Radiochromatographic analysis of fresh water resources

MISSING IMAGE

Material Information

Title:
Radiochromatographic analysis of fresh water resources
Series Title:
Florida Water Resources Research Center Publication Number 15
Physical Description:
Book
Creator:
Cram, Stuart P.
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Notes

Abstract:
The study of radiochromatographic separations for the neutron activation analysis of trace level metals in fresh water sources is reported. Chromatographic separations of metal beta-diketonates were developed for post-irradiation separations. When the trace metals were complexed before irradiation, radiation degradation was found to be a function of the solution concentration, presence of excess ligand, irradiation time, and the neutron flux spectrum. Trifluoroacetylacetone and heptafluorodimethyloctanedione complexes of chromium were quantitatively recovered after irradiation. Quantitative elution from the chromatographic column of the beta-diketonate complexes of Cr, Mn, Fe, Be, Lu, Gd, and Cu were studied and found to yield recoveries between 52 and 98%. Extensive studies are reported which describe the optimum conditions for separation and account for decomposition and adsorption losses in the system. The development of sampling systems, counting geometries, and sample transfer lines are reported which must be carefully considered when analyzing multicomponent metal chelate mixtures.
General Note:
Florida Water Resources Research Center 15

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the source institution and holding location.
System ID:
AA00001465:00001


This item is only available as the following downloads:


Full Text











/ j
I. / IIUk 2. ,


Publication No. 15
The Radiochromatographic Analysis of
Fresh Water Resources


By
Stuart P. Cram
Department of Chemistry






Water Resources Research
University of Florida


















THE RADIOCHROMATOGRAPHIC ANALYSIS OF
FRESH WATER RESOURCES

by



STUART P. CRAM
(Principal Investigator)


PUBLICATION NO. 15

FLORIDA WATER RESOURCES RESEARCH CENTER






RESEARCH PROJECT TECHNICAL COMPLETION REPORT


OWRR Project Number A-012-FLA


Annual Allotment Agreement Numbers

14-01-0001-1628 (1969)
14-31-0001-3009 (1970)
14-31-0001-3209 (1971)

Report Submitted: October 28, 1971






The work upon which this report is based was supported in part
by funds provided by the United States Department of the
Interior, Office of Water Resources Research as
Authorized under the Water Resources
Research Act of 1964.











TABLE OF CONTENTS

Page

ABSTRACT ----------------------------------------------- iii

PROJECT PUBLICATIONS -------------------------------- iv

INTRODUCTION ----------------------------------------- 1

PART I. CHARACTERIZATION OF THE IRRADIATION STABILITY
OF SOME CHROMIUM BETA-DIKETONATES FOR GAS CHROMATO-
GRAPHIC SEPARATIONS IN NEUTRON ACTIVATION ANALYSIS 8

1. Experimental ----------------------------- 8
2. Results and Discussion ------------------- 13

PART II. THE QUANTITATIVE GAS CHROMATOGRAPHIC
SEPARATION AND ANALYSIS OF RADIOACTIVE, VOLATILE
METAL BETA-DIKETONATES ----------------------------- 21

1. Experimental ---------------------------- 21
2. Quantitative Column Elution -------------- 22
3. Counting System -------------------------- 25
4. Chelation Studies ------------------------ 29
5. Results and Discussion ------------------- 31

ACKNOWLEDGEMENTS ------------------------------------- 44

LITERATURE CITED ------------------------------------- 45











ABSTRACT


The study of radiochromatographic separations for
the neutron activation analysis of trace level metals in
fresh water sources is reported. Chromatographic separations
of metal beta-diketonates were developed for post-irradiation
separations. When the trace metals were completed before
irradiation, radiation degradation was found to be a function
of the solution concentration, presence of excess ligand,
irradiation time, and the neutron flux spectrum. Trifluoro-
acetylacetone and heptafluorodimethyloctanedione complexes
of chromium were quantitatively recovered after irradiation.

Quantitative elution from the chromatographic column
of the beta-diketonate complexes of Cr, Mn, Fe, Be, Lu, Gd,
and Cu were studied and found to yield recoveries between
52 and 98%. Extensive studies are reported which describe
the optimum conditions for separation and account for de-
composition and adsorption losses in the system. The
development of sampling systems, counting geometries, and
sample transfer lines are reported which must be carefully
considered when analyzing multicomponent metal chelate
mixtures,




















Cram, Stuart P.
THE RADIOCHROMATOGRAPHIC ANALYSIS OF FRESH WATER RESOURCES
Completion Report to the Office of Water Resources Research,
Department of the Interior, October, 1971, Washington, D.C. 20240
KEYWORDS: neutron activation*/ gas chromatography*/
activation-chromatography coupling*/ fresh water/ trace analysis.


iii











PROJECT PUBLICATIONS


1. Cram, S. P., and Varcoe, F. T., "Gas Chromatographic
Separations in Neutron Activation Analysis," Proc.
Int'l. Symp. on Modern Trends in Act. Anal.,
Gaithersburg,_ Maryland.

2. Wade, R. L., and Cram, S. P., "Quantitative Interpreta-
tion of Semilogarithmic Gas Chromatographic Data,"
Anal. Chem., 41, 893 (1969).

3. Glenn, T. H., and Cram, S. P., "A Digital Logic System
for the Evaluation of Instrumental Contributions to
Chromatographic Band Broadening, J. Chromatog. Sci.,
8, 46 (1970).

4. Juvet, R. S., and Cram, S. P., "Gas Chromatography,"
Anal. Chem., 42, 1R (1970).

5. Booher, T. R., and Cram, S. P., "Characterization of
the Irradiation Stability of Some Chromium Beta-
Diketonates for Gas Chromatographic Separations in
Neutron Activation Analysis," J. Radioanal. Chem.,
submitted for publication.

6. McCoy, R. W., and Cram, S. P., "Extention of the Time
Normalization Theory of Gas Chromatography for the
Minimization of Analysis Time," Anal. Chem.,
submitted for publication.

7. McCoy, R. W., and Cram, S. P., "High Speed Gas Chromato-
graphic Separations of Volatile Metal Beta-Diketonates,"
Anal. Chem., submitted for publication.

8. Cottrell, D. B., and Cram, S. P., "A High Precision
Digital Data Acquisition System for Neutron Activa-
tion Analysis," in preparation.

9. Boerner, B. R., and Cram, S. P.', "Absolute Neutron
Activation Analysis," in preparation.











INTRODUCTION


Research in the field of water resources and water
quality is highly dependent upon the development of new
analytical techniques which are specific, sensitive, repro-
ducible, rapid, and suitable for automation. Because of
the shortage of analytical chemists in this country, and
particularly analytical chemists working in water quality
research, new analytical methods of analysis which incorpo-
rate the above criteria have not been designed and developed
to solve many of the existing problems in this field. The
work discussed in this report is directed toward the
development of two such analytical techniques, i.e., neutron
activation analysis and gas-liquid chromatography, and the
coupling of the two complimentary techniques for ultra-
trace metal analysis in aqueous systems.

Numerous instrumental methods for elemental analyses
in water samples have been reported (1-5), although the
number of methods applicable to trace analysis (defined here
as being less than 1 pgr of any given material)-is consider-
ably more restrictive. The methods of elemental analyses
reported to date have generally been very limited in scope,
and the breadth of these techniques at the trace concentra-
tion level in water analysis has been limited by a number
of considerations. First, the wide diversity of matrix
materials represents a limitation on the sensitivity of any
chemical analysis, as the relationship between the chemical
and physical interactions are affected by the environment
of the element to be determined. Further, most techniques
are subject to interference by the presence of other
chemicals. The sensitivity of these methods is therefore
dependent upon the chemical nature of the sample and varies
from one sample to the next. This effect is particularly
pronounced at trace concentrations.

Neutron activation has likewise been investigated
as an analytical tool for the element analysis of aqueous
samples (6-8). These investigations have been summarized
in Table 1 to illustrate the general applicability of the
technique at the trace level. The sensitivities reported
do not represent optimum experimental conditions, and
indicate the sensitivities only for irradiation times of
one hour at a thermal neutron flux of 4.3 x 1012 neutrons/
cm2 sec.












TABLE 1

SENSITIVITY AND APPLICABILITY OF THERMAL NEUTRON
ACTIVATION FOR ELEMENTAL ANALYSES


Minimum Detection
Limit Ranges, pgr


Mn, In, I, Au, Eu, Dy


10-s 10-4

10-4 10-3


10-3 10-2



10-2 10-1



10-1 100


100

101


- 101

- 102


102 103


Ar, Sc, V, Br, Cs, Hf, Ir, Sm,
Ho, Er

Na, Al, Co, Cu, As, Se, Sr, Ga,
Rh, Ag, Cd, Sn, Sb, Ba, W, La,
Nd, Yb, Lu

C1, K, Ti, Zn, Ge, Mo, Pd, Ta,
Pt, Hg, Ce, Ru, Rb, Zr, Te, Pr,
Tb

F, Mg, Cr, Ni, Nb, Pb, Y, Tm


Ca, Os

Si, S

O, Fe










Thus it can be seen that neutron activation offers
the advantages of dynamic applicability and high sensitivity.
Further, the sensitivity of the method is invariant with the
nature, size and concentration of the sample. The sensiti-
vity is determined by the nuclear parameters of the element,
such as the half-life, cross-section, neutron flux, etc.,
and are constant regardless of the chemical system being
investigated.

The inherent sensitivity and precision of neutron
activation analysis at the submicrogram concentration level
is dependent upon the accuracy with which the activated, gamma
emitting radionuclides can be identified and measured. Many
elements have sensitivities that do not differ by more than
a factor of 100, and consequently, the bulk of the photo-
electric peaks are easily masked when multiple activities
are induced by activation (9). This interference is
especially serious when the major elemental constituents
with large thermal neutron cross-sections are activated.

Partial resolution of gamma-ray spectra can be
obtained by variation of the experimental conditions, such
as the irradiation and decay times. This approach is not
practical when a sample contains even trace amounts of Na,
Al, V, Mn, Br, I, Au, etc., because these elements have
large activation cross-sections and reach very large specific
activities during neutron irradiation, subsequently masking
the activities of the other radionuclides in the pulse
height spectrum. Although Ge(Li) scintillation detectors
considerably enhance the resolution of gamma ray spectra,
their depletion depths presently limit their physical size
and thereby their counting efficiency. Numerous proposals
(10-12) have been reported wherein the instrumental
resolution and interpretation of a gamma ray spectrum have
been simplified by post-irradiation chemical separations.

The importance of post-irradiation separations in
neutron activation analysis has been reviewed (9) and
clearly substantiated in the literature (13-16). Currently,
solvent extraction (17, 18), extraction with hydrated
antimony pentoxide (HAP) (19), and-adsorption on MnO2 (20,
21) are the most common non-chromatographic separation
techniques. Paper chromatography (22), thin layer chromato-
graphy (23, 24), ion exchange (25) and ion exchange membranes
(26), and isotopic exchange (27) have been successfully
combined with gamma ray spectrometry in order to give
significant improvements in the analyses.










An automatic group separation system for the determi-
nation of 40 elements was reported by Samsahl, et al. (28),
which is based on ion exchange chromatography an-d solvent
extraction. Recoveries of >90% for 29 elements in biologi-
cal samples in about 2 hours were reported with a mean error
of 3% in the analyses. A number of automatic or semiauto-
matic group separation systems for different kinds of
materials have been reported (28-31). Perhaps the most
extensive and analytically useful scheme has been reported
by Morrison, et al. (32,33). They have developed a group
separation scheme which enabled them to measure 41 trace
level elements in the Apollo 11 lunar samples by NAA. Their
method combines solvent extraction, separation with HAP,
and ion exchange to give six groups which can be counted
without serious interference for trace amounts.

