with the 100% excess ligand sample after it had been annealed 9 hr at 100 °C beginning immediately after irradiation. All chromatographic conditions were the same as those in the previous section describing chelates without annealing. Gas Chromatographic Studies The extent to which various radioactive metal chelates of l,l,l-trifluoro-2,4-pentanedione [H(tfa)], 1,1,1,5,5,5hexafluoro-2,4-pentanedione [H(hfa)], and 1,1,1,2,2,3,3heptaf luoro-7, 7-dimethyl-4 ,6-octanedione [H(fod)] were quantitatively eluted from a gas chromatograph was studied. Each tfa and fod metal chelate was synthesized from radioactive metal and dissolved in benzene. The hfa chelates were purchased. They were irradiated and injected as solids without any chemical processing or purification. The chelate solution (or solid sample) was injected into the gas chromatograph and the chelate eluted and trapped. After trapping the chelate, the system was dismantled, and the packing was placed in a 60by 150-mm Pyrex petri dish. Separate radiation measurements of the metal chelate remaining in the solid sampler, injection port liner, column tubing, column packing, and charcoal trap were made. Where solid samples had been used, the quartz capillary fragments in the injection port

PAGE 96

87 were counted and the injection port was rinsed with acetone and the solution was. counted. Materials . -The H(tfa) was redistilled and the 106-107 °C fraction retained. The H(fod) was also redistilled. Both were obtained from PCR, Inc., Gainesville, Florida. The Cu(hfa)2, Cr(hfa)3, and Mn(hfa)2 were obtained from Pierce Chemical, Rockford, Illinois. The radioactive isotopes used were produced in the UFTR from commonly available reagents, with the exception of the following, which were purchased from New England Nuclear, Boston, Massachusetts: ^ i^^^^ -^^ ^^.-j^ ^ q^q2 rng/ml , 3.7 mCi/ml; carrier-free ^Be in 0.5M HCl, 4 mCi/ml; and ^ ^Fe in IM HCl, 0.165 mg/ml, 2.42 mCi/ml . The benzene (Nanograde, Mallickrodt Chemical Works) was dried over molecular sieve (13 X Linde) . The system was silanized using dimethyldichlorosilane (DMCS, PCR, Inc.). Chelate synthesis Cr (tfa) 3 .-The radioactive metal chelate was prepared by dissolving 40 mg of CrCl3'6H20 in 0.6 ml of H2O. Then 0.9 ml of ^^Cr solution, 300 mg of urea, and 86 mg of H(tfa) were added. The mixture was refluxed for 7 hr at 100-115 °C. The solid product was washed with water and air dried on filter paper. It was then dissolved in ether and sublimed at 150 °C. The improved vapor-phase separation apparatus was used for purifying the chelate. Dark purple Cr (tfa) 3 crystals weighing 47 mg were recovered and redissolved in 1.5 ml of dry Nanograde benzene for chromatographic injection.

PAGE 97

88 Cr (fod) 3 .-The radioactive metal chelate was prepared by dissolving 52.5 mg of Cr (NO 3 ) 3 • 9H2O and 0.3 ml of CrCl 3 solution containing ^^Cr in 11 ml of absolute ethanol and removing the H2O by boiling until only 0.5 ml liquid remained. Next 134.2 mg of H(fod) in 1 ml of absolute ethanol was added and the mixture was heated for 1 hr at 80 °C. The solution was transferred to the vapor-phase separation apparatus, the ethanol was evaporated, and the chelate was sublimed at 140 °C. The 6.5-mg purple crystalline product was dissolved in 0.2 ml of dry Nanograde benzene for chromatographic injection. Be (tfa) 2 . -The radioactive metal chelate was prepared using a procedure similar to that of Berg and Truemper (312) . A 41-mg portion of BeS04*4 H2O and 90.0 mg of NaAc were dissolved in 1.5 ml of H2O. To this 0.25 ml of ^Be solution and H(tfa) dissolved in 1 ml absolute ethanol were added. The Be (tfa) 2 was extracted using dry Nanograde benzene. The product was purified by vacuum sublimation at 65 °C. The pure radioactive chelate weighed 34.5 mg. It was dissolved in 1 ml of dry Nanograde benzene for chromatographic injection . Fe (fod) 3 . -The radioactive metal chelate was prepared by dissolving 1 mg of FeCls in ethanol and adding ^^Fe solution and H(fod). The deep red chelate which formed was extracted into 2.5 ml of hexane . The hexane solution was placed in the improved vapor-phase separation apparatus, the hexane was evaporated, and the chelate v/as sublimed at 14 °C,

PAGE 98

89 The 6-rag deep red product was dissolved in 0.4 ml of dry Nanograde benzene for chromatographic injection. The lower melting pale red portion was discarded. Cu (tf a) 2 .-The radioactive metal chelate was prepared using a procedure similar to that of Berg and Truemper (312) . A 140.6 mg portion of CUCI2 and 141. 5 -mg of NaAc were dissolved in 10 ml of H2O. Next 305.5 mg of H(tfa) in 1 ml of ethanol was added. The insoluble product was extracted with dry Nanograde benzene and purified by vacuum sublimation at 115 °C. A 65.5-mg portion of the pure Cu(tfa)2 was added to a mixture of 2 ml of dry Nanograde benzene and 65 mg of H(tfa). The solution was sealed in an ampoule and irradiated for 7.5 hr in the UFTR. The radiation damage to the chelate was annealed at 115 °C for 7 hr before opening the sealed irradiation ampoule. Next the chelate was purified by vacuum sublimation at 115 °C. A 54.4 -mg yield of radioactive Cu(tfa)2 was obtained. The pure chelate was dissolved in 1 ml of dry Nanograde benzene for chromatographic injection. Gd (f od) 3 . -The radioactive metal chelate was prepared by dissolving 89.3 mg of Gd(fod)3 87.5 mg of H(fod) in 3 ml of dry Nanograde benzene. One milliliter of this solution was irradiated in a sealed ampoule in port V-43 of the GTRR for 12.33 hr. The radiation damage to the chelate was annealed for 9.5 hr at 100 °C before opening the sealed irradiation ampoule. Vacuum sublimation at 150 °C yielded a yellow oil weighing 15 mg . This pure radioactive chelate was dissolved in 1 ml dry ::anograde benzene for chromatographic injection.

PAGE 99

90 Lu (fod) 3 .-The radioactive metal chelate was prepared by sealing 43.5 mg of Lu(fod)3 and 42.1 mg of H(fod) in 1.5 ml of dry Nanograde benzene in an ampoule and irradiating in port V-43 of the GTRR for 14 hr . The radiation damage to the chelate was annealed at 100 °C for 10 hr before opening the sealed irradiation ampoule. A 33.5-mg yield was obtained from vacuum sublimation at 115 °C. The purified chelate was dissolved in 0.5 ml of dry Nanograde benzene for chromatographic injection. Gas chromatography Cr (tfa) 3 . -The column, No. 1, Table 13, was conditioned overnight at 175 °C, treated with 10 yl of DMCS, and conditioned two more hours. Ten 5-yl injections of the radioactive Cr(tfa)3 solution were made 1 min apart. The carrier flow and oven heat were kept on for 30 min after the last injection before dismantling the system for counting. The chelate solution was injected into three identical columns. Cr(fod) 3 . -The column, No. 1, Table 13, was conditioned overnight at 175 °C, treated with 10 yl of DMCS, and conditioned two more hours. Ten 5-yl injections of the radioactive Cr(fod)3 solution were made 1 min apart. The carrier flow and oven heat were kept on for 30 min after the last injection before the system was dismantled for counting. Be (tfa) 2 .-The gas chromatographic conditions were taken from Ross and Sievers (181). After conditioning column No. 2, Table 13, for 30 min at 80 °C, twenty 5-pl injections of the radioactive Be (f ta) 2 solution were made 1 min apart.

PAGE 100

91 u -P o
PAGE 101

92 m

PAGE 102

93 The carrier flow and oven heat were left on for 30 min after the last injection before dismantling the system for counting, Injections were made into three identical columns. Fe (fod) 3 .-The column. No. 3, Table 13, was conditioned at 200 °C for 15 min. Five 5-yl samples of the chelate solution were injected into the gas chromatograph 1 min apart, The carrier flow and oven heat were kept on for 45 min after the last injection before dismantling the system for Y~^3-y counting. Injections were made into three identical columns. Cu(tfa) 2 .-The gas chromatographic conditions were those used by Tanaka et al. (313). After conditioning column No. 4, Table 13, for 40 min at 100 °C, twenty 5-yl injections were made 1 min apart. After waiting 10 min after the last injection, the system was examined. The trap showed a well defined zone of blue chelate. Another less well defined blue region was seen through the Teflon tubing at the head of the column just beyond the heated injection port region. Presumably a cis trans isomeric separation had taken place. The column oven was then reheated to 13 °C to elute the remaining chelate. After 10 min this was accomplished and after an additional 15 min the oven heat and carrier flow were cut off and the system dismantled for counting. The chelate solution was injected in the same manner into two more columns identical to the first, except operated exclusively at 130 °C. Gd (fod) 3 . -The experiment was done with three columns, No. 5, Table 13 . One run was done with 10 injections of 5 yl

PAGE 103

94 each. Two runs were done with 20 injections of 5 yl each. There was a 1-min interval between injections. The column oven heat and carrier flow were left on for 15 min after the last injection before dismantling the system for counting. Lu (fod) 3 . -The column, No. 5, Table 13, was conditioned for 1 hr at 170 °C. Ten S-yl injections of the chelate solution were made 1 min apart. The oven heat and carrier flow were left on for 15 min after the last injection before dismantling the system for counting. Injections were made into three identical columns. Cu(hfa)2 and Cr(hfa)3 .-A 1-mg sample of each chelate was sealed in separate 2-mra o.d., 2-cm long quartz capillaries using a dry ice-acetone bath to cool the sample, thus preventing decomposition from the heat of the torch. The samples were irradiated for 4 hr in the UFTR. The gas chromatographic column used for these injections was No. 6, Table 13. The sample was injected; the chelate eluted, V7as trapped, and the system v;as dismantled for counting, Mn(hfa)2 . -A 1-mg solid sample of the chelate was sealed in a 2-mm o.d., 2-cm long quartz capillary and irradiated for 15 min in the UFTR. The sample was then injected as a solid into column No. 7, Table 13. After elution and trapping, the system was dismantled for counting. Chelate residue exchange An experiment was carried out to determine whether the incompletely eluted chelate residue which remained on the gas chromatographic column was inert to subsequently

PAGE 104

95 injected samples of the same chelate, or whether there was exchange between samples injected at different times. A 1.09-mg solid sample of Cu(hfa)2 was irradiated in the UFTR for 4 hr. It was then injected into column No. 8, Table 13 and eluted. Two hours later a similar solid sample of Cu(hfa)2 containing no radioactive material was injected into the same column and the effluent was trapped and counted using the Nal(Tl) scintillation spectrometer to see if any ^ '*Cu from the first sample were present. Silanization An experiment was carried out to determine whether silanization of the injection port and column would significantly decrease metal chelate decomposition in the gas chromatograph or increase the amount of chelate detected in a given injection. Two chelate solutions were prepared. One contained 29.0 mg of Cr(fod)3 in 1 ml of dry Nanograde benzene; the other contained 29.0 mg of Cr (tfa) 3 in 1 ml of dry Nanograde benzene. Dimethyldichlorosilane (DMCS, PCR, Inc., Gainesville, FL) was used neat. Two identical columns, No. 1, Table 13, were prepared. The column for the Cr(fod)3 study was conditioned at 190 °C. The column for the Cr(tfa)3 study was conditioned at 175 °C. A series of 1-pl injections was made using a l-yl syringe (Hamilton) . Then a DMCS injection (10-40 yl) was made using a 100-yl syringe with a Teflon tipped plunger (Precision Sampling Corp., Baton Rouge, Louisiana). After

PAGE 105

96 a short interval, typically 15 min, another series of 1-yl chelate injections and another DMCS injection were made. This was repeated three or four times until 70-80 yl of DMCS had been injected. A separate column was used for each of the chelates . Counting geometry operation Stainless steel spiral . -The operation of the geometry was evaluated using a chelate mixture consisting of 2.03 mg of Cr(hfa)3 and 4.72 mg of Cu(hfa)2. The mixture was sealed in a 2-mm o.d. , 2-cm long quartz capillary and irradiated in the UFTR for 4 hr. The solid mixture was injected onto column No. 6, Table 13. Elution was monitored using the Nal(Tl) scintillation spectrometer operating in the time sequenced store (TSS) mode. Each channel in the memory is assigned for storage sequentially; the number of input pulses which arrive during its assignment is stored. Thus a plot of gross counts versus elapsed time is obtained. The length of time each channel was active was 1.2 sec in this particular experiment. The solid sample injection and the beginning of TSS operation occurred at the same time. Charcoal traps . -A fresh trap was attached before each component was eluted. After elution, the flow was stopped, the trap changed, and elution continued. The trap was then transferred to the counting equipment for measurement. Simple reve rsible geometry . -The operation of the simple reversible geometry v/as evaluated by injecting a 2.26-mg solid