Gas and liquid chromatographic separations can be
expected to become of increasing importance in neutron
activation analysis (NAA) for the separation of radionuclide
interference in complex gamma ray scintillation spectra.
Chromatographic separations are rapid, highly efficient, and
can be developed without the use of carriers. Further, these
separations can be optimized with respect to time, resolution,
or sample size as required so that the separation -f short-
lived isotopes, multicomponent mixtures, or trace constitu-
ents, respectively, may be effected. These separations are
very important in trace analysis as they effectively give
an infinite resolution of the pulse height spectra and enable
the theoretical lower limits of detection to be realized.
The popularity of the Ge(Li) detector is attributed to its
high resolution and the minimum amount of "wet" chemical
separations required. Gas chromatography with large volume
NaI(T1) detectors, however, can be automated, and even
operated remotely, and therefore possesses all of the
advantages of Ge(Li) detectors and has the additional
advantage of having a much higher counting efficiency.

Gas chromatography provides a quantitative post-
irradiation separation and effectively enchances the
sensitivity of activation analysis by:

1. Removing matrix activities
2. Resolving overlapping photopeaks
3. Reducing the dead time of the analyzer
4. Isolating single radionuclide species in order
to increase the quantitative precision of
pulse height analysis and reduce the amount of
interpretation of pulse height spectra and
data reduction.









Only a few papers have sought to develop gas and/or
liquid partition chromatography for the separation step in
NAA. From a recent review of radiochromatographic separa-
tions and analyses, it is clear that radiochromatographic
separations have been used extensively in radiation chemistry,
hot atom chemistry, and biochemistry but not for the separa-
tion of photon emitting radionuclides (34). Cram and
Brownlee (35-37) first reported the use of gas-liquid chro-
matography for the clean-up of pulse height spectra in the
functional group analyses of the halogens by (n,y) reactions.
The advantages and workability of gas chromatography for
high speed post-irradiation separations was demonstrated by
eliminating the interference Na24, Al28, Br80, and 1128 from
the F20 spectrum. Further, this previous work resolved the
1.63 Mev photopeak of F20 from the 1.64 Mev peak of Cl38 in
less than 5 seconds so that the analysis for fluorine with
the 10.7 second half-life could be used directly. For the
activation analysis of fluorine with an (n,y) reaction,
high speed and high efficiency separations are imperative
if meaningful analytical results are to be obtained. Even
though the higher atomic weight members of the halogen series
(i.e., Br and I) have considerably longer half-lives and
larger activation cross-sections, they too must be chromato-
graphically separated for analysis at trace concentrations.

However, organic analyses are rather limited in scope
if one is restricted to (n,y) reactions, and thus it has
been necessary to develop gas phase, partition separations
for the activation analysis of the metallic elements. (As
in the case of the organic functional group analyses, the
materials to be separated must have a vapor pressure of at
least 1.0 mm Hg at the separation temperature, and must
be thermally stable at that temperature.)

Metal analysis by gas chromatography has been shown
to be both practical and useful (38). Volatile metal
chelates such as the beta-diketonates, have been most often
used in these analyses. The use of metal chelates has
enabled more than three quarters of the elements in the
periodic table to be separated and eluted from a gas
chromatograph easily, quickly, and with high resolution.
Trace analysis studies (39-41) and'the development of new
ligands (42) have helped make metal analysis by gas
chromatography a powerful tool with a wide range of applic-
ability. It is the inherent speed and selectivity of gas
chromatography combined with the chemistry of the metal
chelates which makes the technique such an excellent
separation method for neutron activation analysis.










The synthesis of the metal chelates using beta-diketone
ligands such as 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)] is
usually simple. The chelates can be made either before or
after irradiation of the metal mixture. Post-irradiation
synthesis has the advantage of avoiding irradiation of the
ligand and chelate. Each radionuclide is chelated quantita-
tively and is therefore in a form suitable for efficient
separation. This advantage is often outweighed by the
disadvantage of increased decay time due to additional sample
treatment of the radionuclide between irradiation and
counting. In order to take full advantage of the speed
inherent in the technique, chelation before irradiation is
usually necessary.

The irradiation stability of the metal chelates is
important and corrections must be made for decomposition in
the reactor. The radiolysis of reactor irradiated Cr
chelates of [H(tfa)], [H(hfa)], and [H(fod)] has been studied
(43). No thermal degradation in the reactor was found, but
radiolysis losses of up to 10% were reported when the
irradiation time was only 1%- of one half-life-of chromium.
It has been shown that Szilard-Chalmers reactions involving
the central metal atom of a metal beta-diketonate yield
retentions (percentage of radioactive atoms in the original
chemical compound) as low as 4% (44).

Szilard-Chalmers reactions involving solid metal
chelates and the use of annealing to increase their reten-
tions have been studied (44-58). The recoil and metal-
ligand bond cleavage of chelates after (n,y) reactions have
been used for iostope enrichment (45,46) and the metal is
believed to exist as a cation in solution (44). Szilard-
Chalmers reactions following (y,n) reactions have also been
discussed (47). Lazzarini (48) has studied the effect of
the ligand field on Szilard-Chalmers reactions by studying
the retention for 28 complexes of Co(III). Annealing
studies have been carried out on Cr tris-acetylacetonate
(49-51) and Co tris-acetylacetonate T(5-56), which are
similar to the chelates used in this work. Repair of
reactor irradiation damage can result from heating the
chelate (44, 49-58) or exposing it to y-radiation (51,54).
Thus, there is a moderate amount of annealing before a
sample leaves the reactor. For the Co tris-acetylacetonate,
each type of treatment has about the same effect (54), which
is reported to be due to localized heating during y-
irradiation. Preheating of solid samples before neutron
irradiation will also increase retention (55), and the
presence of an electron donating atmosphere around a solid
chelate during annealing has been shown to increase
retention (44,57).









Gainer and Ponta (58) have studied thermal annealing
of Cr tris-acetylacetonate in methanol solutions. The
retention was 32% after 2 hours of irradiation at 2.3 x 1011
neutrons/cm2-sec, and increased with heating. In the plots
of retention versus annealing time at constant temperature,
the plateous had heights which were temperature dependent.
At 950 C, 80% retention was obtained, but above 95 C, the
retention reached a peak and fell off due to decomposition.
Gainer and Ponta also found exchange between 51Cr3+ in
solution and the chromium in the chelate, The work described
in this report used both the annealing described by Gainer
and Ponta and the technique of adding excess ligand to the
solution to be irradiated in an effort to obtain quantita-
tive yields of the irradiated chelate prior to gas chromato-
graphic separation.












PART I


CHARACTERIZATION OF THE IRRADIATION STABILITY OF SOME
CHROMIUM BETA-DIKETONATES FOR GAS CHROMATOGRAPHIC
SEPARATIONS IN NEUTRON ACTIVATION ANALYSIS


Chromium beta-diketonates were chosen as model
compounds for this study because the gas chromatographic
separations of these materials have been studied extensively
(38,40,59-64) and because of their physiological and
toxicological importance at trace concentrations (65-81).
Trifluoroacetylacetonate [Cr(tfa) ], hexafluoroacetylacetonate
[Cr(hfa)3] and heptafluorodimethyloctanedione [Cr(fod)3]
complexes of chromium were studied in order to measure the
ligand effect on the stability during and after the irradia-
tion, and the reactivity towards recombination reactions
in the presence of excess ligand. The quantitative
recovery of the chelates is then reported as a function of
the type of ligand, irradiation time, solution concentra-
tion, and the effect of the presence of excess ligand.


1. Experimental


The chromium chelates were irradiated as standard
benzene (Nanograde, Mallinckrodt Chemical Works) solutions.
The Cr(hfa)3 was purchased from Pierce Chemical Company,
and the Cr(tfa)3 and Cr(fod)3 were synthesized following
published procedures (63,64). The ligands for the
syntheses were purchased from Peninsular Chem. Research
Corp., and the metals from Fisher Scientific Company.

Quartz (General Electric Company) break seal vials
were drawn from 9 mm tubing and used as sample containers.
These vials were leached, filled with the solution to be
irradiated, and sealed with a torch after the solutions
were frozen out. In this manner the vials were sealed
air tight and, because of the low sodium concentration,
could be handled immediately after irradiation.

The irradiations were performed in the University
of Florida Training Reactor (UFTR) and the Georgia
Institute of Technology Research Reactor (GTRR). The UFTR
is a water cooled, MTR type reactor. At a steady state
power level of 60 Kw and a thermal neutron flux of 1 x 1012
neutrons/cm2-sec, the temperature in the sample position
was 60.5' C as measured with an iron-constantan thermocouple.









The GTRR is a heterogeneous, 1 Mw, D20 moderated and cooled
reactor. Since the outlet coolant temperature of the GTRR
was only 1 C higher than the UFTR, the irradiation
temperatures were assumed to be the same in both reactors.
A comparison of the irradiation facilities used is given
in Table 2. It should be noted that the irradiation times
were adjusted to give the same integrated flux in both
reactors.

The gamma ray pulse height spectra were measured
with two 3 inch x 3 inch Nal (Tl) crystals (Type 12S12/E-X,
Harshaw Chemical Company). Nuclear Chicago Model 10-17
scintillation preamplifiers were used with RC amplifiers
(Model 27001, Nuclear Chicago Corp.), superimposed in a
mixer amplifier (Model 30-35, Nuclear Chicago Corp.), and
counted in a 400 channel pulse height analyzer (Model
34-27, Nuclear Chicago). Data was readout on an ASR Model
33 Teletype and an X-Y plotter (Model 850, Data Equipment
Company).

A four column gas chromatograph (Series 4000,
Victoreen Instrument Company) with an electron capture
detector (Model 4008-3, Victoreen Instrument Company) was
used for the separations. The electron capture -detector
was operated in the DC mode because this model was at
least six times more sensitive for Cr(tfa)3 in the DC mode
than in a pulsed mode, even though the pulse period was
varied from 25 to 125 isec with a pulse width of 0.5 psec.
Also the background noise in the DC mode was less (ca,
3.5 x 10-12 amps). Peak areas were integrated with a
digital integrator (CRS-100, Infortronics Corp.) and a
Hewlett Packard Model 7127-A strip chart recorder was
connected in series with the digital integrator for an
analog display. The chromatograph was modified to bring
the Teflon column into the injection port and to run up to
the sensitive area of the detector in order to avoid
decomposition or adsorption of the metal chelate in the
instrument. The chromatographic system is shown in Figure
1.

An extensive study of column materials and condi-
tions showed that the conditions reported in Table 3 gave
the most reproducible and quantitative results for the
metal chelates and their degradation reaction products.

When the column effluent was to be analyzed by mass
spectrometry, the samples were trapped in V-shaped glass
melting point tubes in a dry ice-acetone bath. For
activation analysis of the effluents, low sodium quartz
melting point tubes (General Electric Company) were used
for the traps. At all other times a two stage absolute
trap was connected to the detector outlet to prevent radio-
active or toxic chelates from being vented to the room.












TABLE 2

IRRADIATION FACILITIES


Irradiation time

Thermal flux

Power level

Moderator

Irradiation position


Georgia Tech University Florida
Research Reactor Training Reactor

7.0 hrs 4.0 hrs

5.6 x 1011 neut/cm2-sec 1 X 1012 neut/cm2-sec

1 Mw 60 Kw

D20 and Graphite Graphite

Vertical thimble Center vertical port
(reflector)


Cd ratio

Gamma flux


2000:1

1 x 106 R/hr


1:1

2.4 x 106 R/hr

























?~'~ 'ITEF i
w I


= TWO STAGE REGULATOR


= NEEDLE VALVE


Figure 1. Block diagram of the gas chromatographic separation system


CARRIER
GAS


r7
E. L 1 C 5! P













TABLE 3

CHROMATOGRAPHIC CONDITIONS FOR THE
SEPARATION OF THE METAL CHELATES


Column = 20 feet of 1/8 inch o.d. Teflon

= 5% QF-l on 80/100 mesh Chromosorb W, AW (DMCS)

Carrier gas flow rate = 60 ml/min of N2

Detector polarizing voltage = 30 volts


Temperature

Injection port

Column

Detector


Cr(hfa) 3


1460 C

1000 C

1500 C


Cr(tfa) 3


Cr(fod) 3


2000 C

1600 C

2080 C


2000 C

1750 C

2080 C










Mass spectra were run at 70 ev. on an RMU-6E mass
spectrometer (Hitachi Div., Perkin Elmer Corp.). The batch
liquid inlet system was heated to 750 C for Cr(hfa)3 and
125 C for the other chelates.


2. Results and Discussion


Since these metal chelates are known to be thermally
unstable, standard benzene solutions of each of the chelates
were prepared, sealed in the quartz vials, and heated in a
laboratory drying oven at 610 C for a period of time equal
to the time of irradiation. Chromatographic analysis of the
solutions with the electron capture detector showed that
there was no measurable thermal degradation at this tempera-
ture. Therefore the degradation which is reported for the
irradiated samples must be due to radiolysis effects alone.

In order to further isolate the variable parameters
which effect chelate degradation, a Cr(hfa)3 solution was
sampled periodically from 1 to 168 hours following irradia-
tion in order to-follow the Szilard-Chalmers effect due to
decay recoil reactions. This study showed that all of the
measurable degradation appears to take place in the reactor,
within the limits of the measurement error.

The effect of the irradiation time on the radiolysis
degradation was also studied. Cr(hfa)3 was chosen for this
study because it was the most labile of the beta-diketonates
studied. Figure 2 shows that the Cr(hfa)3 is markedly more
unstable in a reactor under a thermal neutron flux of
% 1012 neutrons/cm2-sec. Therefore it is not at all feasible
to use the hexafluoroacetylacetone ligand for chromium
analyses by GC-NAA as only about 50% of the chromium will be
recovered for counting after an irradiation corresponding
to b 0.03% of the saturation activity for Cr51.

The determination of the concentration dependence of
the radiation degradation was of paramount importance in
choosing the most stable ligand system and in order to do
quantitative analyses. Concentrations from 2.5 x 10-2
mg/ml to 5.3 mg/ml were used. This range was chosen in
order to give a signal to noise ratio of >100:1 at the low
concentrations (which corresponds to 1.2 x 10-2 pgr of chelate
per 0.5 pl sample injected) and the maximum concentration
was limited by the solubility of the chelate in benzene.
The samples were chromatographically analyzed before and
after irradiation at each concentration and the fraction of
metal chelate remaining was calculated at the average of the
peak areas for at least five replicate determinations.
















11
7A


II V
I is



04 II











0 2 4 6 8 10 12 14

TIME in.
Z3

WI



















curve) in an integrated thermal neutron flux of 1.4 x 1016
neutrons/cm2 -ec
I- -













neutrons/cm2 -sec










Typical chromatograms of Cr(hfa)3 before and after irradia-
tion are shown in Figures 2 and 3. After irradiation of
the low concentration solutions of Cr(hfa)3, three major
peaks were found in the chromatogram. Peak A was identified
as hfa by a comparison of retention times. Peaks B and C
were trapped for analysis by mass spectrometry and gamma
ray spectrometry. The former was identified as Cr(hfa)3
but peak C was not positively identified. The mass spectra
showed a parent peak at a mass to charge ratio of 644 but no
chromium photopeak was found in the pulse height spectra
after extended counting periods.

The ratio of the peak heights of peaks B and C was
found to be a function of concentration as the ratio of B
to C increased with increasing concentration. This is the
expected result and consistent with the Szilard-Chalmers
theory if the major degradation products resulted from
cleavage of the metal-ligand bond. Such an equilibrium
should be affected by the presence of an excess of ligand
in favor of the recombination reaction and give larger
recoveries of the metal chelate.

The results of the CrC(hfa) 3 study are shown in
Figure 4. The percent of metal chelate recovered from a
benzene solution ranges from 81 to 96% and the recovery is
seen to be significantly less in the presence of excess
hfa (from 40.5 to 90.0% recovery). These results are seen
to be consistent at all concentrations but cannot be
explained for this system.

A series of Cr(tfa)3 solutions were also studied in
order to determine the ligand effect. These solutions were
similarly analyzed chromatographically before and after
irradiation. The chromatogram of the Cr(tfa)3 and
Nanogram benzene showed only the solvent and chelate peaks
and no other peaks were discernable above the baseline
noise (3.5 x 1012 amps). Figure 5 shows the results of
this series of experiments. The chelate recoveries ranged
from 90 to 98% with increasing metal chelate concentration
which is in marked comparison with the less stable hfa
chelate. The aliquots containing excess ligand were not
analyzed because the tails of the excess ligand peaks
obscured the remaining chromatogram due to detector
poisoning. Savory et al. (81) used sodium hydroxide to
hydrolyze the excess ligand before analyzing by electron
capture. However, he reported recovering only 80 85%
of the chelate. Since at least 90% of the irradiated
chelate could be recovered without excess ligand and
because of the uncertainty of the hydrolysis effect, the
effect of the excess ligand was not measured.














6.0 r


Q.



cI




Z
0-







0







1-
I-

Q


4.5 -


3.0 --


1.5







r I -- I
I j I I I


I


I881


TIME, min.

Figure 3. Chromatogram of a saturated solution of Cr(hfa)3
in benzene before (dashed line) and after irradiation
(solid line) in an integrated thermal neutron flux of 1.4 x
1016 neutrons/cm2-sec


__l___lLl___l_^________m____L I_~___~~_~_~___ILi-1~11-~-~~11~111~




























- 0














-0- NONIRRADIATED Cr(hfa)3
--a-- IRRADIATED W/O XS. LIGAND
.------ IRRADIATED W/ XS. LIGAND

3--- PERCENT CHELATE RECOVERED W/O XS. LIGAND
--?-- PERCENT CHELATE RECOVERED W/ XS. LIGAND


I a I I


I I I1 I I I Ig L I Is I -r


0.01 .02 .03 .04 .06 .08 0.1 0.2 7," 0.4 0.6 6 1.0 2 3

CONCENTRATICF-N OF Cr(hf- m -,-,-,

Figure 4. Experimental data for the recovery of Cr(lhfa) 3 following thermal
neutron irradiation


4


0
w
aJ








o
0
0
a:

100 W
I-

80 j

60 go

-J
40 c


F-
Z
20
C-)

W
aL


6 8 10


I I if


60


40


20-


fO
0



H--
z
0
0




a
LJ




Q-


IL


." "


' ""











Sif(1. --


20


10o


-0-- NONIRRADIATED Cri;)3

---A-- IRRADIATED CHELATE

-0-- CALCULATED CHELATE RECOVERY


I a I I I_ a a a a a a~ E a I


.02 .03 .04 .0f .08 0.1


0.2 0.3 0.4 0.6 0.8 .O


6 8 10


Ci'OCENTRATION OF


Figure 5. Experimental data for the recovery of Ct(tfa)3 following thermal neutron
irradiation


ro
0

x
:,
I-

0
0



W
O:L
Ld





LU


0
W


0
W
IU




80
w

60 O


40 I-
s


0.01


2--


Cr -':, ,mg/ml









Chromium heptafluorodimethyloctanedione solutions
were run under similar conditions and analyzed as before.
These solutions were also found to be clean before
irradiation within the limits of detection. The recoveries
shown in Figure 6, are seen to increase with concentration
from 90.0 to 100.0%. The results of the Cr(tfa)3 and
Cr(fod)3 studies were particularly pleasing as they
indicate that these chelates are stable to irradiation
times up to 7 hours and are analytically useful for
chromatographic post-irradiation separations in neutron
activation analysis.

The relative irradiation stabilities of the three
chromium beta-diketonates was found to be Cr(fod)3 = Cr(tfa)3
> Cr(hfa)3. The amount of Cr(hfa)3 recovered is more a
function of concentration than are the metal chelates of
the other ligands. The presence of excess ligand appears
to be detrimental to the quantitative recovery of the
Cr(hfa)3. The chromatograms showed no radiolysis peaks to
obscure the chromatogram with either Cr(tfa)3 or Cr(fod)3
and this study has clearly shown these to be the chelates
of choice for chromium analyses.








































-- --- NONIRRADIATED Cr(fod)5

--6-- IRRADIATED CHELATE

----- CALCULATED CHELATE RECOVERY


a a aI a


I a I I I I I


.02 .03 .04 .06 .08 0.1


0.2 0.3 0.4 0.6 0.8 1.0


2 3 4


6 8 10


Figure 6. Ex]
irradiation


perimental data


CONCENT.,.T',T OF Cr(,o'Omg /ml
for the recovery of Cr(fod) 3 following thermal neutron


600


400-


200-





100-

80-

60-


40-


20F


-J

60 -
w
W
m


0.01


I I I


0oo0


VJ 1 I I I I a n n I .


SI-'


m i I












PART II


THE QUANTITATIVE GAS CHROMATOGRAPHIC SEPARATION AND ANALYSIS
OF RADIOACTIVE, VOLATILE METAL BETA-DIKETONATES


It has been shown that in order to obtain reproducible
chromatographic peak areas with the metal chelates, several
injections of the chelate solution must be made to "condition"
the coulmn before quantitative measurements are made (38).
This suggests that the first few injections leave chelate in
the instrument, and only after these chelate adsorption
sites are filled can a sample be eluted quantitatively.
Gainer and Ponta (58) have shown that radioactive chromium
ions in solution will exchange with chelated chromium.
Exchange reactions might also take place between the chelate
in the vapor phase and the chelate held on active sites in
the chromatographic column. Under these conditions, the mass
of chelate may be quantitatively recovered, but there is no
assurance that the activity eluted is representative of the
original- sample.

In order to keep these reactions to a minimum, glass
and Teflon systems have been used. It has been further
shown that metal chelates will react with metal columns,
and it is suspected that metal injection ports will aid and
even catalyze the decomposition. Treatment of the chromato-
graphic system with silanizing agents has also been used in
practice to decrease on-column adsorption.

Radiotracer techniques have been known for a long
time and applied to metal chelate synthesis and extraction
studies (39). However, there has been no application of
the technique to the determination of metal chelate residues
in gas chromatography. This determination of metal chelate
residues is a necessary step in the activation analysis-gas
chromatographic technique.


1. Experimental


The gas chromatograph used in these studies was a
Varian Model 1200-1 equipped with a 1/4 inch injection port,
a hydrogen flame ionization detector and a linear temperature
programmer. The columns were 1/8 inch Teflon (Alpha type,
TFT-250/11), 0.095 inch I.D. and 0.016 inch in wall thickness.
The seals at each end of the column were made by forcing the










column inside a constriction in a Pyrex tube. The injection
port linear was a 3 mm O.D. Pyrex tube tapered to fit inside
the Teflon column. All of the columns and their operating
conditions are given in Table 4.

Liquid samples were injected with a Hamilton 10 ip
syringe and a solid sampler was used to inject the sealed
ampules (64). Samples to be injected using solid sampler
were sealed in quartz tubes 2 cm x 2 mm. Both ends were
sealed with a methane-oxygen torch. After filling, the tube
was placed inside a 1/8 inch copper tube which was placed
in a dry-ice-acetone bath. This prevented decomposition of
the chelate by the heat from the torch.

Solid sample injection has many advantages. There is
no solvent to interfere with trapping the effluent or
obscure the chelate peaks when using a standard gas chromato-
graphic detector. The maximum amount of chelate which can
be placed in the solid sampler without overloading the column
would require too much solvent for a satisfactory injection.
In addition, the sealed quartz tube which contains the
sample during irradiation can be placed directly in the
solid sampling device without-any of the-handling or open
vessels required for the more conventional liquid injection.
Several sampling techniques have been developed for gas
chromatography. The type which uses a plunger to break a
quartz tube containing the sample is a logical choice for
the gas chromatography of irradiated metal chelates.