PAGE 106

97 sample of Cu(hfa)2 which has been irradiated in its injection capillary in the UFTR for 1 hr . The column used was No. 9, Table 13. The chelate was injected into the gas chroma tograph as the multichannel analyzer reached channel 45 in the TSS mode. The dwell time was 1.2 s per channel. The geometry temperature was 60 °C at injection, but was raised to 150 °C at channel 364 to release the chelate and purge the geometry. Improved reversible geometry .-The operation of the improved reversible geometry was evaluated in a similar fashion. Twelve milligrams of Cu(tfa)2 and 14 mg of H(tfa) were dissolved in 0.5 ml of benzene and irradiated in the UFTR for 7.5 hr. A 10-ul sample of this radioactive Cr(tfa)2 solution was injected onto column No. 10, Table 13. The geometry inlet was operated at 150 °C, the geometry outlet was operated at 40 °C, and the CO2 pressure was 20 psi. The multichannel analyzer was started in the TSS mode 6.5 min after sample injection. The dwell time was 1.2 s. After 45 min, the geometry purge was begun; the CO2 flow was stopped, the geometry outlet heat was turned on, and another TSS mode measurement was begun. Ion Exchange A chelating ion exchange resin, Chelex-100 (Bio-Rad Laboratories, Richmond, California), was used in separation studies dealing with activation analysis of metal chelates in human serum. These studies included both preirradiation and poGtirradiation separations, along with both long and

PAGE 107

98 short irradiation periods. The amount of water required for optimal sodium removal was also determined. Postirradiation sepa ration Sodium separation optimization . -A 7 . 0-g serum sample was ashed and irradiated in the UFTR for 30 min. After irradiation, the ash was dissolved in 1 ml of IM HNO3 and the pH was adjusted to 7 with 15M NH^^OH. The Chelex-100 was washed with deionized water and packed in a 0.8-cm i.d, Pyrex tube to a height of 6 . cm. The column was eluted with 2 ml of 2M HNO3 and washed with 60 ml of deionized water. The radioactive ash solution was placed on the column along with 3 ml of deionized water and allowed to soak in for 5 min. Next the column was rinsed with 30 ml of deionized water at . 5/min to remove the sodium. Each 5-ml portion of eluate v/as collected separately and its radioactivity, primarily due to ^''Na, was measured using the Ge(Li) detector and 400 channel analyzer. Trace element determination After sodium removal, the resin, which contained the trace elements, was placed in a counting vial and a y-ray spectrum was recorded. Preirradiation separation Short irradiation . -The final location of the trace elements in the separation scheme was determined by analyzing the various system components: the resin before use, a solution blank, the eluate from the resin, and the resin used for the separation.

PAGE 108

99 Resin blank . -The ion exchange column was prepared as described above and washed with 50 ml of deionized water, eluted with 50 ml of 2M HNO3, and washed again with 50 ml of deionized water. At this point the column was ready for sample introduction. The top 8 0% of the resin was transferred to a quartz ampoule and sealed. Solution blank . -The ion exchange column was prepared as described above. A blank was prepared by mixing . 5 ml of 2M HNO3, 0.5 ml of 2M KOH, and 2 drops of 15M NH^OH and placing the mixture on the column. The mixture was allowed to soak in for 2 min. The column was then rinsed with 3 ml of deionized water and eluted with 30 ml of 2M HNO3. All flow rates were 0.5 ml^in. The eluate was evaporated to 1 ml on a hotplate and sealed in a quartz ampoule. Serum solution . -The ion exchange column used for the solution blank was rinsed with 50 ml of deionized water to prepare it for the serum ash solution. A 7 . -g serum sample was ashed in the low temperature dry asher for 12 hr at an rf power level of 200 W. The ash was dissolved in 1 ml of 2M HNO3; the pH was adjusted to 7 with 2M KOH. A small amount of undissolved ash remained. The solution was decanted and placed on the column. The solids were dissolved, the solution was neutralized, and it was placed on the column along with 2 ml of deionized water. The mixture was allov/ed to soak in for 2 min. The column was then rinsed with 30 ml of deionized water and eluted with 25 ml of 2M HNO3. The

PAGE 109

100 flow rates were 0.5 ml/min. The eluate was evaporated to 1 ml on a hotplate and sealed in a quartz ampoule. Finally, 80% of the ion exchange resin was removed from the column and sealed in a quartz ampoule. Irradiation and counting . -All four ampoules were irradiated for 3 hr in the UFTR. Their y-ray spectra were recorded using the high-resolution Y-ray spectrometer system. Each of the four samples was counted within 4 hr of the irradiation. Long irradiation . -The procedure was essentially the same as for the short irradiation, except that triplicate serum samples v/ere included and no resin blank was processed. The resin preparation was described above. Each serum sample weighed 6.2 g. The ash was dissolved in 0.8 ml of 2M HNO3, and the pH was adjusted to 7 with 15M NH.*OH. The solution was placed on the Chelex-100 column and allowed to soak in for 2 min. The resin was rinsed with 25 ml of deionized water, eluted with 2 5 ml of 2M HNO3, and the eluate was evaporated to 1 ml on a hotplate. The eluate, the blank, and 80% of the resin were then sealed in separate quartz ampoules. All five were irradiated in V-37 of the GTRR for 52.1 hr . After irradiation, the samples were removed from the irradiation ampoules and placed in glass counting vials. The Y-ray spectra were recorded using the high resolution Y-ray spectrometer system described earlier.

PAGE 110

CHAPTER III RESULTS AND DISCUSSION Objectives The overall objective of this research is to expand the usefulness of NAA for trace analysis by developing new separation methods. These methods are in the areas of chelation, gas chromatography, and ion exchange. Work in these areas is unified by the goal of improved serum analysis, a powerful tool in medical and biochemical research. The specific objective of the chelation studies is the demonstration of the principles involved in using chelates for trace element separation from matrices such as serum quantitatively, rapidly, reproducibly , and in a form suitable for further purification and resolution by gas chromatography. Efficient application of this technique to NAA requires the judicious choice of ligands, experimental conditions, and the proper sequence of operations, all of which are affected by both the elements of interest and the external demands on the analysis. The objectives of the gas chromatographic studies are the demonstration and optimization of metal chelate separation as it applies to NAA. Goals of maximum resolution and counting efficiency, complete sample elution, ease of operation, and speed of analysis are sought through studying various methods 101

PAGE 111

102 sample introduction and determining the elution characteristics of the chelates, with particular attention being paid to the extent to which each sample is eluted. Vapor and solid phase y-ray counting devices must also be developed as an integral part of the gas chromatographic system to facilitate such analyses. The specific objective of the ion exchange studies is the evaluation of chelating ion exchange resins as separation tools in trace element analysis in matrices such as seriom. This includes optimizing conditions for sodium separation and evaluating subsequent trace element recovery. Chelation Studies The need for chemical separation in serum trace element analysis by NAA is apparent when y-^ay spectra recorded at various stages of an analysis are compared. Figure 5 from the first extraction experiment shows a 1-min yi^^Y spectrum of untreated serum ash immediately after irradiation. The radiation from the ^ "^Na and ^ ®C1 obscures all other photopeaks. Figure 6 shows a 1-min Y~i^ay spectrum of the combined chelate solutions after direct chelation followed by chelation-extraction was performed on the irradiated ash. The interfering activity is greatly reduced, but ^ "^Na and ^ ®C1 still predominate. Figure 7 shows a 1-hr yi^^y spectrum of the volatile material trapped during vapor-phase separation of the chelate mixture. Some of the trace element photopeaks are now visible It is clear, then, that under these irradiation and decay conditions, chemical separation of some sort is needed.

PAGE 112

0) u en •H Em

PAGE 113

104 01/ 3inNll^ d3d siNnoo

PAGE 114

0) u •H

PAGE 115

10. 8C DNfrZ 106 O o O in CM liJ — UJ o So NOIlVllHrNNV NOdllSOd O If) 6 00 ID ^01/ ainNitM a3d in CM d

PAGE 116

a o •H 4J ta u (h Ui (U (0 nJ s: I u o a > u Q) -P < Q) -P (0 rH •H

PAGE 117

108 10, 8£ NOIiVllHINNV uounsod O q Cvj IT) o lO in

PAGE 118

109 Chelation provides an effective means of separating metal ions which form stable chelates from other ions, such as sodium, which do not. The separation makes use of the change from ion to neutral complex during chelation. The low solubility of matrix ions in the organic solvent containing the chelates yields high separation factors. Many ligands have been used successfully for this type of separation, but ligands which form volatile metal chelates can undergo further purification in the vapor phase. This type of separation can be used with many elements; B-diketone chelates of about 60 metals have been reported (314) . The volatility and thermal stability of the f luorine-substitited chelate allows even further sodium decontamination, since whatever sodium-containing material might be extracted will most likely not be sufficiently volatile to interfere in vapor-phase separation techniques. This is illustrated in Figures 5-7, where a sodium decontamination factor of about 1000 is shown. H(tfa) and H(fod) were chosen for reasons discussed in the Introduction: ease of chelate formation, lack of hydrolysis, excellent thermal stability, and when used together, ability to form complexes with a large nuiT\ber of elements. Chelation Procedure The low temperature dry asher used in these studies provided complete decomposition and oxidation of the dried serum samples in 12-14 hr with no apparent sample losses. Freeze drying was accomplished by placing frozen serum in the asher and immediately reducing the pressure to 1 Torr.

PAGE 119

110 This eliminated foaming with accoiupaning sample loss, since only sublimation from the solid surface, not boiling, could take place. This evaporation combined with the low thermal conductivity of the low pressure atmosphere surrounding the serum to keep it frozen until all the water had been removed. A 6-ml serum sample yields about 4 mg of white, hygroscopic ash, a convenient size for processing. This ash can be dissolved, the metal chelates being subsequently synthesized and isolated, or a chelation procedure using a direct reaction of the solid ash with the organic ligand solution can be used. Chelation from solid ash Advantages . -This direct approach, like the microreactor technique described by Sievers et al . for forming fod chelates directly from solid samples (192) , has the advantages of being simple and rapid. In the preirradiation separations often required when long irradiations are used, additional sources of contamination are introduced by ash dissolution. In addition, since the extraction efficiencies of some metal chelates depend on pH to some extent, good pH control may be desirable. This is difficult in the small liquid volumes resulting from ash dissolution. Of course, since the ash is hygroscopic, care must be taken if all solvent extraction character is to be removed from the experiment. Procedure . -The metal chelates were formed directly from the dry ash by mixing the ash v;ith a benzene solution containing H(tfa) and 11 (fod). Since the bulk of the ash consisted of NaOH, no dissolution was apparent. The trace metals with

PAGE 120

Ill which the liquid came in contact did form soluble chelates. The trace elements were thus physically separated from the sodium in the ash when the mixture was centrifuged and the solution containing only trace metal chelates and ligand was decanted. At this point, the sodium concentration in the trace metal mixture had been lowered to a level which allowed trace metal determination by NAA without significant sodium interference. Preirradiation separation . -Figure 8 shows the y^ay spectrum of the benzene solution of the chelated trace metals after a 49-hr irradiation. Table 14 lists the radioisotopes found in this and two other spectra of the same mixture recorded after different decay periods. A total of 23 photopeaks representing 11 elements were identified. The later spectra show the complete decay of some iostopes. The decay of ®^Br and ^ "^Na allows identification of ^^"^Sb, '^Se, ^^Fe, and Co photopeak. More complete separation is thus needed where the delay required to detect these radioisotopes is unacceptable. The y^ay spectrum from the solid residue recorded after a three week decay is shown in Figure 9. A total of 35 photopeaks representing 17 elements were identified, Table 15 contains a complete list of the radioisotopes which were detected in this and three other spectra of the same sample recorded after different decay periods. In an attempt to chelate a larger number of metals, the experiment v;as repeated with additional mixing and more time for chelation. Figure 10 shows the Y~i^ay spectrum of the

PAGE 121

c o •rt +> (0 H 0) x: u +j o 0) M •H Q U 0) +J M-l < C O •H +> r-i o CO +> (0 H 0) u iw o »H 4J O 0) Gi CO >i « I (0 0) en •H

PAGE 122

113 Bh £02 3S 9fr tl> lO ^ ro cj 3inNllM d3d siNnoo

PAGE 123

114 Table 14. Spectra from Direct Chelation Organic Layer Isotope

PAGE 124

i •H

PAGE 125

116 "Z. S9 Ofr •J 6S NOIlVniHINNV 08 NOHllSOd og icr "V, '661 o

PAGE 126

117 Table 15. Spectra from Direct Chelation Solid Residue Isotope

PAGE 127

118 Table 15.

PAGE 128

U cn

PAGE 129

120 O o cvi •30 0N^2 qs. in 01 ON fr2 Ofrl 0* U7 3S •J '6S 9J ^^29 — =^C^ ""7 %s_!!tfc= ••a Z8 DT ^2 9 'S, on ow9*o in in CVJ > — • UJ >o o q: o iiJ — z UJ o o in b in CVJ d o in cvi Q CVI in in d 3inNiN d3d siNnoo

PAGE 130

121 benzene solution of chelated trace elements thus obtained. A total of 32 photopeaks representing 13 elements can be identified. Table 16 contains a complete list of the isotopes which were found in this and two other spectra of the same material recorded after longer decay periods. The improved results after additional mixing were partly due to a shorter decay period, 4 days rather than 2 6 days, before counting. The total integrated flux was higher for the second sample. The enhanced sensitivity observed may have resulted to some extent from these factors rather than entirely from improved chelation conditions. Figure 11 shows the y-ray spectrum of the solid residue after additional mixing. The 3 identified photopeaks from 16 radionuclides are listed in Table 17. They include all those identified in the preceding experiment and ^^^Hf. The reagent blank spectrum photopeaks and radionuclides are listed in Table 18. Six elements present in the above spectra were found in the blank. Of these, however, only ^^g^^ ^"Hg, ^^agj^^ la^gj^ ^^^ 46 g^ originated entirely in the reagents used in the blank. The rest were also present in the serum. Postirradiation separation . -An experiment involving a 30-min irradiation followed by direct chelation and a short Y-ray count revealed no trace metals. ^^^I, ^^Cl, and ^ ^Br were the only radionuclides detected, other than small amounts of ^"^Na contamination. One would expect that at least a few trace metals can be identified this way in spite of the short irradiation time. At this point, however, the data indicate