In order to study the effect of sample size, a series
of injections from 1 10 pl of a benzene solution of 32
mg/ml of Cr(tfa)3 were run on column 6. The H2 flow rate
on the flame ionization detector was 46 ml/min and the 02
flow rate was 135 ml/min.

All peak areas were measured with a digital integra-
tion Model CRS-100 (Infortronics, Inc.).


2. Quantitative Column Elution


To evaluate the effectiveness of silanization of the
gas chromatographic system, a series of 1 pl injections of
a solution with 29 mg/ml of Cr(tfa)3 in benzene were
injected into column 5. The study was repeated with a
Cr(fod)3 solution of the same concentration on column 7.
After 4 to 6 chelate injections, 10 pl of DMCS was injected
to silanize the system.










TABLE 4


CHROMATOGRAPHIC COLUMNS AND CONDITIONS


Literat. Length
Ref. ft


Liquid
Phase


4 5% SE-52


2 20% High
vacuum
grease

4 5% QF-l


4 5% QF-1


4 5% QF-1


4 5% QF-l


4 5% QF-1


4 5% QF-l


3 15% QF-1


Column
No.


He Flow
Rate,
ml/min


Temperature, C
IP. Col. Det.


50 170 80


75 170 150


Chelates


Be(tfa)2


Cr(hfa)3


Solid
Support

60/80 mesh
Gas Chrom. Z

100/120 mesh
AW, DMCS
Chrom. P

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W

80/100 mesh
AW, DMCS
Chrom. W


30 200 175


Cr(fod)3


20 165 100 165 Mn(hfa)2


60 120 100 150 Cr(hfa)3


30 150 100 150 Cr(hfa)3


60 160 135 160 Cr(tfa)3


30 170 150 170 Cr(tfa)3


60 170 150 170 Cr(fod)3











TABLE 4 CONTINUED


Literat.
Ref.


Length
ft


Liquid
Phase


7.5 10% SE-30


2 20% High
vacuum
grease

2 5% SE-30


1.5 20% SE-30



6 10% SE-30


6 10% SE-30


Solid
Support

60/80 mesh
Gas Chrom. Z

100/120 mesh
AW, DMCS
Chrom. P

100/200 mesh,
Chrom. W

100/120 mesh
AW, DMCS
Chrom. P

80/100 mesh
Chrom. W

80/100 mesh
Chrom. W


He Flow
Rate
ml/min


Temperature, C
IP. Col. Det.


30 205 170


75 170 150



60 170 130


30 150 150



30 230 170


30 230 170


Chelates


--- Fe(fod)3


Cu(hfa)2



--- Cu(tfa) 2


--- Cu(hfa) 2



--- Gd(fod)3


--- Lu(fod)3


Column
No.










In order to evaluate the elution of irradiated chelates,
solid samples of Cr(hfa)3 and Cu(hfa)2 were irradiated for 4
hours in the UFTR at %/.5 x 1012 neutrons/cm2-sec and
injected into columns 2 and 11, respectively. One mg solid
samples of Mn(hfa)2 were irradiated 15 min at the same flux
and run on column 9. After the chelates eluted, the solid
sampler, charcoal trap, and column inlet, column, and outlet
system were counted. The quartz fragments in the injection
port were counted, the injection port rinsed with acetone,
and the solution counted. Similar system counting was done
after irradiated solutions had been injected: Cr(hfa)3 in
column 4, Cr(tfa)3 in column 6 and Cr(fod)3 in column 8.

In order to separate the effects of adsorption and on-
column reaction from the radiolysis decomposition of the
metal chelates, pure radioactive chelate solutions were
prepared. These were injected into their respective columns
(see Table 4) in a series of 5 pl injections, 1 min apart.
The series of injections totaled 50 pl and 100 pl. After the
carrier gas had been allowed to flow after the last injection
for 3 times the retention time of the chelate, the system
was dismantled and the components counted. The packing was
counted in a 60 x 15 mm Pyrex petri dish.


3. Counting System


All radioactivity measurements were made using two
3 x 3 inch NaI(TI) detectors connected to the multichannel
pulse height analyzer system previously described. A block
diagram of the counting system is~shown in Figure 7. The
three different counting geometries shown in Figure 8 were
studied and they are described in detail elsewhere (34).
The exponential dilution flask exhibited sever adsorption
properties and was difficult to heat uniformly, and conse-
quently, will not be considered here.

The spiral coil flow-through geometry consisted of a
57 inch piece of 1/8 inch O.D. #316 stainless steel tubing
with a 0.005 inch wall (Superior Tube Company). It was
wound in a spiral 2 7/8 inch O.D. x 1 1/8 inch I.D. The
inlet and outlet arms each consisted of 6 1/2 inches of
tubing. The charcoal traps were made from 5 inch pieces of
tubing packed with 1 inch of 40-50 mesh charcoal. The
packing began at the center and extended toward the outlet.
Some of the charcoal traps were made from the same stock as
the spiral. Others were 1/8 inch O.D. Teflon lined aluminum
tubing or 3 mm O.D. Pyrex.
















































Figure 7. Block diagram of the NaI(TI) detector and counting system used for
radiochromatography
















-300 v.D.C.


GAS


CHROMATOGRAPH


INJECTION
-PORT


COLUMN


HE
:ARRIER
GAS


IIIZZ]IIZIIZIIZ
JIZEIIZ IIZ


LIQ N2 TRAP


Figure 8. Counting geometries studies for measuring the radiochromatographic
effluent activity


--ELECTROMETER R-ECORDER










The reversible counting geometry consisted of a 6 inch
piece of Teflon tubing packed with 2.5 inches of 20% SE-30
on Chromosorb P, (AW DMCS). 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 geometry. A He purge line was connected
upstream to help flush the counting geometry.

The sample counting in geometries connected to the
system was done in a cave beside the gas chromatograph. A
heated Teflon tube passed through the oven wall into the cave
The cave had 8 inches of lead between the detectors and the
gas chromatograph, 4 inches of lead on the side, top, and
back, and no lead on the bottom and front. The inside
dimensions were 24 x 20 x 8 inches.

Sample counting, where the traps were removed from
the gas chromatograph system, was done in a lead cave 40 x
22 x 12 inches with 4 inches of lead on 5 sides and 2 inches
on the bottom.

The response of the NaI(TI) detector as a function of
position was obtained by passing a 0.2 inch long by 0.061
inch diameter piece of molecular sieve containing Cs137
across the surface of the can containing the detector. Two
perpendicular passes were made. The time sequence store
mode of the analyzer was used to obtain a plot of relative
response versus position.

Three different modes of counting radioactive metal
chelates were used. Where two or more radionuclides were
present, a pulse height spectrum was taken and the areas
under the photopeaks integrated using a PDP-8/L laboratory
computer (Digital Equipment Corp.). When only one radio-
nuclide was present, a single channel analyzer was used and
the entire spectrum above a few key was measured. When
radiochromatograms were made, the time sequence store mode
of the analyzer was used.

The flow through spiral counting geometry was used
at 1700 C with a continuous He carrier gas flow of 30 ml/
min to obtain radiochromatograms of Cu(hfa)2 Cr(hfa)3
mixtures using column 2. The time sequence store mode of
the analyzer was used with a dwell time of 0.02 min/channel.

The stop-flow counting geometry was operated in a
flow-through mode at a constant temperature of 115 C with
a sorbent identical to column 13 inside. The geometry was
held at 600 C to trap effluents, such as the Cu(hfa)2, and
then heated to 1500 C to desorb the complex. The procedure
for repetitive use of the reversible stop-flow counting
system was:










1. Start carrier gas, apply cool air to geometry
2. Inject sample
3. Wait 2.5 3.0 min, stop carrier gas flow
4. Count sample 10 min
5. Heat geometry and allow purge gas to flow for
15 min
6. Cool geometry, stop purge gas, and make a 10 min
background count
7. Restart carrier gas and trap the next component.

Irreversible charcoal traps were used to obtain inte-
gral radiochromatograms. The metallic traps were cooled
with dry ice, but the 3 mm Pyrex traps required no cooling.
The integral chromatograms were obtained by stopping the
carrier gas flow and counting the accumulated activity in
the trap. Cu(hfa)2 Cr(hfa)3mixtures were injected as
solids and counted using this technique.


4. Chelation Studies


The '(tfa) ,H(hfa)-, and -H(fod- ligands (Peninsular Chem.
Research) were all redistilled prior to use. For H(tfa),
the 1070 C fraction was collected. The Cu(hfa)2 and
Cr(hfa)3 were used as purchased (Pierce Chemical Company).
The Mallinckrodt Nanograde benzene was dried over a 13X
molecular sieve. The Vitreoseal quartz for the ampules
was obtained from the Thermal American Fused Quartz Company.

Crs5C1l in 1 N HC1, 0.02 mg/ml, 3.7 mCi/ml (Union
Carbide), Fe59C13 in 1 N HC1, 2 mCi in 0.83 ml (New England
Nuclear Corp.) and Be7C12 in 0.5 N HC1, 2 mCi in 0.5 ml
(New England Nuclear) were used as tracers in these experi-
ments.

The atmospheric sublimation apparatus consisted of an
8 cm x 6 mm O.D. Pyrex tube bent into a "U" shape with a con-
necting tube on each end. The apparatus was heated in the
gas chromatographic oven. The He flow through the tube and
the heat applied by the oven were sufficient to sublime the
chelates and move them from the bottom of the "U" through a
short piece of Teflon tubing to a 3 mm O.D. Pyrex trap
outside the oven. This trap was maintained at room tempera-
ture.


The vacuum sublimation apparatus was a standard
design with a 15 mm O.D. Pyrex pot and an 8 mm O.D. cold
finger. It was connected to a mechanical forepump.










Mixtures of 2 mg of solid Cr(hfa)3 and 1 mg of solid
Cu(hfa)2 were sealed in 2 mm x 2 cm quartz tubes and
irradiated for 4 hours at 1012 neutrons/cm2-sec in the
University of Florida Training Reactor (UFTR). Solid 1 mg
samples of Mn(hfa)2 were sealed and irradiated under the
same conditions for 15 min. The Cr(hfa)3 and Cu(hfa)2
mixture was injected into columns 2 and 11 (see Table 4),
and the Mn(hfa)2 was injected into column 9. No annealing
treatment was used on either sample. The chromatographic
column was then dismantled and each component counted.

Cr(tfa)3, Cr(fod)3, and Cr(hfa)3 were dissolved in
Nanograde benzene, irradiated for 14 hours at 6.5 x 1011
neutrons/cm2-sec in the Georgia Technology Research Reactor
(GTRR), and injected without further treatment into columns
5, 7, and 3, respectively. Each of these columns, charcoal
traps, and injection port liners were similarly counted.

Thermal annealing of irradiation damage in Cr(tfa)3
was studied by irradiating two benzene solutions: one
containing 30.2 mg Cr(tfa)3 and 2.7 mg H(tfa) (10% excess)
in 1 ml of benzene, and the other contained 30.4 mg Cr(tfa)3
and 27 mg H(tfa) (100% excess) in 1 ml of benzene for 14
hours at 6.20 x 1011 neutrons/cm2-sec in the GTRR. Part of
the sample with 10% excess ligand was injected into column
5 without treatment. The rest was annealed (heated) 9 hours
at 100 C and then injected. The sample with 100% excess
ligand was annealed 9 hours at 1000 C in its sealed irradia-
tion vial and injected. All three columns were dismantled
and counted.

Hot Cr(tfa)3 was prepared by synthesizing the chelate
using Cr spiked with Cr as the starting material. Forty
mg CrC13. 6 H20 was spiked and dissolved in 1.5 ml H20
in a 5 ml round bottom flask. Three hundred mg of urea and
86 mg (10% excess) of H(tfa) were added, and the mixture
refluxed at 100 1150 C for 7 hours. The insoluble product
was filtered, washed with H20, and air dried. The product
was dissolved in ether and placed in the atmospheric sublima-
tion apparatus. After the ether had evaporated, the tempera-
ture was raised to 1500 C for 15 min. Forty-seven mg of
purple product was recovered from the trap under a flow rate
of 10 ml/min.