PAGE 131

122 Table 16. Spectra from Longer Direct Chelation Organic Layer Spectra Isotope Y-Ray Energy 3 day 5 day 7 day keV (315) Decay Decay Decay «^Br 1 9 9 Au 8 2, Br 5^Cr 1 ii 0'La 2 4, 'Na 1 9 8 1 4 Au La 'La Br ^Sb Sb Br ^Br 1 k 8 2, La Br 14 La 8 2, Br ^''Na 1 h La h 6 Sc 'La 92

PAGE 132

123 Table 16. continued, Isotope

PAGE 133

c •r^ +j (0 H 0) i3 u •p o 0) M •H Q U 0) Cn O ^:^ u (U -p m < 3 13 •H U) •H H O CO o u +J o 0) CO >i (^ I en •H

PAGE 134

125 O O cvi <5S lO ifZ\ — o-\ Ofrl -s o

PAGE 135

126 Table 17. Spectrum from Longer Direct Chelation Solid Isotope Y-Ray Energy keV (315) Isotope y-Ray Energy keV 'Se 'Se 'Au 'Se 'Se ^Cr Ba 'Se 'Au 'La Ba 8 2 Br ^^Zn ^^Br 1 1 om Ag 121.1 136.0 158.3 264.6 279.6 320.0 373.1 400.7 411.8 486.8 496.3 554.3 604.4 619.0 657.8 «2Br

PAGE 136

127 Table 18. Spectra from Longer Direct Chelation Blank Isotope Y-Ray keV

PAGE 137

128 Table 18. continued... Isotope

PAGE 138

129 that a preirradiation separation with its option of a long irradiation time offers the most promise for multielement determination . Extraction from solution Advantages . -The exposure of the atoms in the aqueous solution to the ligand is more uniform and more thorough, resulting in greater reproducibility and accuracy than can be obtained v/ith the solid ash. Dissolving the ash allows metals which form oxides with differing stabilities to be uniformly available for chelation. Another advantage in dissolving the ash lies in being able to use a carrier solution. The carrier must be in the same chemical state as the sample. The difficulty of homogenous sampling of solid carrier along with the impracticality of duplicating the chemical state of the serum ash make solid carrier mixtures undesirable. Prodedure.-The first extraction experiment used a procedure for determining a limited number of elements using a postirradiation separation with a simple carrier mixture. Spectra from this experiment appear in Figures 5-7. Ash dissolution and chelate extraction were preceded by a direct chelation step. Any metals which are preferentially chelated this way are removed under optimum conditions, v/hile the others are unaffected and remain with the ash to be chelated after dissolution. The results of a procedure for determining a large number of elements using a postirradiation separation with a complex

PAGE 139

130 carrier mixture appear in Table 19. It lists photopeaks present in the y-ray spectra recorded at various stages of each of three experiments using the multielement carrier solution. The items whose activities were measured are listed in the source column. The designation total organic refers to the combined organic extract solutions. Counting took place after chelate extraction but before vaporization. The traps were charcoal filled; their numbers indicate the sequence in which they were attached to the outlet of the vaporization system. The pot contained the nonvolatile residues after vaporization was complete. The time and temperature column entries are the length of time each trap was attached at the specified vaporization temperature. The other two columns contain the major and minor radionuclides observed. The underlined components make up an unusually large proportion of the total activity present. Adjustment of pH in the ash solution after acid dissolution is important for several reasons. As mentioned in the introduction, the pH should be between 5 and 8 for maximum extraction efficiency using the ligands H (tf a) and H(fod). Perhaps more important than the accuracy of the adjustment are the complexing characteristics of the base used for neutralization. In experiments B and C, KOH was used for neutralization. A substantial amount of precipitate was formed in both cases. The precipitated metals are closer to being completely available for chelation than were their oxides in the ash, yet chelate formation still cannot take place

PAGE 140

X31 -p c e •H U Q) X w 0) iH U nJ O Q) •-{ CU e o u c o •H 4-) o +» M +J O
PAGE 141

132 0) a H +> o o w

PAGE 142

133 directly. The same thing is true when NH^OH is used for -neutralization. Here, some of the metals are present as complexes. Assuming the pH is kept low enough to prevent sequestration, these complexes must be capable of rapid dissociation to allow rapid formation of the metal chelates. In such a solution, where none of the molecules can be completely surrounded by less-reactive species and thus kept from reacting, one would expect better results. The presence of Mn, ^^^Ba and Sr in the precipitate in experiment C, even after repeated washing, tends to support this conclusion. B-diketone chelates of all three are known, yet even at a suitable pH they remained as hydroxide precipitates. Carrier Preparation The use of a carrier is sometimes necessary when unfavorable equilibria combine with low concentrations to make efficient recovery of radioisotopes in postirradiation separation procedures difficult. If only a few elements are sought, carrier preparation may consist of simply adding small amounts of salts of the elements of interest to the serum ash just before dissolving it. Results of such an experiment appear in Figures 5-7. Photopeaks from ^"^Cu, ^ ^ Mn , and ^ ^"^Zn appear in Figure 7, the spectrum from the separated chelates. Their presence shows that a carrier solution is neither necessary nor sufficient for the detection of a particular element in every case since no Zn or Mn carrier was present, yet Cr , La, and Sm carrier addition did not result in detection of these elements.

PAGE 143

134 The use of a complex carrier solution in a survey-type NAA experiment requires choosing the carrier elements carefully and systematically. Care must be taken to include every isotope which might be detected under the unique set of irradiation and counting conditions encountered in a particular experiment. Since the iostopes produced by (n, y) reactions have a variety of half-lives as v/ell as different parent cross sections and abundances, it is not always immediately obvious whether or not a particular element will yield significant activity after irradiation under a particular set of conditions, Equation 1 can be used to calcualte this activity. -yt. -yt A = Na(f) (1-e ) (e ) (1) where A activity in disintegrations per second per microgram N = number of target atoms a = thermal neutron cross section (j) = thermal neutron flux Y = decay constant t. = irradiation time t decay time The use of a computer facilitates calculations for many elements under different sets of conditions. A Focal program used with the PDP-8/L computer for these calculations appears in Table 20. This particular program calculates the activity per microgram of parent after a 30 min irradiation at 10 ^^n cm ^ s ^ and a 1 hr decay. The results of the calculations for 105 isotopes appear in Table 21.

PAGE 144

135 Table 20. Activity Calculation Program C-F0CAL8-UF bK3W 01.01 C HhOGnAy TO CALCULATE ACTI^^IT/ IM DI bl VJTF GRAT T J^S 01. 0y C PEh SECOND FEh viChOGKAM AFTEh A 30 MINJUTF 01.03 C IhhADIAlIONJ AI A FLUX OF 10tl2 AMD A 1 HOUh 01.0^ C DECAr. 01. 15 E 01. y0 b L2 = -FLOGCy) 01.25 T " PAhENJT" 1.26 f " PAHENJl •• ! 01»33 T " DPS/UG HALF %ABUM-" 01.31 T •• CROSS ATO"^IC"! 01.^0 1 " ISOTOPE OF PAhEMI LIFE DAMCE" 01.^1 T " SECTIOM WEIGHT"!!! 0i«60 T " ";x =;x :;a w;x e;c rexov/E echo. 01.65 T *'• 01.70 a HL;C half life INJ HOU^.S. 01.75 T %6.02 hl;t •• ";a ab;t %5.0y ab;i • •• 01.80 A xs;c CROSS section inj barms 01.83 T 28.03 XS;T " " 01.85 A WT;T 24.0 1 WT 01.'^0 T *" •• 02.10 S Pl = 6.023E05*XS + < 1-FEXP(L2/(2*HL) )> 02.80 S P2=FEXPCL2/HL)*AB/( 100*WT ) 02. 30 S M=P1*P2; 03.05 S X=0 03.10 I (NJ-l)3.3;i ( 10-M )3. 5.3. 5 03.20 1 %3.02 nj;t "e";t %i x;t ! 03.22 X RiC restore ECHO. "=", AMD ":" 03.25 GOTO 1.6 03.30 s Nj=M + i0;s x=x-i;g 3.1 03.50 s M=>j/i0;s x=x+i;g 3.1

PAGE 145

136 Table 21,

PAGE 146

137 Table 21. continued,

PAGE 147

138 Table 21. continued...

PAGE 148

139 These data were used to choose the elements listed in Table 12, Various salts of these elements were dissolved to make a carrier solution. Concentrations of O.OlMwere chosen so as to minimize the volume of carrier solution added to each ash sample and thus avoid unnecessary ash solution dilution. However, overnight clouding of the carrier solution and the appearance of metallic gold particles from oxidationreduction reactions were noted. Undoubtedly this resulted in some of the elements in the carrier solution being in a nonionic form and therefore not available for chelation. These reactions could have even continued in the serum ash solution after carrier addition. Lower concentrations, along with either use of the carrier solution immediately after preparation or the segregation of the more reactive ions into separate solutions, would probably eliminate this and result in a more homogeneous mixture. Additional Separation Further separation of the fluorine-substituted 6-diketonates by vaporization and gas chromatography is possible because of the chelates' thermal stability and volatility. The first extraction experiment uses the first of two vaporphase separation apparatus designs which were constructed using the oven and carrier gas flow system from a gas chromatograph. The chelates passed easily from the injection port to the charcoal trap, while a significant amount of white crystalline residue remained in the injection port liner and the first inch of the 6-in long open Teflon tube. So even

PAGE 149

140 though the previous solvent extraction step separated the chelate mixture from material which was insoluble in benzene, the useful data in Figure 7 was obtained only after the nonvolatile material had been removed also. The spectra of this white residue left behind show only ^'*Na. The second vapor-phase separation apparatus is a higher capacity version of the first. A liquid reservoir with milliliter capacity acting as a distillation pot in the oven eliminated tedious serial syringe injections into the injection port. Furthermore, uniformly increasing heat applied simultaneously to the entire sample allows chelate vaporization in order of decreasing volatility after solvent removal is complete. This prevents the higher temperatures required for vaporizing certain chelates from causing decomposition of more sensitive, more volatile chelates which can be eluted intact at lower temperatures. Also, solvent condensate in the trap cannot wash condensed chelate out of the trap when solvent removal takes place first. The effectiveness and versatility of the second vaporphase separation apparatus is illustrated by the separation data in Table 19. Experiment A shows that the activated metals, ^ "^Cu in this case, come out in relatively narrow bands rather than being spread out over the entire 27-min heating period. Trap 5, experiment B, shows that the separation is rapid and efficient, as no more radioactive material is eluted after 4 min of high temperature operation. The material of interest in A eluted after ^"'Br and ^'^Cl, but

PAGE 150

141 before ^"^Na. Resolution is also apparent in experiment B, although some ^^Br, ^^Cl, and ^"^Na contaminated the ^ ''Cu spectra. Here, the temperature was raised to the Cu-elution level before all the CI and volatile Br had been eluted. Interference from ^ "^Na is also present in trap 3 because of too rapid heating. The presence of ^ '*Cu and a significant amount of ^^Br in the distillation pot after 8 min at 200 °C in experiment B indicates the presence of both radioisotopes in nonvolatile compounds. Since no attempt was made to influence the Br reactions in the preceding steps it is not surprising that Br compounds with widely varying volatilities were formed. Every attempt was made to chelate quantitatively the Cu, however. The Cu left in the pot is probably the result of decomposition during vaporization. Gentler heating should help reduce this loss. There is another way of obtaining volatile chelate fraction resolution which can be used equally well with either of the vapor-phase separation apparatus designs. If the carrier gas flow, temperature gradient, and diameter of the Pyrex trap tube used for trapping the chelates are properly chosen, the components of the mixture will condense in separate, distinct bands as they encounter successively cooler regions farther from the oven. This procedure does not offer high resolution, but its effectiveness is sufficient for many separation situations.

PAGE 151

142 Interferent Removal Whether the serum spectrum is dominated by the short half-life radioisotopes such as 25.4-min '^^I and 37.29-min ^^Cl, or the longer half-life radioisotopes such as 15-hr ^^Na and 35.82-hr ^^Br, depends on the length of the irradiation and how soon afterward chelation can be performed and they-ray spectrum can be recorded. The preirradiation direct chelation procedure yielded a chelate mixture whose y-^ay spectrum had no ^ "^Na interference after a 3-day decay. The chelate-containing organic layer spectrum summarized in Tables 14 and 16 contain photopeaks from ^^^Au, ^^Br, and La, all of which would have decayed undetected had the sodium not been removed. The spectrum of the chelate extract solution from the postirradiation chelate extraction procedure, Figure 10, contained significant ^ "^Na interference. Vapor-phase separation of this solution reduced the ^ '*Na by a factor of at least 500, leaving ^ ^Cl as the major interferent. At this point, previously obscured ^'*Cu,^^^Zn and ^ ^ Mn photopeaks can be detected. The Na decontamination factors for both these procedures are between 10^ and 10'*. The experimental results from the solution extraction with a complex carrier mixture (Table 19) were not completely without "^a interference even after a final distillation. As was expected, all three organic extracts contained significant 2'*Na contamination, but experiments B and C had additional contamination in the traps after distillation. The

PAGE 152

143 mechanism by which the ^'^Na reached the traps is not known with certainty. Perhaps a volatile sodium-containing compound was formed. What is more likely is that solid particles were blown from the distillation pot into the traps. The use of glass wool plugs in the system may correct this. It must be remembered that the 2'*Na seen in these traps represents an extremely small fraction of the ^'^Na present before separation. Another useful observation is that, rather than flowing continuously into the traps at every temperature, the bulk of the flow took place at higher temperatures. This fact permits further ^'^Na separation from the more volatile chelates where necessary. Once the ^'*Na has been removed, the halogen interference, particularly from the ^^Br with its numerous photopeaks, is the major problem. Serum contains a large amount of CI and lesser amounts of Br and I. During activation these atoms are made highly reactive by Szilard-Chalmers reactions. Their chemical state after irradiation and chelate formation is undoubtedly significantly altered and probably diverse and complex. Collman has reported direct substitution of halogens into 3-diketone chelate rings (316) . Although the number of chelate molecules thus affected is insignificant, the presence of ^^Br in a form with properties similar to the chelates of interest could make interference removal somewhat more complicated. Since the halogens and their compounds are often volatile, the separation techniques described here are not nearly as

PAGE 153

144 effective for them as for sodium. The ^ ^gjinterference problem in the preirradiation direct chelation experiments is shown in Figure 7. Of course it is of nowhere near the same magnitude as the previously illustrated ^'*Na interference. Nevertheless, its removal would facilitate trace element photopeak measurement. The postirradiation extraction experiments, with their short irradiation and decay periods, are particularly vulnerable to interference from 37.39 min '^Cl. Figure 9 shows the prominence of this radioisotope's photopeaks before separation. When Figures 9 and 10 are compared, it can be seen that the separation was much more effective for Na than CI. This is apparent since the relative peak heights are more nearly the same in the later spectrum in spite of the fact that the time which elapsed between them represented several ^^Cl halflives but less than one ^ '*Na half-life. Further decay and separation yield nearly complete ^"^Na removal, but the ^^Cl, even after undergoing a decay of several half-lives, interferes significantly. Further separation of the volatile components is the solution to these halogen interference problems. The three complex carrier experiments show how this separation can be accomplished. For example, experiment A shows that ^®C1 interference with ^"^Cu measurement can be completely eliminated by changing traps after 6 min at 100 °C. Another method is used with the other system. The temperature remains constant; the separated fractions appear as distinct bands in the trap.

PAGE 154

145 Annealing Studies Annealing recoil damage in irradiated metal chelates is necessary when both chelation before irradiation and further chelate separation after irradiation are required. Chelation before irradiation provides a simple means of avoiding the hazardous ^ ''Na radiation levels encountered when irradiating beyond 30 rain. Further chelate separation must be delayed until after irradiation if the separated fractions are to be counted separately. This postirradiation separation allows simultaneous irradiation of the entire sample, reduces irradiation container handling, and makes the use of various gas chromatographic counting geometries possible. The extent of degradation and a means of eliminating it are thus often important considerations in serum analyses. Untreated chelates . -In order to measure the extent and effect of Szilard-Chalmers decomposition, separation was carried out with no attempt to repair the chelate irradiation damage. The result for all three Cr chelates was the same. When the injection ports and traps were visually examined after gas chromatographic separation, a small white residue remained in the injection port, while a large amount of green chelate had condensed in the trap. The bulk of the chelate had passed through the system intact. But when the ^^Cr content of the trap and injection port was measured, incomplete elution was found. Table 22 shows that 85-90% of the ^^Cr activity was found to have never left the injection port.

PAGE 155

146 g 0) -p en >i CO o •H x: a (tj en O -P (d e o v^ u e o s^ en 1 u 0) > o o 0) pi; CO (D •H •o 3 +1 CO tn C •H tH (d (D C O C E-i Eh rH U O (C U Cl4

PAGE 156

147 What happened was that most of the chelate was totally unaffected by the irradiation. It eluted normally. But where ^^Cr neutron capture did occur, Szilard-Chalmers decomposition resulted in at least 85% of the "^Cr atoms remaining in an unchelated form and thus not being eluted. It is clear that some form of effective treatment must be usef if valid data are to be obtained from postirradiation chelate separation techniques applied to irradiated metal chelates. Treated chelates . -Table 22 also contains the results of the experiments involving treatment of the irradiated chelate solution. Only 10% excess ligand and no annealing beyond that which takes place in the reactor increases the eluted ^^Cr activity from 8% to 53%. This figure can be improved even more by annealing. The best and most consistent results come from combining annealing with the addition of 100% excess ligand. When these results are compared with those of an elution study involving chelate synthesis from ^^Cr labeled starting material, the nearly identical results indicate the completeness of radiation damage repair. The measures necessary to prevent loss of chelated radioisotopes in irradiated metal chelate solutions are thus both simple and effective,

PAGE 157

148 Gas Chromatography Those aspects of the gas chromatographic separation of metal chelates which apply to all inorganic gas chromatography are discussed in the Introduction. Two considerations which are not important in ordinary inorganic gas chromatographic analyses but which must be taken into account in radiochemical separations are the radiochemical purity of the eluate and the design of the radiation measurement system. Incomplete Elution Obtaining eluate with the proper radiochemical purity from a gas chromatographic metal chelate separation requires that the activity of each component remain unchanged during elution. The extent to which activated atoms of a given chelate are lost can be determined by measuring the amount of injected radioactive metal chelate which remains in the instrument after the sample has eluted. Residue measurement .-The results of such measurements for several chelates are shown in Table 23. Efficient vaporization with minimal decomposition in the injection port is indicated for all chelates. However, the Teflon tube showed a small amount of chelate retention due to its length and thin liquid phase coating. The packing was the only system component which kept large amounts of chelate from eluting. There is variation in elution behavior among the chelates. The Cr(tfa)3 results were excellent. The higher column temperature used for Cr(fod)3 may have contributed to decomposition,

PAGE 158

149 Table 23. Percent Recovery of Metal Chelates in Chromatographic System

PAGE 159

150 as the high percentage on the column packing indicates. The Be(tfa)2 results were similar to the Cr(tfa)3 results. The Gd(fod)3 data were not as accurate as those from the other experiments because the isotope being measured had gone through three half -lives before counting. This meant that there was a mixture of daughters and other isotopes whose half -lives were unknown, and no correction could be made among the activities measured during each experiment. However, it can be determined that nearly one third of the total activity was found in the column packing. This differs from the Lu(fod)3 results, even though the same column and conditions were used, and both are rare earths. One contributing difference between the chelates lies in their metal ionic , , + 3 13 radii. Gd is 15% larger than Lu ; ions with larger radii form chelates which elute at higher temperatures (172) . Both the Cu(tfa)2 and Fe (fod) 3 left large residues on the packing material. These may be reduced by raising the column temperature, but more extensive changes will probably be necessary before good quantitative results can be obtained with these chelates. To summarize, Cr(tfa)3, Be(tfa)2, and Lu(fod)3 with their respective columns are good choices for radiochromatography. The rest of the chelates will require modification of the chromatographic system described here before they too will be suitable for gas chromatographic separation in NAA. The large residues in some cases would seem to indicate serious deficiencies in the application of gas chromatography

PAGE 160

151 to the quantitative analysis of these metal chelates. This is not necessarily the case. Many workers report the necessity of "conditioning" a freshly prepared column by injecting samples of the particular chelate to be analyzed until reproducible elution is obtained (148, 317). Some laboratories even find this procedure necessary after the system has been idle for only a few hours. This allows quantitative analysis to be performed in spite of the incomplete elution of the first few injections in a long series. If gas chromatography is to be successfully applied to NAA, it may not be sufficient merely to "condition" the column. In a stable isotope experiment the nuclear properties of different samples are the same. If there is exchange between a particular analyte injection and atoms injected prior to analysis for conditioning purposes there is no adverse effect. If, however, there is exchange between two radioisotope injections which have different specific activities, activity measurement of the eluate will yield erroneous results.The exchange experiment with Cu(hfa)2 showed cross contamination to be occurring. The stable sample, injected several hours after the irradiated Cu(hfa)2 sample, eluted carrying with it ^'*Cu from the previous injection. This particular system, then, would not be suitable for carrying out radioactive Cu (hf a) 2 separation. Injection procedures . -Solid injection, liquid injection, and DMCS treatment were evaluated in an effort to promote complete elution. The solid sampling device used in the

PAGE 161

152 Cu(hfa)2 study performed well. A solid sampler allows the injection of a relatively large quantity of chelate without a large amount of solvent which would be detrimental to the separation and interfere with effluent trapping. Also, the sealed quartz ampoule can be taken from the reactor and placed directly in the sampling device without the handling of open containers and evaporation involved in liquid sampling. The measurements of activity from the solid sampler and quartz fragments showed large amounts of metal chelate residue This was partly due to Szilard-Chalmers decomposition. The intact visible chelate residue was present because of low injection port temperatures and incomplete carrier gas purging of the inside of the barrel of the solid sampling device. The length of the quartz ampoules used with the solid sampler was found to be critical. If the ampoule was too short, it was not broken during sampling; instead, it was forced upward, breaking the glass injection port liner. If the ampoule was too long, the large amount of broken quartz fragments prevented all the sample from leaving the sampler as a single narrow band. Lengths between 0.406 in and 0.419 in worked best. Syringe injection of solutions is advantageous when the volume of solvent necessary to dissolve a useful amount of chelate can be tolerated. The extract from a chelation reaction can be introduced directly with no processing or sealing. Furthermore, syringe injection does not entail cleaning quartz fragments from the injection port or

PAGE 162

153 interrupting the carrier flow each time an injection is made. Sometimes excess solvent problems can be eliminated by injecting a relatively large amount of sample into a cold column. After the solvent has eluted, the oven heat is quickly increased, vaporizing the chelate without broadening sufficient to prevent separation. The treatment of the injection port and column with dimethyldichlorosilane (DMCSy showed little effect on the quantity of chelate eluted. Figures 12 and 13 show the result of the DMCS treatment. The chelate peak areas of the 1-yl injections are plotted on the vertical axis; the horizontal axis represents order of injection. Peak area averages for all the injections made at the same stage of DMCS treatment are also plotted. There is no significant increase in slope with treatment. This indicates DMCS treatment is of no value in decreasing incomplete elution in this case. Counting Geometry After the radioactive metal chelate has eluted, it must be held in a reproducible manner in front of a radiation detector long enough to allow good counting statistics and detection of low activity components. Factors to be considered in choosing a counting geometry are reproducibility and ease of restoration before a subsequent sample arrives to be counted, The relative response of the Nal (Tl) crystal as a function of sample position is an important consideration in the design of counting geometries. Consecutive samples need to be held

PAGE 163

t3

PAGE 164

155 — 1

PAGE 165

c

PAGE 166

^^^rn 157 o GO o o CD O UJ O tiJ < O o o o o -J < Io o to o CVJ CM CM GO '(Vdl)J0*V3dV>
PAGE 167

158 in regions with the same response if reproducible results are to be obtained. Figure 14 shows the plot of response versus position obtained by slowly passing a ' ^ ^Cs source across the face of the detector. The region where the response decreased by no more than 10% from the maximum was found to be a 33-mm diameter circle concentric with the cylindrical detector. The active part of each geometry was kept within this region. Stainless steel spiral . -The stainless steel geometry can yield a differential radiochromatogram with characteristics similar to a conventional chromatogram. Figure 15 shows such a chromatogram obtained from ^ '*Cu enriched Cu(hfa)2 which had been separated from an irradiated Cu (hf a) 2-Cr (hfa) 3 mixture. The tailing is pronounced, indicating either adsorption or condensation on the walls of the counting geometry tubing. The only way to obtain good counting data with this geometry is to stop the flow of carrier gas while counting. Otherwise, the residence time of the radionuclide in the geometry is too short. Figure 15 shows that the peak maximum is only one or two channels wide. In this case the maximum counting time would be 2.4 s. Since stopping the flow at the optimum time would require precise timing as well as an extremely stable chormatographic system, this type of geometry is not best for quantitative analysis of lov;-activity radionuclides .

PAGE 168

.

PAGE 169

160 a: o Li. o UJ \liJ O o a: u. u o 2 3SN0dS3d 3AllV13d %

PAGE 170

X5 "

PAGE 171

162 .01/ 3inNIW d3d siNnoo

PAGE 172

163 Charcoal trap . -This was by far the simplest and least susceptible to error of all the geometries tested. Moreover, counting the activity from the traps can be done in a location remote from the gas chromatograph and free from its attendant background. Traps from a multicomponent separation of short half-life radionuclides can be counted simultaneously, thus maintaining maximum sensitivity and minimum decay correction uncertainty. The data for the incomplete elution experiments were taken in this way. There were no problems associated with this method of trapping. Reversible counting geometry . -The most elegant counting geometry constructed was the reversible counting geometry packed with column packing. It combines the advantages of the spiral flowthrough and the charcoal trap, yet has none of their disadvantages. Like the charcoal trap, the reversible geometry holds the radionuclide for an indefinite counting time. Unlike the flowthrough type, however, the timing of the carrier gas flow stop is not critical, since the sample moves along the packed tube much more slowly than through empty tubing. The reversible geometry does not hold the sample permanently. After the count is finished, heat and purge gas are applied to flush the geometry. After it is cool again, it is ready for the next component of the sample to be eluted and trapped. Here is complete flexibility, indefinite counting time, freedom from critical stop flow timing, and, ideally, complete reversibility.