Radioactive Cr(fod)3 was also prepared by synthesizing
the chelate using chromium spiked with Cr51 as the starting
material. Fifty-two and one-half mg Cr(N03)3 9 H20 was
spiked and dissolved in 0.5 ml absolute ethanol in a 10 ml
round bottom flask. One hundred thirty-four and two tenths
mg (10% excess) H(fod) dissolved in 1 ml absolute ethanol










was added. After heating for 1 hour at 800 C, the mixture
was transferred to the atmospheric sublimation apparatus.
Heating for 1 hour at 1400 C with 10 ml/min He flow
transferred 6.5 mg chelate to the trap.

A solution containing 43.5 mg Lu(fod)3 and 42.1 mg
H(fod) in 1.5 ml benzene was prepared. One ml of this was
irradiated for 14 hours at 5 x 1011 neutrons/cm2-sec in the
GTRR. It was annealed 10 hours at 1000 C. The sample was
vacuum sublimed at 1150 C for 15 min. Thirty-three and one-
half mg pale yellow chelate was obtained.

The same procedure was applied to Gd(fod)3, except
that the sublimation was for 45 min at 1500 C.

Radioactive Fe(fod)3 was synthesized from FeC13
spiked with Fe59 dissolved in ethanol. H(fod) was added
and immediately after mixing, the deep red chelate was
formed. It was extracted into 2.5 ml hexane and sublimed
at atmospheric pressure at 1400 C. Six mg of deep red
chelate was obtained. The pale red crystals were discarded.

Radioactive Be(tfa)2 was synthesized from BeSO4
4 H20 spiked with Be7, Forty-one mg BeSO4 4 H20 and 90.0
mg NaAc were dissolved in 1.5 ml H20. One hundred mg (50%
excess) H(tfa) dissolved in absolute ethanol was added.
The white product was extracted into benzene and vacuum
sublimed at 650 for 15 min. Thirty-four and a half mg chelate
was obtained.

The Cu(tfa)2 was synthesized according to the method
of Berg and Truemper (84). A saturated benzene solution
with excess solid chelate, 65.5 mg total weight, was
irradiated for 7.5 hours at 1012 neutrons/cm -sec in the
UFTR. The sample was annealed at 1150 C for 7 hours and
vacuum sublimed at 1150 C.


5. Results and Discussion

Both solid and liquid samples were evaluated in this
study. The solid sampler's large sample size (1-2 mg) was
ideal from the standpoint of counting, especially when
using the flow-through type geometries. However, the
activity measurements made on the solid sampler and quartz
fragments in the injection port showed large amounts of
radioactive residue, most of which was due to Szilard-
Chalmers decomposition. There were visible chelate residues
in the sampler barrel and among the quartz fragments,
indicating either that the injection port was not hot enough,
or that the solid sampling technique needed to be improved.
This contributed to the injection port activity to a small
extent.










The length of the sealed sample tubes which were
broken by the solid sampler was found to be critical. If
the tubes were too long, large amounts of glass (or quartz)
fragments prevented all the sample from leaving the sampler
barrel. If the sample tubes were too short, they were not
broken at all, and were forced upward,.breaking the injec-
tion port liner which surrounded the solid sampler barrel.
For 0.060 inch O.D. melting point capillaries, the best
lengths were found to be between 0.406 inches and 0.469
inches. For liquid injections of 5 pl or less, the total
sample activity was so low that.long counting times were
required. When a total of 100 Ip of Cu(tfa)2, injected
5 pl at a time (in order not to overload the column) 7 x 10"
cpm were measured in the trap. If all factors affecting
recovery are linear, a 1 pl injection should have at least
7 x 104 cpm activity which is roughly 70 times background.

The results of the injected sample size study are
shown in Figure 9. The benzene Cr(tfa)3 chromatograms gave
one peak for benzene and two chelate peaks. The second
chelate peak area was only 10% of the first and it was
necessary to integrate both chelate peaks together. From
these results it is seen that the 10 pl sample deviates
significantly from linearity and is therefore considered as
the upper limit of liquid sample injections.

In terms of eluting the chelates quantitatively, the
silanization did not produce any measurable improvement.
Even after 70 pl of DMCS has been injected, the chelate
peak areas were the same as those observed before any DMCS
had been injected.

For all of the solid injections the elution efficiency
was poor. Eighty ninty% of the activity was found in the
injection port, the solid sampler, and the column inlet.
Both the quartz fragments and the acetone rinse were radio-
active. Most of the activity was removed from the injection
port by this procedure. The only time any radioactive chelate
was measured was when the sample activity was so high that
the 5% which was eluted was hot enough to be counted.

The color and appearance of the solid Cr(hfa)3,
Cu(hfa)2, and Mn(hfa)2 did not change during irradiation.
Since only 2 atoms in 109 of the chelate undergo the (n,y)
reaction during a 15 hour irradiation at 1012 neutrons/cm2-
sec, only a small fraction of the sample is effected by the
Szilard-Chalmers reaction. Thus, even though these solids
may have been essentially unaffected by irradiation, the
chelation of the radioactive atoms was by no means quantita-
tive. This was shown by the results of the gas chromatograph
residue measurements. When solid Cr(hfa)3 and Cu(hfa)2 were

































-4-


I I I I I I I1


~a 2.-


Figure 9. Effect of sample size on the quantitative elution of Cr(tfa) ~ on Column 6


10


<



01
' -
W


W
Ld
w.


S0


0
0



0
cr


_ __ _I _____ I


:;IZ-, 3;8










injected into columns 2 and 11, respectively, large amounts
of activity were found in the solid sampler, the injection
port, and the head of the column. Smaller amounts of
activity were found at the middle of the column and at the
outlet. When a charcoal counting trap was used for the
above experiment, approximately 50% of the radioactive
sample was found in the charcoal trap. Undoubtedly, there
was some annealing in the reactor and/or the injection port.
After the irradiated solid Mn(hfa)2 was injected, no trace
of Mn6" was found beyond the injection port. There was
apparently insufficient annealing to rechelate any Mn56

The Cr(tfa)3, Cr(fod)3, and Cr(hfa)3 solutions
behaved the same way as the solids. Very small amounts of
activity were found in the counting geometry and large
amounts stayed in the injection port.

The use of a Teflon injection port liner showed no
improvement over quartz. The injection port retention was
not due to the liner material, since hot chelate passed
through the rest of the Teflon tube satisfactorily.

The results of the Cr(tfa)3 annealing experiment are
shown in Table 5. The residue study was used to evaluate
the purity of the radioactive chelate. The figures from
an untreated sample similar to this one are included for
comparison. It can be seen that heating at 1000 C for 9
hours in conjunction with irradiation with a 100% excess
amount of ligand is quite effective in minimizing the amount
of unchelated radioactive metal. When only 10% excess
ligand was used the results were much better than when no
treatment was used. Still, the results were poor. Heating
caused some improvement, but only with 100% excess ligand
were the results optimum.

The solutions which were irradiated and not annealed
were equally unsatisfactory. Here, there was no significant
solid residue in the injection port as most of the metal
chelate was eluted through the column. Only the small
number of radioactive nucleii apparently underwent radiolysis
and were consequently not in a form suitable for chromato-
graphic elution. Only after annealing can irradiated
chelate solutions be used for gas chromatography (see Table
5).

The solution of Cr(tfa)3 which was irradiated with
100% excess ligand and then annealed gave excellent chromato-
graphic results. The sublimation step used in the residue
study for purifying the irradiated mixture is apparently
not necessary for efficient elution. Table 6 lists the
results of the residue studies of the chelates.













TABLE 5

EFFECT OF POST-IRRADIATION ANNEALING
OF Cr(tfa)3 IN BENZENE


Annealing
Excess Ligand

Injection port
liner

Column packing

Column tubing

Charcoal trap


Activity
None None
0% 10%


89.7 11.0

3.8 35.8

0.25

7.0 52.9


Recovered, %
9 hrs, 100C
10%


2.80


3.0

0.7


93.5


of Total
9 hrs, 100C
100%


0.88

1.43

0.38

97.30


NOTE: Chromatographic analysis run on Column 5 (Table 4).












TABLE 6


PERCENT RECOVERY OF METAL CHELATES
IN CHROMATOGRAPHIC SYSTEM


Chelate

Cr(tfa) 3

Cr(tfa)3a

Cr(tfa) 3b

Cr(tfa) 3

Cr(fod) 3

Be (tfa) 2

Fe(fod) 3

Cu(tfa) 2

Gd(fod) 3

Lu(fod) 3


Column
No.

5

5

5

5

7

1

10

12

14

15


Residual
Injection
Port Liner

0.35

11.06

2.80

0.88

0.35

0.07

0.30

1.30

1.0

0.51


Activity,
Column
Packing

1.36

35.81

2.97

1.43

15.7

2.7

23.0

23.0

35

7.0


% of Total
Teflon
Tubing Trap

0.38 98.16

0.25 52.87

0.70 93.50

0.38 97.30

0.75 83.0

3.0 94.2

2.48 74.3

3.5 72.2

1.3 63

0.43 92.1


aIrradiated
bIrradiated

CIrradiated


with

with

with


10% excess ligand

10% excess ligand and annealed

100% excess ligand and annealed










The Cr(tfa)3 results were quite good. The poorer
Cr(fod)3 results may have been due to a higher column
temperature which would have contributed to decomposition
on the column packing, since a large portion of the
activity was found there.

The Be(tfa)z results were similar to those of the
Cr(tfa)3. The third column, however, appeared to hold a
large portion of the activity. There was no explanation
for this. It is not a problem with sample purity because
the injection port did not contain any more activity than
it did in the other two runs.

The Gd(fod)3 data was less accurate than the other
experiments because the sample had decayed through 3 half-
lives considerably before it was counted. This meant that
a mixture of daughters and other Gd isotopes were present,
the half-life of which was unknown, and therefore no decay
correction could be made. The results are accurate to the
extent that they show nearly 1/3 of the total activity on
the column packing. This is different from the Lu(fod)3
result, even though they are both rare earths and both
experiments-used identical columns. -This is similar to-the
Cr(tfa)3 Cr(fod)3 results.

The Fe(fod)3 also left a large residue on the packing
material. Some was also present on the column tubing, which
would indicate that the chelate and liquid phase did interact
and that accurate quantitative results will be difficult to
realize. The Cu(tfa)2 behaved similarly. The larger
amounts of residue in the column is due to the low operating
temperature for that particular experiment.

In summary, the Cr(tfa)3, Be(tfa)2, and Lu(fod)3 with
the columns which were used with them seem to be good choices
for gas chromatography of radioactive metals. The rest of
the chelates seem to be restricted by the nature of the
column packing. A better choice of chromatographic condi-
tions, then, will probably make activation analysis work
with these chelates feasible also. Further, it is seen that
reaction irradiation of the metal chelates requires post-
irradiation annealing of both solid and liquid samples.

The relative response of the Na(Tl) crystal as a
function of sample position is an important consideration
in the design of counting geometries. Consecutive samples
need to be held in regions of the same response if repro-
ducible results are to be obtained. Figure 10 shows the
plot of response versus position for the 3 x 3 inch NaI
detectors used in this study. The region where the response
decreased by no more than 10% from the maximum was found












100



90- qL


w 80-
z

( 70-


> 60-


dw OUTSIDE RADIUS
: 50-
OF CRWS7-.L '-"


40


30-
I I I I I I

S 5 10 15 20 25 15 35 40 45 0

DlST A F. 'T F Ti OF C--'' : M.
Figure 10. Response of the NaI(Tl) detector as 4 function of position on the crystal
face. Points denoted by circles represent the response of the facing side of the
detector face






Pages
Missing
or
Unavailable










0
O


-J


I~~~~U~ I


5.0

4.5


4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0


1*
h -!