PAGE 173

164 Figure 16 shows a 400 channel radiochromatogram which illustrates the reversibility of the geometry. A solid Cu(hfa)2 sample was injected. The activity suddenly rises as the radioactive plug enters the geometry. Since flow was not stopped, the tail of the chromatographic peak accumulates and the count rate slowly increases. At channel 364, heat was applied to the geometry and the temperature started to rise from 6 °C at about 2 °C per minute. The trapped chelate started to move as the geometry began to heat. First it moved to a more sensitive counting position, as indicated by the steep rise in count rate. Then the rate drops to the background level as the radioactive material is flushed from the geometry. The performance of the improved reversible counting geometry can be evaluated using the two 400-channel radiochromatograms in Figure 17. The left-hand portion follows the activity increase as the radioactive chelate accumulates in the packed region. Some tailing is present, but after 12 min the activity remains constant. Fifteen minutes after injection, the flow was stopped for an extended period to permit y-ray counting under static conditions. The halflife of the radioisotope is the only parameter which limits counting time. The right-tend portion of the curve shows the chelate being purged from the geometry. The two curves have plateaus at the same activity levels, indicating no change in the measured activity while the flow was stopped, and a complete purge with background activity returning to the

PAGE 174

(U

PAGE 175

166 ,01/ 3inNllAl d3d SlNflOO

PAGE 176

0)

PAGE 177

168 o

PAGE 178

169 same level as before injection. The improved circular design with its extended region of uniform detector sensitivity eliminated the sensitivity change which caused the sharp peak at the far right of Figure 16. Lack of uncertainty in sample position sensitivity results in better reproducibility with this design. Similar results were obtained with Cr(tfa)3. Samples with activities of several hundred counts per minute were successfully counted and purged from the system. Dependence of photopeak area on injected sample volume was also established,

PAGE 179

170 Ion Exchange Chelating ion exchange resins such as Chelex-100 are potentially quite useful for sodium removal in serum trace element analysis. The strong bond between the metal and the chelating resin enables extraction from very dilute solutions of high ionic strength such as dissolved serum ash. Rapid, quantitative sodium and potassium removal from a large number of serum constituents by this method should be extremely valuable in NAA. Postirradiation Separation Sodium separation optimization . -The sodium separation is optimized when the minimum amount of rinse water necessary to remove all the interfering ^-^Na has been eluted from the coliomn. At this point, sodium removal is adequate, while trace element loss is minimized. The sodium removal process can be followed in Figure 18. The total activity of each 5-ml portion of eluate decreased rapidly during rinsing; after 25 ml had been eluted a sodium decontamination factor of 10^ had been achieved and very little ^ '*Na was present in the eluate. However, ^ ^Mn and ^^g^ ^^^.^ ^-^.so present. The next three 5-ml eluate • samples each showed ^'^Na and ^^gjactivity at equal levels. Both the 2 "^Na and 823^activities decreased as rinsing continued. In this case, a compromise between total ^ "^Na removal and total trace element retention must be made. In view of the sudden drop in the contamination rate around 25 ml, less rinsing is not advisable. However,

PAGE 180

CO -a 0) (U o o u 04 o o H I >:
PAGE 181

17 2 10 15 20 25 30 MILLILITERS ELUTED 35 40

PAGE 182

173 additional rinsing is beneficial only if the radionuclide of interest v/ill not be removed along with the ^"^Na. Serum analysis . -The absence of a blank in postirradiation separations is especially appealing in ion exchange. Another advantage in irradiating before an ion exchange separation such as this is that no final elution step is necessary. The activity measurement can be made while the radioactive trace metals are still on the resin, thus elimiating possible loss during elution. The 1-hr Y-r^ay spectrum recorded 2 . 5 hr after a 30-min irradiation shov/ed moderate amounts of ^'*Na, ®^Br, and ^ ^Cl activity, along with small amounts of ^^®I, ^ ** °La, ^^Ito, and Cu activity in the resin. The measurement was made while the radioactive metals chelated by the resin were still on the column. No elution had taken place, except for rinsing the sodium from the resin. Preirradiation Separation Short irradiation . -The 3-hr irradiation time used in this experiment gives the sample a much greater neutron exposure than the 30-min irradiations used with postirradiation separations, yet it is not so long that turnaround time for the determination is inordinately long or tliat there is unnecessary delay in the determination of nuclides which reach saturation quickly. The purpose of the experiment was to evaluate this intermediate time period as well as find out more about the applicability of this type of separation to trace metal analysis in serum.

PAGE 183

174 NAA of the resins after complete elution showed that the same elements were present in the resin blank and the resin used in the separation experiment. Apparently none was lost in the separation due to permanent retention on the resin, The blank containing the reagents used in the separation did contain a significant amount of potassium. There was enough to obscure small photopeaks which might otherwise have been detected. The potassium apparently came from the KOH used to adjust the pH of the ash solution, even though all potassium should have been removed during the sodium separation. Ammonium hydroxide could have been substituted for KOH, but some ammine complex sequestration of the trace metals might have occurred. The best results are obtained using KOH along with thorough removal of alkali metal ions. The radionuclides detected in the irradiated effluent containing the trace metals were ^"^Na, "^^K, ^^^iZn, ^ ^Cl and ^Mn. More would undoubtedly have been found if a higher flux had been used. The 3-hr irradiation showed little improvement over the 30-min irradiation as far as the number of elements detected. Of course, the activities of the radionuclides detected were greater, and therefore more accurate quantitative data could be determined. Long irradiation . -When used with a Chelex-100 separation, a long irradiation yields a much greater number of detectable radionuclides than do the shorter irradiation times discussed in previous sections. For survey work, or for determining the concentrations of trace elements with long half-lives, a long irradiation is necessary.

PAGE 184

175 Figure 19 shows the yi^^y spectrum of one of the three serum samples which underwent ashing, Na separation using Chelex-100, and irradiation for 52 hr. This 15-hr count followed a 7-d decay. The radionuclides identified are listed in Table 24. The solution blank contained a small amount of ^ ^Cr and ^ °Co. ^ ^ Zn and ^ ^ ®Au were present in significant amounts in the blank, although none of the metals found in the serum originated entirely in the blank. Of all the various procedures used here for the ion exchange separation of sodium from the trace elements in seriom, the preirradiation separation with a long irradiation yielded the best results in terms of the number of elements detected. A postirradiation separation following a neutron exposure time of several hours may also be productive , but remote handling of the sample will be required.

PAGE 185

C

PAGE 186

177 ro (vj 3inNII^ d3d siNnoo

PAGE 187

178 Table 24. Long Irradiation Chelex-100 Separation Isotope Y~R^y Energy, keV Isotope (315) Y-Ray Energy, keV ^Hf 'OS 1 1 7m ^^Br 2 3 Sn Hg 5 1, ^Cr 1 < La 5 9, 'Fe 1 't La ^^Br 8 2i Br 1 1 om Ag 133.1 139.0 158.4 222 279.1 320.0 328.6 335 411.8 482.2 486.8 527.7 554.3 564.3 619.0 657.8 Br Br Br Sc Br 'Fe Zn Co Fe 'Ca Br 'Co *Na ^Br ^La 698.3 776.6 827.8 889.4 1043.9 1098.6 1115.4 1173.1 1291.5 1296.9 1317.2 1332.4 1368.4 1474.7 1595.4

PAGE 188

179 Conclusions Sequence of Operations Figure 20 shows the procedural options v;hich are available when the separation techniques evaluated here are used in NAA of trace elements in serum. The two fundamental choices are which sodium separation technique to use and when to irradiate. Preirradiation separation .-The primary advantage in sodium removal before irradiation is that after leaving the reactor the sample needs no further processing before counting. This avoids decay of short half-life radioisotopes during separation and eliminates personnel exposure to the high radiation levels present after sodium irradiation. The latter is quite important, since irradiation of untreated serxom is limited to a maximum of 30 min unless special shielding and sample manipulating equipment are used. Postirradiation separation . -The chief advantages in delaying separation until after irradiation are the option of carrier addition and the elimination of the reagent blank. Thus, in situations v-here both the irradiation and separation can be carried out quickly, a postirradiation separation is more effective. When choosing the best sequence of operations for determining the serum concentrations of a particular element, its nuclear properties, particularly half-life, are considered. Each element which can be determined using NAA has isotopes

PAGE 189

180 SHORT IRRADIATION ION EXCHANGE LONG OR SHORT IRRAD lATION SERUFl SHORT IRRAD ATION CHELATION EXTRACTION LONG OR SHORT IRRAD GC ATION LONG IRRAD ATION DECAY SHORT IRRADIATION SUBLIMATION COUNTING Figure 20. Procedural Options

PAGE 190

181 which can be placed into at least one of the four categories listed in Table 25. However, a wide range in half-lives may prevent simultaneous analysis of certain combinations of elements. Separation Techniques Once the length of the irrradiation and its location in the analysis scheme have been established, the separation technique options can be evaluated. Chelation . -Both the chelation from solid ash and the chelate extraction from solution worked well. Each is effective in removing ^ '*Na contamination and permitting further separation if needed. Although the better pH control of the ash solution allows selective chelation of some metals and thereby gives greater selectivity, direct chelation from the solid ash is simpler and much faster. The failure of the chelate extraction from solution procedure to yield a large number of elements when used after irradiation is due to the delay in counting caused by the additional sample manipulation. Such procedures must be restricted to preirradiation separations, Chelate separation. -Both gas chromatography and the vapor-phase separation techniques were effective. GC has much greater resolution, but the increased throughput and lack of column retention make the simpler method valuable too. The trapping systems, which can be used with either method, were satisfactory. A charcoal trap or an open Pyrex tube at room temperature v7ork best where the sample is not complex and the analysis time is not critical. There is

PAGE 191

182

PAGE 192

183 more potential versatility with the reversible geometry. It makes yi^ay counting for a series of eluted components from a complex sample quite simple. The information about chelate retention by the gas chromatographic system is applicable to trace element analysis as well as to radiochemical separations in NAA. Some chelates are eluted almost completely and some are not. When a chelate injection is not entirely eluted from a fresh column, exchange with the species needed to condition the column for complete elution can preclude the system's use in NAA. Before such a separation is attempted, this aspect must be considered. Ion exchange . -The Chelex-100 separation yielded excellent ^ '*Na decontamination rapidly with a minimum amount of sample manipulation. It was the only method evaluated here which was suitable for postirradiation separation. The trace metal resolution after sodium removal is poor, especially when compared to chelate methods involving further separation using GC. However, when highresolution counting equipment is available, chelating ion exchange separations can yield resolution sufficient for a wide range of determinations. Future V?ork Automating the experiment will increase the accuracy, the speed, and the number of elements analyzed. Remote sample handling will allov/ postirradiation separation to be used most effectively. On-line computer control of the experiment will allow such improvements as eluent volume optimization

PAGE 193

184 by continuous effluent monitoring in ion exchange separations. The reversible counting geometry and GC stop-flow system were designed for such automated operation. Szilard-Chalmers reactions in the naturally occurring complexes in serum can replace ashing. Ligands forming highly stable complexes can be added to serum before irradiation; chelates with very high specific activities can then be extracted after irradiation. Metal complex separation using liquid chromatography permits the use of thermally unstable complexes. The wider range of ligands available can make multielement quantitative separation of a greater number of elements possible.

PAGE 194

REFERENCES 1. W.G. Hoekstra, Ann. N.Y. Acad. Sci . , 199 , 182 (1972). 2. J.T. McCall, N.P. Goldstein, and L.H. Smith, Fed. Proc. , 20, 1011 (1971) . 3. M.A. Schroeder, J. Chron. Pis. , 18 , 217 (1965). 4. W. Mertz, Ann. N.Y. Acad. Sci. , 199 , 191 (1972). 5. K.G. Kjellin, Int. J. Appl. Radiat . Isotopes, 15, 461 (1964). 6. W.F. Beyer, Clin. Chim. Acta , 38 , 119 (1972). 7. D. Kolihova and V. Sychra, Anal. Chim. Acta, 36 , 479 (1973) 8. I.L. Liplavk and L.N. Ushkova , Lab. Delo , 7, 399 (1969); Chem. Abstr ., 71 , 78049h (1969)T ~ 9. G.S. Fell, H. Smith, and R.A. Howie, J. Clin. Pathol ., 21, 8 (1968). 10. G.C. Battistone, E. Levri, and R. Lofberg, Clin. Chj m. Acta, 20, 429 (1970). 11. J.D. Khandekar, D. Mukerji, and G.C. Sepahn, Indian J . Med. Sci . , 2e_, 813 (1972). 12. F.B. DeJorge, J. Sampaio Goes, Jr., J.L. Guedes, and and A.B. DeUlhoa Cintra, Clin. Chim. Acta, 12, 403 (1965). 13. T. Anovski, T. Memedovic, and M. Rastvovac, J. Radioanal . Chem . , 12_, 483 (1972); Chem. Abstr . , 78, 107858y (1973). 14. E.L. Kanabrocki, T. Fields, C.F. Decker, L.F. Case, E. B. Miller, E. Kaplan, and Y.T. Oester, Int. J. Appl . Radiat. Isotopes , 15 , 175 (1964). 15. E.L. Kanabrocki, L.F. Case, L. Graham, T. Fields, E.B. Miller. Y.T. Oester, and E. Kaplan, J. Nucl. Med . , 8, 166 (1967). 16. E.L. Kanabrocki, L.F. Case, T. Fields, L. Graham, E.B. Miller, Y.T. Oester, and E. Kaplan, in "Proceedings of the 1965 International Conference on riodern Trends in Activation Analysis," College Station, Texas, 1965, p. 220. 185