~I .;


Figure 11. Cu(hfa)2 activity counted in a stainless steel flow-through counting geometry, after
sepn r;tion from a (u(h Fa) 2: -Cr(hf-a) 3 mixture


I I I I I I I I I I I


-


-l


5 0
o .


c, .,,.


T .--- -s '


"I I ,g < ... .
,i; fr ^ .; I ": ,' ,











tL 10 min.


3.9 Kev/Chan.


Cr51

0.322 Mev.






10 K


Cu64
0.511 Mev.


c 64
Cu6
1.34 Mov.


PHOTON
Figure 12. Multichannel pulse-height
Cr f followed by the elution of (ICu


ENERGY, Mov.
spectrum taken after the gas chromato aphic separation of


* ~1~









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 flexibi-
lity, indefinite counting time, freedom from critical stop
flow timing, and complete reversibility.

Figure 13 shows a 400 channel radiochromatogram which
illustrates the reversibility of the geometry. A solid
Cu(hfa)2 sample was injected into column 13. The activity
suddenly rises as the radioactive plug enters the geometry.
Since flow is not stopped the tail of the chromatographic
peak accumulates and the count rate slowly increases. At
channel 364, heat was applied to the 600 geometry and the
trapped chelate started to move. 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.







I--~-~--~ I I~ I~~p~~~m9""" N8m(8a I N N N--


ro

x

-J





0
0.





CO
Th


I 0


TI g` 2 '~:


)% r


- 4


~T ~7I iI 9


Figure 13. IRdiochromatogramr illustrating the county ng of Cul(hfa) 2
counting geometry


in the reversible


I I,


5.0


4.5

4-.0


3.0


2.5


2.0


1.5


1.5


0.5


~Gm .~..r:(I~o~"\r~a~,r1b; b
-9~~~- ~: b
a
.


0 ~


tsfss _
' ..












ACKNOWLEDGEMENTS


The diligence and dedication of the members of this
research group during the period of this grant has made
this work possible. They are the ones who are responsible
for the productivity of this project. Particularly, the
efforts of Dr. R. W. McCoy, Mr, F. T. Varcoe, Mr. T. R.
Booher and Miss J. Durham should be recognized. The
collaborative efforts and stimulating discussions of the
entire group are sincerely appreciated.












LITERATURE CITED


1. Fishman, M. J., Robinson, B. P., and Midgett, M. R.,
Anal. Chem., 39, 261R (1967).

2. Okun, D. A., et al., J. Water Pollution Control Feder-
ation, 37, 587 (1965).

3. Christman, R. F., et al., J. Water Pollution Control
Federation, 38, 685 (1966).

4. Kitano, Y., and Nasu, Y., Bunsek, Kagaku, Shinpo Sosetsu,
1964, 140R, C.A., 63, 12873c (1965).

5. Monnier, D., Arch. Sci. (Geneva), 18, 273 (1965). C.A.,
64, 4253b (1966).

6. Landstrom, 0., and Wenner, C. G., Aktiebolaget Atomenergi,
Stockholm AE-204 (1965). C.A., 12361g (1966).

7. Yule, H. P., Anal. Chem., 38, 818 (1966).

8. Leddicotte, G. W., and Moeller, D. W., U.S. At. Energy
Comm., CF-61-5-118 (1961).

9. Ross, W. J., Anal. Chem., 36, 1114 (1964).

10. Albert, P., Pure Appl. Chem., 1, 111 (1960).

11. Brownlee, J. L., U.S. AEC Rept. TID-6311 (1960).

12. Kusaka, Y., and Meinke, W. W., Natl. Acad. Sci., Natl.
Res. Council, Nucl. Sci. Ser. NAS-NS-3104 (1961).

13. Ruch, R. R., Tera, F., and Morrison, G. H., Anal. Chem.,
37, 1565 (1965).

14. Brune, D.,and Landstroem, 0., Radiochim. Acta, 5, 228
(1966).

15. Smales, A. A., Mapper, D., and Fouche, K. F., Geochim.
Cosmochim. Acta, 31, 673 (1967).

16. Thompson, B. A., and LaFleur, P. D., Anal. Chem., 41,
852 (1969).

17. Hahn, K. J., Tuma, D. J., and Quaife, M. A., Anal. Chem.,
39, 1169 (1967).










18. Hahn, K. J., Tuma, D. J., and Sullivan, J. L., Anal.
Chem., 40, 974 (1968).

19. Girardi, F.,and Sabbioni, E., J. Radioanal. Chem., 1,
169 (1968).

20. Shipman, W. H., Anal. Chem., 38, 1175 (1966).

21. Bigliocca, C., Girardi, F., Pauly, J., Sabbioni, E.,
Meloni, S., and Provasoli, A., Anal. Chem., 39,
1634 (1967).

22. Patek, P.,and Sorantin, H., Anal. Chem., 39, 1458
.(1967).

23. Houtmann, J. P. W., Proc. Intern. Conf. on Activation
Techniques in the Life Sciences, IAEA, Vienna, p.
171 (1967).

24. Bark, L. S., Duncan, G., and Graham, R. J. T., Analyst,
92, 347 (1967).

25. Tera, F., and Morrison, G; H., Anal. Chem.,- 38, 959
(1966).

26. Eisner, U., Rottschafer, J. M., Berlandi, F. J., and
Mark, H. B., Anal. Chem., 39, 1466 (1967).

27. Tang, C. W., and Maletskos, C. J., Science, 167, 52
(1970).

28. Samsahl, K., Wester, P. 0., and Landstrom, 0., Anal.
Chem., 40, 181 (1968).

29. Aubouin, G., Diebolt, J., Junod, E., and Laverlochere,
J., Proc. Intern. Conf. Modern Trends in Activation
Analysis, College Station, Texas, 1965.

30. Girardi, F., Merlini, M., Pauly, J., and Pietra, R.,
Proc. Symp. Radiochemical Methods of Analysis, Vol.
2, Salzburg, 1964, p. 3.

31. Moiseev, V. V., Kusnetsov, R. A., and Kalinin, A. J.,
Proc. Intern. Conf. Modern Trends in Activation
Analysis, College Station, Texas, 1965.

32. Morrison, G. H., Gerard, J. T., Travesi, A., Aurie,
R. L., Peterson, S. F., and Potter, N. M., Anal.
Chem., 41, 1633 (1969).










33. Morrison, G. H., Gerard, J. T., Kashuba, A. T.,
Gangadharm, E. V., Rothenberg, A. M., Potter, N. M.,
and Miller, G. B., Science, 167, 505 (1970).

34. Cram, S. P., Advances in Chromatography, Vol. 9, J.C.
Giddings and R.A. Keller, eds., Marcel Dekker,
New York, 1970, p. 243.

35. Cram, S. P., and Brownlee, J. L., J. Gas Chromatog.,
5, 353 ( 67).

36. Cram, S. P., and Brownlee, J. L., J. Gas Chromatog.,
6, 305 (1968).

37. Cram, S. P., and Brownlee, J. L., J. Gas Chromatog.,
6, 313 (1968).

38. Sievers, R. E., and Moshier, R. W., Gas Chromatography
of Metal Chelates, Pergamon Press, New York, 1965.

39. Ross, W. D., and Sievers, R. E., Talanta, 15, 87 (1968).

40. Ross, W. D., Sievers, R. E., and Whe-eler, G., Anal.
Chem., 37, 598 (1965).

41. Ross, W. D., and Sievers, R. E., Anal. Chem. 41, 1109
(1969).

42. Scribner, W. G., Smith, B. H., Moshier, R. W., and
Sievers, R. E., J. Org. Chem., 35, 1696 (1970).

43. Booher, T. R., M.S. Thesis, University of Florida,
Gainesville, Florida, 1969.

44. Shankar, J., India At. Energy Comm., BARC-348, Bhabha
At. Resch. Centre, 1968.

45. Lin, T. K., and Yeh, S. J., J. Nucl. Sci. Technol.
(Tokyo), 3, 289 (1966).

46. Torko, J., Energ. Atomtech, 20, 200 (1967).

47. Nath, A., and Nesmeyanov, Radiokhimiya, 5, 125 (1963).

48. Lazzarini, E., J. Inorg. Nucl. Chem., 29, 7 (1967).

49. Gainer, I., and Ponta, A., Rev. Roum. Phys., 13, 645
(1968).

50. Gainer, I., and Ponta, A., Rev. Roum. Phys., 13, 887
(1968).










51. Machado, J., Machado, R. M., Vargas, J., Chem.
Effects of Nucl. Transformations, Proc. Symp.,
Vienna, 1964, 2, 195 (1965).

52. Thomas, V. G., Nath, A., and Shankar, J., USAEC
Tech. Rept., NP-12330, p. 415 (1962).

53. Thomas, V. G., Nath, A., and Shankar, J., USAEC Tech.
Rept., NP-12330, p. 407 (1962).

54. Shankar, J. Venkateswarlu, K. S., and Lol, M., J.
Inorg. :',.. Chem. 28,11 (1966).

55. Shankar, J., Nath, A., and Thomas, V. G., J. Inorg.
-Nucl. Chem., 30, 1361 (1968).

56. Thomas, V. G., Indian J. Chem., 1, 247 (1969).

57. Rao, K. A., and Nath, A., Radiochimica Acta, 5, 162
(1966).

58. Gainer, I., and Ponta, A., Rev. Roum. Chim., 13,
401 (19681.

59. Sievers, R. E., Moshier, R. W., and Morris, M. L.,
Inorg. Chem., 1, 966 (1962).

60. Sievers, R. E., Ponder, B. W., Moshier, R. W., and
Morris, M. L., Inorg. Chem., 2, 693 (1963).

61. Ross, W. D., Anal. Chem. 35, 1596 (1963).

62. Ross, W. D., and Wheeler, G., Anal. Chem., 36, 266
(1964).

63. Hill, R. D., and Gesser, H., J. Gas Chromatog., 1,
10 (1963).

64. Sievers, R. E., Connolly, J. W., and Ross, W. D., J.
Gas Chromatog., 5, 241 (1967).

65. Baetzer, A. M., Damron, C., and Budaez, V., A.M.A.
Arch. Indust. Health, 20, 136 (1959).

66. Brieger, H., Ztschr, F. Exper. Path. V. Therap., 21,
393 (1920).

67. Goldman, M., and Karotkin, R. H., Am. J.M.Sc., 189,
400 (1935) .

68. Major, R. H., Bull. John Hopkins Hosp., 33, 56 (1922).










69. Mancuso, T. F., Industrial Med. and Surgery, 20,
393 (1951).

70. Morris, G. E., A.M.A. Archivis of Dermatology, 78,
612 (1958).

71. Pascale, L. R., Waldstein, S. S., Engbring, G.,
Dubin, A., and Szanto, P. B., J.A.M.A., 149, 1385
(1952).

72. Curran, G. L., J. Biol. Chem., 210, 765 (1954).

73. Mertz, W., Roginski, E. E., and Reba, R. C., Amer.
J. Physiol., 209, 489 (1965).

74. Mertz, W., and Schwarz, K., Measurements of Exocrine
and Endocrine Functions ofth-e Pancreas, J. B.
Lippincott Company, Philadelphia, Penn., pp. 123-
237 (1961).