PAGE 195

186 17. A. Jacobson, S. Brar, T. Fields, I.G. Fels, E. Kaplan, P. Gustafson, and Y.T. Oester, J. Nucl. Med . , 2, 289 (1961) 18. M.C. Haven, G.T. Haven, and A.L. Dunn, Anal. Chem . , 38 , 141 (1966). 19. D.A. Olehy, R.A. Schmitt, and VJ.F. Bethard, J. Nucl . Med . , 7, 917 (1966) . 20. W.F. Bethard, D.A. Olehy, and R.A. Schmitt, in "L'Analyse par Radioactivation et ses Applications aux Sciences Biologiques ," Presses Universitaires de France, 1964, p 379; Chem. Abstr . , 6J^, 3305g (1965). 21. W.F. Bethard, R.A. Schmitt, and D.A. Olehy, Strahlen therapie Sonderb . , 5, 222 (1962); Chem. Abstr . , 61 , 9766c (1964). 22. N.A. Dubinskay, H.H. Mihelsons, and L. Pelekis, Neitrono aktiv. Anal., Akad. Nauk Latv. SSR , Inst. Fiz., 109, 1966. 23. D. Comar, G. Girault, and C. Kellershohn, Bull. Soc . Chim. Fr ., 1, 1766 (1962). 24. THAI Office of Atomic Energy for Peace, Bangkok, Report THAI-AEC-22 , 1969. 25. T. Asai, E. Meren, M. Okumoto , Y. Sato, and T. Azuma, Annu. Rep. Radiat. Center Osaka Prefect . , 1_, 99 (1966) ; Chem. Abstr ., 6_7, 105916q (1967). 26. D. Brune, B. Frykberg, K. Samsahl , and P.O. Wester, "Determination of Elements in Normal and Leukemic Human V7h61e Blood by Neutron Activation Analysis," AE-60, Aktiebolaget Atomenergi , Stockholm, Sweden, 1961; Chem . Abstr . _59, 12013h (1963) . 27. W.A. Wolstenholrae, Nature , 203 , 1284 (1964). 28. C.C. Thomas, Jr., G.P. Tercho, and J. A. Sondel, Nucl . Appl . , 3, 53 (1967); Chem . Abstr . , 66, 92354g (1967). 29. C.C. Thomas, Jr., G.P. Tercho, and J. A. Sondel, Trans . Amer. Nucl. Soc , 9, 69 (1966). 30. M.M. Molokhia and B. Portnoy, Br. J. Dermatol . , 83, 376 (1970) . 31. L.O. Plantin and P.O. Strandberg, Acta Rheumatol. Scand . , 11, 30 (1965); Chem. Abstr . £3, 15362e (1965). 32. S. Meret and R.I. Henken, Clin. Chem., 17, 369 (1971).

PAGE 196

187 33. C. Pfeiffer, Rev. Can. Biol . , 31, Suppl . 73 (1972). 34. S. Sprague and W. Slavin, At. Absorption Newslett . , 4, 228 (1965). ~~ ~ 35. J. P. Matousek and B.J. Stevens, Clin . Chem. , 17, 363 (1971). — 36. D.M. Taylor, Biochem. J . , 91 , 424 (1964). 37. S. Giovannetti, Q. Maggiore , and R. Malvano, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p. 511. 38. D. Brune, K. Samsahl, and P.O. Wester, "The Amounts of Arsenic, Gold, Bromine, Copper, Iron, Molybdenum, Selenium, and Zinc in Normal and Uremic Human Whole Blood. A Comparison by Means of Neutron Activation Analysis," AE-134, Aktiebolaget Atomenergi, Stockholm, Sweden, 1964; Chem. Abstr . 60 , 14956c (1964). 39. D. Brune, K. Samsahl, and P.O. V'/ester, Clin. Chem . 13, 285 (1966). 40. W. Niedermeier and J.H. Griggs, J. Ch ron. Dis., 23, 527 (1971) . — 41. R. Malvano, U. Rosa, and P. Grosso, Int. J. Appl . Radiat . Isotopes , 18, 121 (1967). 42. K. Kasperek, H. Schicha, V. Siller, and L.E. Feinendegen, Strahlentherapie , 143 , 468 (1972). 43. J. Pijck, J. Gillis, and J. Hoste, Int. J. Appl. Radiat . Isotopes , 1_0, 149 (1961). 44. J. Pijck and J. Hoste, "L' Analyse par Radioactivation et ses Applications aux Sciences Biologiques," Presses Universitaires de France, Paris, 1964, p. 361. 45. L.M. Paixao and J.H. Yoe , Clin. Chim. Acta , _4, 507 (1959), 46. R. Monacelli, H. Tanaka, and J.H. Yoe, Clin. Chim. Acta . , 1, 577 (1956). 47. D. Comar, C. LePoec, M. Joly, and C. Kellershohn, Bull. Soc. Chim. Fr. , 1, 56 (1962). 48. R.P. Spencer, T.G. Mitchell, and E.R. Xing, Amer. J . Roentgenol . , 7_9, 1053 (1958). 49. S.M. Prigozhina, A. A. Kist, and E.M. Lobanov, "Aktiv. Anal. Biol. Ob'ektov 1967 ," p 92; Chem. Abstr., 70, 45594h (1969) . —

PAGE 197

188 50. S,K. Teshabaev and E.N. Mavaeva, Tr. Samarkand. Gos. Univ. , 193, 142 (1972), Chem. Abstr. 78, 107847v (1973). 51. K. Kasperek, H. Schica, A. Hoeck, V. Siller, and L.E. Feinendegen, Strahlentherapie , 145 , 229 (1973). 52. R.E. Fredericks, K.R. Tanaka, and W.N. Valentine, J. Clin. Invest ., 43 , 304 (1964). 53. T.F. Budinger, J.R. Farwell, A.R. Smith, and H. Bichsel, Int. J. Appl. Radiat. Isotopes , 2_3, 49 (1972). 54. H.J.M. Bowen and P. A. Cawse , "The Determination of Inorganic Elements in Biological Tissue by Activation Anaysis," AERE-R-4 309, Atomic Energy Research Establishment, Wantage, England, 1963. 55. R.A. Nadkarni, D.E. Fliedner, and VJ.D. Ehmann, Radiochim . Acta , 11, 97 (1969). 56. D. Brune, K. Samsahl, and P.O. Wester, Atompraxis , 9, 368 (1963); Chem. Abstr . £0, 7134c (1964). 57. D. Behne, J. Radioanal. Chem . , 3, 17 (1969); Chem. Abstr . 71, 78061f (1969). 58. T. Asai, Y. Iwai, R. Miki , Y. Kondo , Y. Sato, and T. Azuma, Ann. Rep. Radiat. Center Osaka Prefect . , 8_, 120 (1967); Chem. Abstr . 69 , 84012r (1968). 59. W.A. Haller, R. Filby, and L.A. Rancitelli, Nucl . Appl . , 6, 365 (1969); Chem. Abstr . 71^, 10161h (1969). 60. E.L. Kanabrodci, J. Nucl. Med . , 6, 780 (1965). 61. G.C. Cotzias, S.T. Miller, and J. Edwards, J. Lab . Clin. Med ., 67, 836 (1966). 62. H.J.M. Bowen, J. Nucl. Energy , 3, 18 (1956). 63. M. Simkova and M. Krivanek, J. Radioanal. Chem . , 2_, 229 (1969); Chem. Abstr . 7_0, 112270q (1969). 64. S. Meloni, A. Brandone, and V. Maxia, Int. J. Appl. Radiat . Isotopes , 20 , 757 (1969). 65. H.J.M. Bowen, Analyst (London), 89_, 658 (1964). 66. P.F. Johnson, P. Tothill and G.W.K. Donaldson, Intern . J. Appl. Radiat. Isotopes , 20 , 103 (1969). 67. J. Savory, P. Mushak , F.W. Sunderman , Jr., R. Estes and N.O. Roszel, Anal. Chem., 42, 294 (1970).

PAGE 198

189 68. J. Savory, P. Mushak and F.W. Sunderman, Jr., J. C hrom. Sci. , 7, 674 (1969). 69. L.C. Hansen, W.G. Scribner, T.W. Gilbert, and R.E. Sievers, Anal. Chem . , 43 , 349 (1971). 70. W.R. Wolf, M.L. Taylor, 3.M. Hughes, T.O. Tierman, and R.E. Sievers, Anal. Chem . , 44 , 616 (1972). 71. F.J. Feldman, G.D. Christian, and W.C. Purdy , Amer. J. Clin. Pathol . , 49, 826 (1968). 72. S. Hamada, K. Torizuka, K. Hamamoto, T. Mori, R. Morita, T. Nakagawa, and J. Konishi , Radioisotopes , 18, 354 (1969); Chem. Abstr ., 7_2, 63120w (1970). — 73. R. Grunewald, P.L. Ziemer, and J.E. Christian, Anal . Chem . , 36.' 1138 (1964). 74. A. Dimitriadou, P.C.R. Turner, and T.R. Eraser, Nature, 198 , 576 (1963). 75. A. Dimitriadou, P.C.R. Turner, & T.R. Fraser, Nature, 197, 446 (1963). 76. R.H. Tomlinson and M.W. Billinghurst , Trans. Amer. Nucl . Soc. , 12, 458 (1969). 77. J.F. Balcius, D.M. Linekin and G.L. Brov;nell, Trans . Amer. Nucl. Soc . , 12, 458 (1969). 78. R. McG. Harden and C.H. Bastomsky, Clin. Chem . , 17, 1020 (1971). 79. H.J.M. Bowen, Biochem. J . , 73 , 381 (1959). 80. Y.K. Plotnikov, Terapevt. Arkh . , 35, 79 (1963); Chem. Abstr . , 59_, 12013b (1963). 81. J.L.A. Hunteler, W. van der Silk, and J. P. Persijin, Clin. Chim. Acta , 36 , 485 (1972). 82. H.Y. Yee and A. Zin, Clin. Chem. , 17, 950 (1971). 83. L.O. Plantin, "L'Analyse par Radioactivation et ses Applications aux Sciences Biologiques , " Presses Universitaries de France, Paris, 1964, p 211. 84. H.R. Lukens, K. Heydorn, and T. Choy, Trans. Amer. Nucl . Soc. , 8^, 331 (1965). 85. T.E.F. Carr and A. Sutton, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p 445.

PAGE 199

190 86. C. Jansen, G.W. Leddicotte, and M. Navarrette, Radio logy, 91, 813 (1968). 87. T.M. Terree and S.H. Cohn , J. Nucl. Med . , 1_, 848 (1966). 88. B.S. Carpenter and C.H. Cheek, Anal. Chem . , 42 , 121 (1970). 89. B.S. Carpenter, in "Proceedings of the 1968 International Conference on Modern Trends in Activation Analysis," National Bureau of Standards, Gaithersburg Md. , 1968, p 942. 90. M. Okuda and H. Sasmoto, Rinsho Byori , 17, 380 (1969); Chem. Abstr . , 71, 78954f (1969). 91. H.D. Schwartz, B.C. McConville, and E.F. Christopherson, Clin. Chim. Acta , 31, 97 (1971) . 92. S.O. Hansen and L. Theodorsen, Clin. Chim. Acta, 31, 119 (1971). — 93. G.K. Davis, Ann. N.Y. Acad. Sci . , 199 , 92 (1972). 94. J. Pybus, F.J. Feldman, and G.N. Bowers, Jr., Clin. Chem., 16, 998 (1970). 95. P. A. Cawse and M. Daglish, Analy st (London), 89, 266 (1964). — 96. C.K. Kim and W.W. Meinke, in "Proceedings of the 1965 International Conference on Modern Trends in Activation Analysis," College Station Texas, 1965, p 316. 97. D. Brune, Acta Chem. Scand . , 20 , 1200 (1966). 98. C. Kellershohn, D. Comar, and C. LePoec , J. Lab. Clin . Med., 6_6, 168 (1965). 99. B. Sjostrand, Anal. Chem ., 36, 814 (1964). 100. J.Y. Kwang, P. A. Ullucci, and C.J. Mokeler, Anal. Chem. , 45, 795 (1973). 101. G.A. Rose and E.G. Willden, Analyst (London), 98, 243 (1973) 102. M.M. Joselow and J.D. Bogden, At. Absorption Newslet t. , IJ., 127 (1972). 103. H.J.M. Bowen and P. A. Cawse, Analyst (London), 88. 721 (1963). . — ' 104. R.C. Dickson and R.H. Tomlinson, Int. J. Appl. Radiat . Isotopes, 18, 153 (1967).

PAGE 200

191 105. R.H. Tomlinson and R.C. Dickson, in "Proceedings of the 1965 International Conference on Modern Trends in Activation Analysis," College Station, Texas, 1965, p 66. 106. N.V. Bagdavadze, L.V. Barbakadze, E.N. Ginturi , N.E. Kuchava, L.M. Mosulishvili and N.E. Kharabadze, Soobshch. Akad. Nauk Gruz . SSR , 39, 287 (1965); Chem. Abstr, 6£, 5439b (1966). 107. E. Sh. Botzvadze, L.M. Mosulishvili, N.E. Kuchava, and E.N. Ghinturi , Phys. Med. Biol ., 14 , 19 (1969). 108. J.V. Dunckley, Clin. Chem ., 17, 992 (1971). 109. A. Abu-Samra and G.W. Leddicotte, in "Proceedings of the 1968 International Conference on Modern Trends in Activation Analysis," National Bureau of Standards, Gaithersburg, Md., 1968, p 134. 110. J. Pybus and G.N. Bowers, Jr., Clin. Chem . , 16 , 139 (1970). 111. H.J.M. Bov/en, Biochem. J. , 77, 79 (1960). 112. M.L. Taylor and E.L. Arnold, Anal. Chem. , 43 , 1328 (1971) 113. M.L. Taylor, E.L. Arnold, and R.E. Sievers, Anal. Lett . , 1, 735 (1968). 114. J.F. Goodwin, Clin. Chem . , 11_, 544 (1971). 115. P. Venkateswarlu, L. Singer, and W.D. Armstrong, Anal. Biochem ., £2, 350 (1971). 116. G.W. Leddicotte in "Methods of Biochemical Analysis," D. Click, Ed., Interscience, New York, N.Y., 1971. 117. H.A. Schroeder and A. P. Nason, Clin. Chem. , 17, 461 (1971) . 118. H.J.M. Bowen , "Elementary Composition of Mammalian Blood," AERE-R-4196, Atomic Energy Research Establishment, Wantage, England, 1963; Chem. Abstr . , 59 , 6784c (1963) . 119. R.A. Levine, D.H.P. Streeten, and R.J. Doisy, Me tab . , Clin. Exp. , 17 , 114 (1968). 120. W.B. Herring, B.S. Leavell, L.M. Paixao, and J.H. Yoe , Amer. J. Clin. Nutr . , 8_, 846 (1960) . 121. P.O. Wester, "Trace Elements in Human Myocardial Infarction Determined by Neutron Activation Analysis," AE-188, Ziktiebolaget Atomenergi, Stockholm, Sweden, 1965.

PAGE 201

192 122. P. Correa and J. P. Strong, Ann. N.Y. Acad. Sci ., 199 , 217 (1972). 123. P.O. Wester, Acta Med. Scand . , 178 , 765 (1965). 124. H.A. Schroeder, J. Amer. Med. Ass . , 172 , 1902 (1960). 125. G. Biorck, H. Bostrom, and A. Widstrom, Acta Med. Scand , 178 , 239 (1965). 126. H.M. Perry, Jr., Ann. N.Y. Acad. Sci . , 199 , 202 (1972). 127. G.L. Curran, J. Biol. Chem. , 210, 765 (1954). 128. J.W. Berg and F. Burbank, Ann. N.Y. Acad. Sci ., 199 , 249 (1972). 129. W.J. Pories, E.G. Mansour , and W.H. Strain, Ann. N.Y. Acad. Sci., 199, 265 (1972). 130. L. de Galan, "Analytical Spectroscopy," Adam Hilger, Ltd., London, 1971, p 127. 131. H.L. Kahn in "Trace Inorganics in Water," Advances in Chemistry Series, No. 73, R.F. Gould, Ed., American Chemical Society, Washington, D.C., 1968, p 183. 132. D.C. Manning and F. Fernadez, At. Absorption Newslett. , 9_, 65 (1970) . 133. J.D. Winefordner and R.C. Elser, Anal. Chem . , 43 (4), 24A (1971) . 134. B.V. L'vov, Spectrochim. Acta , 24B , 53 (1969). 135. V.A. Fassel and D.W. Golightly, Anal. Chem . , 39 , 466 (1967). 136. E.E. Pickett and S.R. Koirtyohann, Spectrochim. Acta , 2 3B , 235 (1968). 137. E.E. Pickett and S.R. Koirtyohann, Spectrochim. Acta , 24B, 325 (1969). 138. S. Greenfield and P.B. Smith, Anal. Chim. Acta , 59 , 341 (1972) . 139. J.D. Winefordner, V. Svoboda, and L.J. Cline, Crit . Rev. Anal. Chem . , 1, 233 (1970). 140. A. J. Mitteldorf, in "Trace Analysis, Physical Methods," G.H. Morrison, Ed., Interscience , Hew York, N.Y., 1965, p 193.

PAGE 202

193 141. J.D. Winefordner, S.G. Schulman, and T.C. O'Haver, "Luminescence Spectrometry in Analytical Chemistry," Interscience, New York, N.Y. , 1972. 142. F.S. Goulding and J.M. Jaklevic, "Trace Element Analysis by X-Ray Fluorescence," URCL-20625, Lawrence Radiation Laboratory, 1971. 143. R.D. Giauque, F.S. Goulding, J.M. Jaklevic, and R.H. Pehl, Anal. Chem . , 45, 671 (1973). 144. F.C. Young, M.L. Roush , and P.G. Herman, Int . J. Appl. Radiat. Isotopes , 24 , 153 (1973). 145. I. Adler and H.J. Rose, Jr., in "Trace Analysis, Physical Methods," G.H. Morrison, Ed., Interscience, New York, N.Y. , 1965, p 271. 146. L.S. Birks, Anal. Chem ., £4(5), 557R (1972). 147. R.H. Hammerle, R.H. Marsh, K. Rengan, R.D. Giauque, and J.M. Jaklevic, Anal. Chem . , 45 , 1939 (1973). 148. D.E. Leyden, R.E. Channell and C.W. Blount, Anal. Chem . , 4_4, 607 (1972). 149. D.E. Porter and R. Woldseth, Anal. Chem . , 45 , 604, (1973), 150. J. Reboz, in "Trace Analysis, Physical Methods," G.H. Morrison, Ed., Interscience, New York, N.Y., 1965, p 435. 151. L. de Galan, "Analytical Spectrometry," Adam Hilger, Ltd., London, 1971, p 241. 152. N.M. Frew, J.J. Leary, and T.L. Isenhour, Anal. Chem . , 4_4, 665 (1972). 153. P.J. Paulsen, R. Alvarez, and D.E. Kellerher, Spectrochim. Acta , 24B , 535 (1969) . 154. L.J. Moore and L.A. Machlan, Anal. Chem. , 44 , 2291 (1972), 155. N. Gochman and D.S. Young, Anal. Chem. , 45 , llR (1973). 156. H.B. Mark, Jr., J. Pharm. Belg ., 25 , 367 (1967). 157. J.D. Czaban and G.A. Rechnitz, Anal. Chem . , 45 , 471 (1973) . 158. G.J. Patriarche, Anal. Lett . , 5, 45 (1972). 159. F.M. VJachi, Diss. Abstr . , 2_0, 53 (1959). 160. H. Freiser, Anal. Chem,, 31, 1440 (1959).

PAGE 203

194 161. A. A. Duswalt, Jr., Diss. Abstr . , 20 , 52 (1959). 162. N.D. Sokolov, M.A. Baydarovtseva, and N.A. Vakin, Izv. Akad. Nauk SSSR , Ser. Khim . , No. 6, 1396 (1968). 163. R.W. Moshier and R.E. Sievers, "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y. , 1965, p 42. 164. S.T. Sie, J. P. A. Bleumer, and G.W.A. Rijnders, Separ. Sci ., 1, 41 (1966). 165. S.T. Sie, J. P. A. Bleumer, and G.W.A. Rijnders, Separ. Sci . , 2, 645 (1967). 166. S.T. Sie, J. P. A. Bleumer, and G.W.A. Rijnders, Separ. Sci. , 3, 165 (1968). 167. R.W. Moshier and R.E. Sievers, "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y., 1965, p 2. 168. L.M. Brown and K.S. Mazdiyasni, Anal. Chem . , 41, 1243(1969). 169. R.W. Moshier and R.E. Sievers, "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y., 1965, p 40. 170. W.C. Butts, Diss. Abstr . , 29 , 506B (1968). 171. T. Shigematou, M. Matsui , and K. Utsunomiya, Bull. Chem. Soc. Jap ., £1, 763 (1968). 172. R.E. Sievers, K.J. Eisentraut, and C.S. Springer, in "Lanthanide/Actinide Chemistry," Advances in Chemistry, Series, No. 71, R.F. Gould, Ed., American Chemical Society, Washington, D.C., 1967, p 141. 173. M. Tanaka, T. Shono, and K. Shinra, Anal. Chim. Acta, 43, 157 (1968) . 174. C.S. Springer, Jr., D.W. Meek, and R.E. Sievers, Inorg . Chem . , 6, 1105 (1967). 175. W.G. Scribner and R.E. Sievers, Personal Communication. 176. C. Genty, C. Houin, and R. Schott, 7th International Symposium on Gas Chromatography and its Exploitation, Copenhagen, Denmark, June 1968. 177. W.G. Scribner, W.J. Treat, J.D. Weis, and R.W, Moshier, Anal. Chem. , 37, 1136 (1965).

PAGE 204

195 178. W.G. Scribner, M.J. Borchers, and W.J. Treat, Anal. Chem. , 38, 1179 (1966). 179. W.D. Ross and R.E. Sievers , Anal. Chem. , £1, 1109 (1969). 180. R.E. Sievers, G. Wheeler, Jr., and W.D. Ross, Proceedings, 3rd International Symposium on Advances in Gas Chromatography, Houston, Texas, 1965. 181. W.D. Ross and R.E. Sievers, Talanta , 15, 87 (1968). 182. R.W. Moshier and J.E. Schwarberg, Talanta, 13_, 445 (1966). 183. G.P. iMorie and T.R. Sweet, Anal. Chim. Acta, 34, 314 (1966). — 184 186 187 188, 189, 190, 192 196 G.P. Morie and T.R. Sweet, Anal. Chem . , r?/ 1553 (1965). 185. J.E. Schwarberg, R.W. Moshier, and J.H. Walsh, Talanta, 11, 1213 (1964). R.D. Hill and H. Cesser, J. Gas Chromatogr . , 1, 11 (1963) R.E. Sievers, B.W. Ponder, M.L. Morris, and R.W. Moshier, Inorg. Chem ., 2, 693 (1963). W.D. Ross and G. Wheeler, Jr., Anal. Chem., 36, 266 (1964). — W.D. Ross, Anal. Chem ., 35, 1596 (1963). R.W. Moshier and R.E. Sievers, "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y.. 1965. p 41. 191. R.W. Moshier and R.E. Sievers, "Gas Chromatographv of Metal Chelates," Pergamon Press, New York, N.Y., 1965, P 53. R.E. Sievers, J.W. Connolly, and W.D. Ross, J. Gas Chromatogr . , 5, 241 (1967). 193. W.D. Ross, W.G. Scribner, and R.E. Sievers, Personal Communication . 194. D.K. Albert, Anal. Chem . , 36 , 2034 (1964). 195. R.A. Nadkarni and G.H. Morrison, Anal. Chem., 45, 1957 (1973). — T. Asai, Y. Iwai, R. Miki , Y. Kondo , Y. Sato and T. Azuma, Ann. Re p. Radiat . C enter Os aka Prefect ., 6, 111 (1965) ; ~CPiem. Abstr . 6 6" , 62578r (l"9F7T"; ""

PAGE 205

196 197. D.C. Borg, R.E. Segel, P. Kienle and L. Campbell, Int. J. Appl. Radiat. Isotopes , 11 , 10 (1961) . 198. R.D. Cooper and G.L. Brownell, Nuc l. Instrum. Methods, 51, 72 (1967). 199. H.P. Yule, Anal. Chem . , 38.^ 818 (1966). 200. W.W. Meinke, "L'Analyse par Radioactivation et ses Applications aux Sciences Biologiques , " Presses Universitaires de France, Paris, 1964, p 145. 201. T.T. Gorsuch, Analyst (London), 8£, 135 (1959).202. D.S. Robertson, Nature , 210 , 1357 (1966). 203. H.J.M. Bowen, Analyst (London), 92_, 118 (1967). 204. K. Heydorn and E. Damsgaard, Talanta, 20_, 1 (1973). 205. M. Rakovic, Chem. Zvesti , 23, 147 (1969); Chem. Abstr. 71, 687v (19^57"^ — 206. M. Rakovic, I. Langerova and H. Talpova, Atompraxis , 12_, 344 (1966); Chem. Abstr ., 65, 18983e "(1966) . 207. M. Rakovic and Z. Prochazkova, Nucl. Med ., 5_, 436 (1966) 208. M. Rakovic, Ch em. Zvesti , 20, 293 (1966); Chem. Abstr., £5, 2987d (19667"! 209. H.J.M. Bowen and P. A. Cawse, Analyst (London), 86, 506 (1961). — 210. L. Kosta and A.R. Byrne, Talanta , 16, 1297 (1969). 211. W.Gebaur, Radiochim. Acta , 4, 191 (1965). 212. W.J. Ross, Anal. Chem. , 3£, 1114 (1964). 213. K. Samsahl, Anal. Chem. , 39 , 1480 (1967). 214. K. Samsahl, "A Fast Radiochemical Method for the Determination of Some Essential Trace Elements in Biology and Medicine," AE-16 8, Aktiebolaget Atomenergi, Stockholm, Sweden, 1964. 215. F.J. Berlandi, Diss. Abstr . , 27, 2264B (1966). 216. J.I. Kim and H. Stark, Radiochim. Acta , 13, 213 (1970). 217. D. Behne, Radiochem. Radioanal . Lett ., 6_, 39 (1971). 218. D. Behne, Kerntechnik, 12, 112 (1970).