75. Schroeder, H. A., J. Nutr., 88, 439 (1966).

76. Schroeder, H. A., Balassa, J. J., and Tipton, I. H.,
J. Chron. Dis., 15, 941 (1962).

77. Schwarz, K., and Mertz, W., Arch. Biochem. Biophys.,
85, 292 (1959).

78. Glinsmann, W. H., Feldman, G. J., and Mertz, W.,
Science, 152, 1243 (1966).

79. Glinsmann, W. H., and Mertz, W., Metabolism, 15,
510 (1966).

80. Levine, R. A., Doisy, R. J., and Streeten, D. H. P.,
Diabetes, 15, 539 (1966).

81. Savory, J., Mushak, P., and Sunderman, F. W., J.
Chromatog. Sci., 7, 674 (1969).

82. Springer, C. S., Meek, D. W., and Sievers, R. E.
Inorg. Chem., 6, 1105 (1967).

83. Tanaka, N., Shono, T., and Shinra, K., Nippon Kagaku
Zasshi, 89, 669 (1968).

84. Berg, E. W., and Truemper, J. T., J. Phys. Chem.,
64, 482 (1960).




Full Text

PAGE 2

THE; RADIOCHROMATOGRAPHIC ANALYSIS OF FRESH WATER RESOURCES by STUART P. CRAM (Principal Investigator) PUBLICATION NO. 15 FLORIDA WATER RESOURCES RESEARCH CENTER RESEARCH PROJECT TECHNICAL COMPLETION REPORT OWRR Project; Number A-012-FLA Annual Allotment Agreement Numbers 14-01-0001-1628 (1969) 14.31 (1970) 14-31-0001-3209 (1971) Report Submitteq: October 28, 1971 The work upon which this report is based was supported in part by funds provided by the United States Department of the Interior, Office of Water Resources Research as Authorized under the Water Resources Research Act of 1964.

PAGE 3

TABLE OF CONTENTS Page ABSTRACT ---------------------------------------------iii PROJECT PUBLICATIONS ---------------------------------iv INTRODUCTION -----------------------------------------1 PART I. CHARACTERIZATION OF THE IRRADIATION STABILITY OF CHROMIUM BETA-DIKETONATES FOR GAS CHROMATOGRAPHIC SEPARATIONS IN NEUTRON ACTIVATION ANALYSIS -8 1. Experimental -----------------------------8 2. Results and Discussion -------------------13 PART II. THE QUANTITATIVE GAS CHROMATOGRAPHIC SEPARATION AND ANALYSIS OF RADIOACTIVE, VOLATILE METAL BETA-DIKETONATES -----------------------------21 -----------------------------21 2. Quantitative Column Elution --------------22 3. Counting System -------------------------25 4. Chelation Studies -----------------------29 5. Results and Discussion ------------------31 ACKNOWLEDGEMENTS ------------------------------------44 LITERATURE CITED ------------------------------------45 ii

PAGE 4

ABSTRACT The study of radiochromatographic separations for the neutron activation analysis of trace level metals in fresh water sources is reported. Chromatographic separations of metal beta-diketonates were developed for post-irradiation --separatToiis 0 Wlie-ll-the {race metars-were complexed before irradiation, radiation degradation was found to be a function of the solution concentration, presence of excess ligand, irradiation time, and the neutron flux spectrum. Trifluoroacetylacetone and heptafluorodimethyloctanedione complexes of chromium were quantitatively recovered after irradiation. Quantitative elution from the chromatographic column of the beta-diketonate complexes of Cr, Mn, Fe, Be, Lu, Gd, and Cu were studied and found to yield recoveries between 52 and 98%. Extensive studies are reported which describe the optimum conditions for separation and account for decomposition and adsorption losses in the system. The development of sampling systems, counting geometries, and sample trans fer l-iues are reported which must be carefully considered when analyzing mUlticomponent metal chelate mixtures. Cram, Stuart P. THE RADIOCHROMATOGRAPHIC ANALYSIS OF FRESH WATER RESOURCES Completion Report to the Office of Water Resources Research, Department of the Interior, October, 1971, Washington, D.C. 20240 KEYWORDS: neutron activation*/ gas chromatography*/ activation-chromatography coupling*/ fresh water/ trace analysis. iii

PAGE 5

PROJECT PUBLICATIONS 1. Cram, S. P., and Varcoe, F. T., "Gas Chromatographic Separations in Neutron Activation Analysis," Proc. Int'l. Symp. on Modern Trends in Act. Anal., __________________ ________________________________________ 2. Wade, R. L., and Cram, S. P., "Quantitative Interpretation of Semi10garithmic Gas Chromatographic Data," Anal. Chern., 893 (1969). 3. Glenn, T. H., and Cram, S. P., "A Digital Logic System for the Evaluation of Instrumental Contributions to Chromatographic Band Broadening, J. Chromatog. Sci., 46 (1970). 4. Juvet, R. S., and Cram, S. P., "Gas Chromatography," Anal. Chern., lR (1970). 5. Booher, T. R., and Cram, S. P., "Characterization of --------------------the Stab44ity o Some Diketonates for Gas Chromatographic Separations in Neutron Activation Analysis," J. Radioana1. Chern., submitted for publication. 6. McCoy, R. W., and Cram, S. P., "Extention of the Time Normalization Theory of Gas Chromatography for the Minimization of Analysis Time," Anal. Chern., submitted for publication. 7. McCoy, R. W., and Cram, S. P., "High Speed Gas Chromatographic Separations of Volatile Metal Beta-Diketonates," Anal. Chern., submitted for publication. 8. Cottrell, D. B., and Cram, S. P., "A High Precision Digital Data Acquisition System for Neutron Activation Analysis," in preparation. 9. Boerner, B. R., and Cram, S. P.', "Absolute Neutron Activation Analysis," in preparation. iv

PAGE 6

INTRODUCTION Research in the field of water resources and water quality is highly dependent upon the development of new analytical techniques which are specific, sensitive, reproducible, rapid, and suitable for automation. Because of the shortage of analytical chemists in this country, and particularly analytical chemists working in water quality research, new analytical methods of analysis which incorporate the above criteria have not been designed and developed to solve many of the existing problems in this field. The work dis.cussed in this report is directed toward the development of two such analytical techniques, i.e., neutron activation analysis and gas-liquid chromatography, and the coupling of the two complimentary techniques for u1tratrace metal analysis in aqueous systems. Numerous instrumental methods for elemental analyses in water samples have been reported (1-5), although the number of methods applicable to trace analysis (defined here less than 1 of any is ably more restrictive. The methods of elemental analyses reported to date have generally been very limited in scope, and the breadth of these techniques at the trace concentra-tion level in water analysis has been limited by a number of considerations. First, the wide diversity of matrix materials represents a limitation on the sensitivity of any chemical analysis, as the relationship between the chemical and physical interactions are affected by the environment of the element to be determined. Further, most techniques are subject to interferences by the presence of other chemicals. The sensitivity of these methods is therefore dependent upon the chemical nature of the sample and varies from one sample to the next. This effect is particularly pronounced at trace concentrations. Neutron activation has likewise been investigated as an analytical tool for the element analysis of aqueous samples (6-8). These investigations have been summarized in Table 1 to illustrate the general applicability of the technique at the trace level. The sensitivities reported do not represent optimum experimental conditions, and indicate the sensitivities only for irradiation times of one hour at a thermal neutron flux of 4.3 x 1012 neutrons/ cm2 -sec. 1

PAGE 7

TABLE 1 SENSITIVITY AND APPLICABILITY OF THERMAL NEUTRON ACTIVATION FOR ELEMENTAL ANALYSES Minimum Detection Limi t Ranges,. Jlgr 10-5 10-4 10-4 10-3 10-1 10 10 101 101 102 102 103 Mn, Ar, Ho, Na, Rh, Nd, Cl, Pt, Tb F, Ca, Si, 0, 2 In, Sc, Er AI, Ag, Yb, K, Hg, Mg, Os S Fe I, Au, Eu, Dy V, Br, Cs, Hf, Co, Cu, As, Se, Cd, Sn, Sb, Ba, Lu Ti, Zn, Ge, Mo, Ce, Ru, Rb, Zr, Cr, Ni, Nb, Pb, Ir, Sm, Sr, Ga, W, La, Pd, Ta, Te, Pr, Y, Tm

PAGE 8

Thus it can be seen that neutron activation offers the advantages of dynamic applicability and high sensitivity. Further, the sensitivity of the method is invariant with the nature, size and concentration of the sample. The sensitivity is determined by the nuclear parameters of the element, such as the half-life, cross-section, neutron flux, etc., and are constant regardless of the chemical system being investigated. The inherent sensitivity and precision of neutron activation analysis at the submicrogram concentration level is dependent upon the accuracy with which the activated, gamma emitting radionuclides can be identified and measured. Many elements have sensitivities that do not differ by more than a factor" of 100, and consequently, the bulk of the photoelectric peaks are easily masked when multiple activities are induced by activation (9). This interference is espeeially serious when the major el-emental constituents with large thermal neutron cross-sections are activated. Partial resolution of gamma-ray spectra can be obtained by variation of the experimental conditions, such as the irradiation and decay times. This approach is not practlcaI when a sample contains even AI, V, l<1n, Br, I, Au, etc., because these elements have large activation cross-sections and reach very large specific activities during neutron irradiation, subsequently masking the activities of the other radionuc1ides in the pulse height spectrum. Although Ge(Li) scintillation detectors considerably enhance the resolution of gamma ray spectra, their depletion depths presently limit their physical size and thereby their counting efficiency. Numerous proposals (10-12) have been reported wherein the instrumental resolution and interpretation of a gamma ray spectrum have been simplified by chemical separations. The importance of post-irradiation separations in: neutron activation analysis has been reviewed (9) and clearly substantiated in the literature (13-16). Currently, solvent extraction (17, 18), extraction with hydrated antimony pentoxide (HAP) (19), and'adsorption on Mn02 (20, 21) are the most common non-chromatographic separation techniques. Paper chromatography (22), thin layer chromatography (23, 24), ion exchange (25) and ion exchange membranes (26), and isotopic exchange (27) have been successfully combined with gamma ray spectrometry in order to give significant improvements in the analyses. 3

PAGE 9

.An automatic group separation system for the determination of 40 elements was reported by Samsahl, etal. (28), which is based on ion exchange chromatography and solvent extraction. Recoveries of for 29 elements in biological samples in about 2 hours were reported with a mean error of 3% in the analyses. A number of automatic or semiautomatic group separation systems for different kinds of materials have been reported (28-31). Perhaps the most --------------ex-t-en-siveandanalytical1-y useful s chemehasbeen reported by Morrison, etal. (32,33). They have developed a group separation scheme which enabled them to measure 41 trace level elements in the Apollo 11 lunar samples by NAA. Their method combines solvent extraction, separation with HAP, and ion exchange to give six groups which can be counted without serious interferences for trace amounts. Gas and liquid chromatographic separations can be expected to become of increasingimpoI"tance in neutron activation analysis (NAA) for the separation of radionuclide interferences in complex gamma ray scintillation spectra. Chromatographic separations are rapid, highly efficient, and can be developed without the use of carriers. Further, these separations can be optimized with respect to time, resolution, or sample size as required so of lived isotopes, multicomponent mixtures, or trace constitu-ents, respectively, may be effected. These separations are very important in trace analysis as they effectively give an infinite resolution of the pulse height spectra and enable the theoretical lower limits of detection to be realized. The popularity of the Ge(Li) detector is attributed to its high resolution and the minimum amount of "wet" chemical separations required. Gas chromatography with large volume NaI(Tl) detectors, however, can be automated, and even operated remotely, and therefore possesses all of the advantages of Ge(Li) detectors and has the additional advantage of having a much higher counting efficiency. Gas chromatography provides a quantitative postirradiation separation and effectively enchances the sensitivity of activation analysis by: 1. Removing matrix activities 2. Resolving overlapping photopeaks 3. Reducing the dead time of the analyzer 4. Isolating single radionuclide species in order to increase the quantitative precision of pulse height analysis and reduce the amount of interpretation of pulse height spectra and data reduction. 4