PAGE 206

197 219. F. Grass and R. Kittl, Mikrorchim. Acta , 371 (1971). 220. K.G. Kjellin, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p 517. 221. J. Op de Beeck, Anal. Chim. Acta , 40, 221 (1968). 222. E.L. Kanabrocki, L.F. Case, T. Fields, L. Graham, E.B. Miller, Y.T. Oester and E. Kaplan, J. Nucl. Med., 6, 489 (1965) . 223, 224 226 228 229 230 232 234 236 E.L. Kanabrocki, L.F. Case, T. Fields, L. Graham, E.B. Miller, Y.T. Oester, and E. Kaplan, J. Nucl. Med., 6, 780 (1965). E.L. Kanabrocki, L.F. Case, E.B. Miller, E. Kaplan and Y.T. Oester, J. Nucl. Med . , 5, 643 (1964). 225. G.J. Olivares, Acta Cient. Venez., 18, 9 (1967). T. Freeman and J. Smith, Biochem. J ., 118 , 869 (1970). 227. C. Veselsky, M. Nedbalek, and O. Suschny, Int. J . Appl Radiat. Isotopes, 21, 225 (1970). ~ K. Fritze, and R. Robertson, J. Radioanal. Chem. , 1, 463 (1968). T.K. Lin and S.J. Yeh, Nucl. Sci. Technol . (Tokyo). 3, 289 (1966). ^ J. Torko, Energ. Atomtech ., 20 , 200 (1967). 231. D. Brune, "Recoil Reactions in Neutron Activation Analysis. I. The Szilard-Chalmers Effect Applied in. the Analysis of Biological Samples. II. Transfer of Activities from Container Materials to Sample," AE-172, Aktiebolaget Atomenergi , Stockholm, Sweden, 1965; Chem. Abstr. , 63, 2113 (1965). D. Brune, Anal. Chim. Acta, 34, 447 (1966) 233. D. Comar, and C. Le Poec, in "Proceedings of the 1964 Symposium on Radiochemical Methods of Analysis," Salzburg, Austria, 1965, p 15. J. Shankar, K.S. Venkateswarlu and M, Lai, J. Inorg . Nucl. Chem., 28, 11 (1966). 235. K.A. Rao and N. Nath, Radiochim. Acta, 5, 162 (1966). E. Lazzarini, J. Inorg. Nucl. Chem ., 29, 7 (1967).

PAGE 207

198 237. I. Gainar and A. Ponta, Rev. Roum. Chim . , 13 , 401 (1968) 238. J. Shankar, "Hot Atom Chemistry: Annealing of Recoil Damage and Solid State Isotopic Exchange in Cobalt Complexes," BARC-348, Bhabha Atomic Research Centre, Bombay, India, 1968. 239. J. Shankar, A. Nath and V.G. Thomas, J. Inorg. Nucl . Chem. , 20^ 1361 (1968) . 240. G.H. Morrison and H. Preiser, "Solvent Extraction in Analytical Chemistry," J. Wiley and Sons, New York, N.Y., 1962. 241. J. Stary, "The Solvent Extraction of Metal Chelates," Pergamon Press, New York, N.Y., 1964. 242. F. Kukula, B. Mudrova and M. Krivanek, Talanta , 14, 233 (1967). 243. A. Allan and R. Parthasarathy , Anal. Chim. Acta , 35, 69 (1966). 244. K.J. Hahn, D.J. Tuma and J.L. Sullivan, Anal. Chem. , £0, 974 (1968). 245. K.J. Hahn, D.J. Tuma and M.A. Quaife, Anal. Chem . , 39, 1169 (1967). 246. H. Green, Talanta, 20^, 139 (1973). 247. S. Ohno, Analyst (London), 9_6, 423 (1971). 248. H. VJesch, J. Ziramerer and J. Schmuamacher , Int. J. Appl . Radiat. Isotopes , 2_1, 431 (1970) . 249. F.' Girardi, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p 117. 250. F. Girardi, R. Pietra and E. Sabbioni, in "Proceedings of the 1968 International Conference on Modern Trends in Activation Analysis," National Bureau of Standards, Gaithersburg Md. , 1968, p 639. 251. F. Girardi, R. Pietra, and E. Sabbioni, J. Radioanal . Chem . , 5, 141 (1970). 252. F. Girardi, R. Pietra and E. Sabbioni, "Radiochemical Separation by Retention on Ionic Precipitates. Adsorption Tests on 17 Materials," EUR-4287; Chem. Abstr . 72, 73549g (1970).

PAGE 208

199 256 257 258 253. C. Bigliocca, F. Girardi, J. Pauly, E. Sabbioni , S. Meloni and A. Provasoli , Anal. Chem. , 39 , 1634 (1967). 254. F. Girardi and E. Sabbioni, J. R adioanal. Chem., 1, 169 (1968). 255. C.W. Tang and C.J. Maletskos , Science , 167, 52 (1970). C.K. Kim and J. Silverman, Anal. Chem ., 37 , 1616 (1965). C.K. Kim and J. Silverman, Trans. Amer . Nucl . Soc, 7, 332 (1964) . D. Brune, Acta Chem. Scand . , 20 , 1200 (1966). 259. D. Brune and K. Jirlow, Radiochim. Acta , 8^, 161 (1967). 260. R.R. Ruch and R. DeVoe , Anal. Chem ., 39 , 1333 (1967). 261. J.R. DeVoe, C.K. Kim and W.W. Meinke, Talanta, 3, 29 8 (1960) . 262. D.L. Massart, "Cation-Exchange Techniques in Radiochemistry," NAS-NS-3113 (Rev.), Nuclear Science Series of the National Academy of Sciences-National Research Council, Washington, D.C., 1971. 263. J. Inczedy, "Analytical Applications of Ion Exchangers," Pergamon Press, New York, N.Y., 19 66. 264. G. Schmuckler, Talanta , 12 , 281 (1965). 265. J. P. Riley and D. Taylor, Anal. Chim. Acta, 40, 479 tl968). — 266. C.W. Blount, D.E. Leyden , T.L. Thomas and S.M. Guill, Anal. Chem ., 45 , 1045 (1973). 267. K. Fritze, N. Aspin, and T.H. Holmes, Radiochim. Acta., 3, 204 (1964). G. Aubouin and J. Laverlochere , J. Radioanal. Chem., 1, 123 (1968) . ~ R. Griessbach, Angew. Chem ., 67 , 606 (1955). R. Dybczynski, J. Chromatogr ., 31 , 155 (1967). 271. F. Tera, R.R. Ruch and G.H. Morrison, Anal. Chem., 37, 358 (1965). — 272. F, Tera and G.H. Morrison, Anal. Chem ., 38_, 959 (1966) 268 269 270

PAGE 209

200 273. J.C. Ricq, J. Radioanal. Chem. , 1, 443 (1968). 274. V.V. Moiseev, R.A. Kuznetsov and A.I. Kalinin, in "Proceedings of the 19 65 International Conference on Modern Trends in Activation Analysis," College Station, Texas, 1965, p 164. 275. D. Comar and C. Le Poec, in "Proceedings of the 1965 International Conference on Modern Trends in Activation Analysis," College Station, Texas, 1965, p 351. 276 280 282 283 284. K. Samsahl and D. Brune , Int. J . Appl . Radiat . Isotopes, 16, 273 (1965). — 277. K. Samsahl, P.O. Wester and 0. Landstrom, Anal. Chem., 40, 181 (1968). 278. K. Samsahl, "High-Speed, Automatic Radiochemical Separations for Analysis in the Biological and Medical Research Laboratory," AE-389, Aktiebolaget Atomenergi, Stockholm, Sweden, 1970; Chem. Abstr ., 7_3, 95361n (1970) 279. K. Samsahl, Nukleonic, 8, 252 (1966). G.H. Morrison, J.T. Gerard, A.T. Kashuba, E.V. Gangadharam, A.M. Rothenberg, N.M. Potter and G.B. Miller Science , 167, 505 (1970) . ' 281. G.H. Morrison, J.T. Gerard, A. Travesti, R.L. Currie, S.F. Peterson and N.M. Potter, Anal. Chem., 41, 1633 (1969). — F. Girardi, Cron. Chim . , No. 27, 14 (1970); Chem. Abstr., 73, 136447r (1970) . P. Vanderkooi, B. Heagan, and J.T. Lowman , Proc . Soc . Exp. Biol. Med . , 113, 772 (1963). J. Hadzistelios and C. Papadopoulou, Talanta, 16, 337 (1969) . — ' 285. P. Van den Winkel, A. Speecke and J.J. Hoste, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p 569. 286. N. Spronk, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967, p 335. 287. M. Heurtebise, J. Radioanal. Chem . , 7, 227 (1971). 2 88. M.H. Feldman, J. McNamara, R.C. Reba and W. Webster, J. Nucl. Med. , 8, 123 (1967).

PAGE 210

201 289. M. Heurtebise and W.J. Ross, Anal. Chem. , 43, 1438 (1971) . ~ ' — ' 290. R.E. Jervis and K.Y. Wong, in "Nuclear Activation Techniques in the Life Sciences," Proceedings of a Symposium, Amsterdam, 1967, IAEA, Vienna, 1967 p 137. 291. U. Eisner, J.M. Rottschafer, F.J. Berlandi and H.B. Mark, Jr., Anal. Chem ., 38 , 1466 (1967). 292. A.Z. Budzynski and J.Z. Beer, Anal. Chim. Acta , 46_, 293. T. Nascutiu, Rev. Roum. Chim ., 9, 283 (1964). 294. T. Nascutiu, Acad Repub. Pop. Rom. . Stud. Cercet, Chinu_, 10, 27b (1962; Chem. Abstr . 55, 2841q (1^6 3). 295. W. Bock-Werthmann, Anal. Chim. Acta , 28_, 519 (1963). 296. M. Qureshi and S.D. Sharma, Anal. Chem. , 45_, 1283 (1973). 297. J.M. Steim and A. A. Benson, Anal. Biochem , 9, 21 (1964). 298. J.M. Steim, Diss. Abstr ., 23, 3622 (1963). 299. A. A. Benson, W.W. Miller, and J.M. Steim, Nippon Isotope Kaigi Hobun-shu , 119 (1963) . 300. R Blomstrand and F. Nakayama, J. Neurochem., 8, 230 (1961) . " — — 301. J B. Smathers, D. Duffy and S. Lakshamanan , Anal. Chim. Acta , 4_6, 9 (1969) . 302. H. Staerk and D. Knorr, Atomkernenergie , 6, 408 (1961)Chem. Abstr . , 56_, 3765b (1962) . " " 303. J.B. Smathers, D. Duffy and S. Lakshmanan, Anal. Chim. Acta , 39, 529 (1967). W.T. Burnett, Jr., and M.D. Cohan, Trans. Amer. Nucl Soc. , 7, 333 (1964). " 92^"34ra967)°''''''^'' ^""^ R-J.T. Graham, Analyst (London ), L.S. Bark, G. Duncan, and R.J.T. Graham, Analyst (London), 9_2, 31 (1967). 304 305, 306. 307. S.P. Cram, Advan . Chromatogr ., 9, 243 (1970)

PAGE 211

202 30 8. S.P. Cram and J.L. Brownlee, Jr., J. Gas Chr oma togr . , 5, 353 (1967) . 309. S.P. Cram and J.L. Brownlee, Jr., J. G as Chromatogr , , 6, 305 (1968) . 310. S.P. Cram and J.L, Brownlee, Jr., J. Ga s. Chromatogr., 6, 313 (1968) . 311. D.B. Cottrell, Ph.D. Thesis, The University of Florida, Gainesville, Florida, 1973. 312. E.W. Berg and J.T. Truemper, J. Phys , Chem. , 64, 487 (1960). — 313. M. Tanaka, T. Shono , and K. Shinra, Nippon Kagaku Zasshi , 89, 669 (1968). 314. R.W. Moshier and R.E. Sievers , "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y. 1965, p. 4. 315. C.E. Crouthamel, "Applied Gamma-Ray Spectrometry," 2nd ed, completely revised and enlarged by F. Adams and R. Dams, Pergamon Press, Ltd., Elmsford, N.Y. 1970. 316. J. P. Collman, in "Inorganic Synthesis," Vol 7, McGraw-Hill Book Company, New York, N.Y., 1963, p 134. 317. R.W. Moshier and R.E. Sievers, "Gas Chromatography of Metal Chelates," Pergamon Press, New York, N.Y., 1965, p 118.

PAGE 212

BIOGRAPHICAL SKETCH Frederick Turner Varcoe, Jr., was born April 27, 1944, at Chattanooga, Tennessee. In 1953 he and his parents, an actuary and a mathematics instructor, moved to Indianapolis, Indiana. He returned to Chattanooga to attend The McCallie School, graduating in May, 1962. He and his family then moved to Jackson, Mississippi, where he received the degree of Bachelor of Science cum laude from Millsaps College in June, 1966. He entered the University of Florida in September, 1966, and since that time he has been working toward the degree of Doctor of Philosophy. While in Gainesville he was active in the Gator Sail Club, the Gator Ski Club, and worked in a drug abuse rehabilitation program at the Drug Project. He is a member of the American Chemical Society and Alpha Chi Sigma. 203

PAGE 213

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Roger (^J. Bates, Chairman Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. "TT' ( ^\r Jaraes/D. Wmefordner Processor of Chemistry V^ I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. P.obert J. HaiYranan Professor o» Chem.istry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. '^A, Hs '"7:^ ^ t<-?T^ ^ Je^n/H. J^ Fames Ks0bc . Profess61:^ of Management

PAGE 214

This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June, 1974 Dean, Graduate School

PAGE 215

UNIVERSITY OF FLORIDA 3 1262 08553 0375