PAGE 10

Only a few papers have sought to develop gas and/or liquid partition chromatography for the separation step in NAA. From a recent review of radiochromatographic separations and analyses, it is clear that radiochromatographic separations have been used extensively in radiation chemistry, hot atom chemistry, and biochemistry but not for the separation of photon emitting radionuclides (34), Cram and Brownlee (35-37) first reported the use of gas-liquid chro matography for the clean-up of pulse height spectra in the functional group analyses of the halogens by (n,y) reactionso The advantages and workability of gas chromatography for high speed post-irradiation separations was demonstrated by eliminating the interferences Na24, A128, Br8D, and 1128 from the F20 spectrum. Further, this previous work resolved the 1.63 Mev photopeak of F20 from the 1.64 Mev peak of C138 in less than 5 seconds so that the analysis for fluorine with the 10,7 second half-life could be used directly. For the acti vation analysis of fluorine wi th an (n, y) reaction, nigh speed and high efficiency separations are imperative if meaningful analytical results are to be obtained. Even though the higher atomic weight members of the halogen series (i.e" Br and I) have considerably longer half-lives and larger activation cross-sections, they too must be chromato.. graph i Lally s e parat e d f Qx_an a 1 y sis ___ a t trace con ceI1"tX9t ions. However, organic analyses are rather limited in scope if one is restricted to (n,y) reactions, and thus it has been necessary to develop gas phase, partition separations for the activation analysis of the metallic elements. (As in the case of the organic functional group analyses, the materials to be separated must have a vapor pressure of at least 1.0 mID Hg at the separation temperature, and must be thermally stable at that temperature.) Metal analysis by gas chromatography has been shown to be both practical and useful (38). Volatile metal chelates such as the beta-diketonates, have been most often used in these analyses. The use of metal chelates has enabled more than three quarters of the elements in the periodic table to be separated and eluted from a gas chromatograph easily, quickly, and with high resolution. Trace analysis studies (39-41) and'the development of new ligands (42) have helped make metal analysis by gas chromatography a powerful tool with a wide range of applicabili ty. It is the inherent speed and selecti vi ty of gas chromatography combined with the chemistry of the metal chelates which makes the technique such an excellent separation method for neutron activation analysis. 5

PAGE 11

The synthesis of the metal chelates using beta-diketone ligands such as 1,1,1-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)] is usually simple. The chelates can be made either before or after irradiation of the metal mixture. Post-irradiation synthesis has the advantage of avoiding irradiation of the ligand and chelate. Each radionuclide is chelated quantitatively and is therefore in a form suitable for efficient separation. This advantage is often outweighed by the disadvantage of increased decay time due to additional sample treatment of the radionuclide between irradiation and counting. In order to take full advantage of the speed inherent. in the technique, chelation before irradiation is usually necessary. The irradiation stability of the metal chelates is important and corrections must be made for decomposition in the reactor. The radiolysis of reactor irradiated Cr chelates of [H(tfa)], [H(hfa)], and [H(fod)] has been studied (43). No thermal degradation in the reactor was found, but radiolysis losses of up to 10% were reported when the was It has been shown that Szilard-Chalmers reactions involving the central metal atom of a metal beta-diketonate yield retentions (percentage of radioactive atoms in the original chemical compound) as low as 4% (44). Szilard-Chalmers reactions involving solid metal chelates and the use of annealing to increase their retentions have been studied (44-58). The recoil and metalligand bond cleavage of chelates after (n,y) reactions have been used for iostope enrichment (45,46) and the metal is believed to exist as a cation in solution (44). SzilardChalmers reactions following (Y,n) reactions have also been discussed (47). Lazzarini (48) has studied the effect of the ligand field on Szilard-Chalmers reactions by studying the retention for 28 complexes of Co(III). Annealing studies have been carried out on Cr tris-acetylacetonate (49-51) and Co tris-acetylacetonate which are similar to the chelates used in this work. Repair of reactor irradiation damage can result from heating the chelate (44, 49-58) or exposing it to y-radiation (51,54). Thus, there is a moderate amount of annealing before a sample leaves the reactor. For the Co tris-acetylacetonate, each type of treatment has about the same effect (54), which is reported to be due to localized heating during yirradiation. Preheating of solid samples before neutron irradiation will also increase retention (55), and the presence of an electron donating atmosphere around a solid chelate during annealing has been shown to increase retention (44,57). 6

PAGE 12

Gainer and Ponta (58) have studied thermal annealing of Cr tris-acetylacetonate in methanol solutions. The retention was 32% after 2 hours of irradiation at 2.3 x lOll neutrons/cm2-sec, and increased with heating. In the plots of retention versus annealing time at constant temperature, the plateous had heights which were temperature dependent. At 95 C, 80% retention was obtained, but above 95 C, the retention reached a peak and fell off due to decomposition. Gainer and Ponta also found exchange bet,veen 51 Cr3+ in solution and the chromium in the chelate. The work described in this report used both the annealing described by Gainer and Ponta and the technique of adding excess ligand to the solution to be irradiated in an effort to obtain quantitative yields of the irradiated chelate prior to gas chromatographic separation. 7

PAGE 13

PART I CHARACTERIZATION OF THE IRRADIATION STABILITY OF SOME CHROMIUM BETA-DIKETONATES FOR GAS CHROMATOGRAPHIC SEPARATIONS IN NEUTRON ACTIVATION ANALYSIS Chromium beta-diketonates l,'lere chosen as model compounds for this study because the gas chromatographic separations of these materials have been studied extensively (38,40,59-64) and because of their physiological and toxicological importance at trace concentrations (65-81). Trifluoroacetylacetonate [Cr(tfa)3], hexafluoroacetylacetonate [Cr(hfa)3] and heptafluorodimethyloctanedione [Cr(fod)3] complexes of chromium were studied in order to measure the ligand effect on the stability during and after the irradiation, and the reactivity towards recombination reactions in the presence of excess ligand. The quantitative recovery of the chelates is then reported as a function of the type of ligand, irradiation time, solution concentration, and the effect of the presence of excess ligand. 1. Experimental The chromium chelates were irradiated as standard benzene (Nanograde, Mallinckrodt Chemical Works) solutions. The Cr(hfa) 3 was purchased from Pierce Chemical Company, and the Cr (tfa) 3 and Cr(fod) 3 were synthesized follm'ling published procedures (63,64). The ligands for the syntheses were purchased from Peninsular Chern. Research Corp., and the metals from Fisher Scientific Company. Quartz (General Electric Company) break seal vials were dral,'ln from 9 mm tubing and used as sample containers. These vials were leached, filled with the solution to be irradiated, and sealed with a torch after the solutions were frozen out. In this manner the vials were sealed air tight and, because of the lm'l sodium concentration, could be handled immediately after irradiation. The irradiations were performed in the University of Florida Training Reactor (UFTR) and the Georgia Institute of Technology Research Reactor (GTRR). The UFTR is a water cooled, MTR type reactor. At a steady state power level of 60 Kw and a thermal neutron flux of 1 x 1012 neutrons/cm2-sec, the temperature in the sample position was 60.5 C as measured with an iron-constantan thermocouple. 8

PAGE 14

The GTRR is a heterogeneous, 1 Mw, D20 moderated and cooled reactor. Since the outlet coolant temperature of the GTRR was only 1 C higher than the UFTR, the irradiation temperatures were assumed to be the same in both reactors. A comparison of the irradiation facilities used is given in Table 2. It should be noted that the irradiation times were adjusted to give the same integrated flux in both reactors. The gamma ray pulse height spectra were measured with two 3 inch x 3 inch NaI (Tl) crystals (Type l2Sl2/E-X, Harshaw Chemical Company). Nuclear Chicago Model 10-17 scintillation preamplifiers were used with RC amplifiers (Model 27001, Nuclear Chicago Corp.), superimposed in a mixer amplifier (Model 30-35, Nuclear Chicago Corp.), and counted in a 400 channel pulse height analyzer (Model 34-27, Nuclear Chicago). Data was readout on an ASR Model 33 Teletype and an X-Y plotter (Model 850, Data Equipment Company) A four column gas chromatograph (Series 4000, Victoreen Instrument Company) with an electron capture detector (Model 4008-3, Victoreen Instrument Company) was ------------lJu.,lS-ad-fn-r--the-s-B_pzI" at i on s Ih e-----Ble c t ron cap tll re de t e _______ was operated in the DC mode because this model was at least six times more sensitive for Cr(tfa)3 in the DC mode than in a pulsed mode, even though the pulse period 'vas varied from 25 to 125 with a pulse width of 0.5 Also the background noise in the DC mode was less (ca, 3.5 x 10-12 amps). Peak areas were integrated with a digital integrator (CRS-lOO, Infortronics Corp.) and a Hewlett Packard Model 7l27-A strip chart recorder was connected in series with the digital integrator for an analog display. The chromatograph was modified to bring the Teflon column into the injection port and to run up to the sensitive area of the detector in order to avoid decomposition or adsorption of the metal chelate in the instrument. The chromatographic system is shown in Figure 1. An extensive study of column materials and conditions showed that the conditions reported in Table 3 gave the most reproducible and quantitat.ive results for the metal chelates and their degradation reaction products. When the column effluent 'vas to be analyzed by mass spectrometry, the samples were trapped in V-shaped glass melting point tubes in a dry ice-acetone bath. For activation analysis of the effluents, low sodium quartz melting point tubes (General Electric Company) were used for the traps. At all other times a two stage absolute trap was connected to the detector outlet to prevent radioactive or toxic chelates from being vented to the room. 9

PAGE 15

TABLE 2 IRRADIATION FACILITIES Georgia Tech Research Reactor University Florida Training Reactor Irradiation time 7.0 hrs 4.0 hrs Thermal flux 5.6 X.l011 neut/cm2-sec 1 X 1012 neut/cm2-sec Power level 1 Mw 60 Kw Moderator D20 and Graphite Graphite Irradiation position Vertical thimble Center vertical port (reflector) Cd ratio 2000:1 1:1 Gamma flux 1 x 106 R/hr 2.4 x 106 R/hr 10

PAGE 16

Nzi CARRIER f-I GAS f-I JC.F Q GAS HELECTRON CAPTUREUjVIBRATING REEDH DIGITAL 1--1 RECORDER CHROMATOGRAPH DETECTOR I I IELECTROMETER INTEGRATOR 50 v. POWER I SUPPLY COLD I TRAP I PRINTER = TWO STAGE REGULATOR = NEEDLE VALVE Figure 1. Block diagram of the gas chromatotraphic separation system

PAGE 17

TABLE 3 CHROMATOGRAPHIC CONDITIONS FOR THE SEPARATION OF THE METAL CHELATES Column = 20 feet of 1/8 inch o.d. Teflon = 5% QF-l on 80/100 mesh Chromosorb W, AW (DMCS) Carrier gas flow rate = 60 ml/min of N2 Detector polarizing voltage = 30 volts Temperature Injection port Column Detector Cr(hfa) 3 1000 C 1500 C 12 Cr(tfa)3 2000 C 1600 C Cr(fod) 3 2000 C

PAGE 18

Mass spectra were run at 70 eVa on an RMU-6E mass spectrometer (Hitachi Div., Perkin Elmer Corp.). The batch liquid inlet system was heated to 750 C for Cr(hfa) 3 and 1250 C for the other chelates. 2. Results and Discussion Since these metal chelates are knmvn to be thermally unstable, standard benzene solutions of each of the chelates were prepared, sealed in the quartz vials, and heated in a laboratory drying oven at 610 C for a period of time equal to the time of irradiation. Chromatographic analysis of the solutions with the electron capture detector showed that there was no measurable thermal degradation at this temperature. Therefore the degradation which is reported for the irradiated samples must be due to radiolysis effects alone. In order to further isolate the variable parameters which effect chelate degradation, a Cr(hfa)3 solution was sampled periodically from 1 to 168 hours following irradiaorder to Szilard-Chalmers effect due to decay recoil reactions. This study showed that all of the measurable degradation appears to take place in the reactor, within the limits of the measurement error. J The effect of the irradiation time on the radiolysis degradation was also studied. Cr(hfa) 3 was chosen for this study because it was the most labile of the beta-diketonates studied. Figure 2 shows that the Cr(hfa) 3 is markedly more unstable in a reactor under a thermal neutron flux of 1012 neutrons/cm2-sec. Therefore it is not at all feasible to use the hexafluoroacetylacetone ligand for chromium analyses by GC-NAA as only about 50% of the chromium will be recovered for counting after an irradiation corresponding to 0.03% of the saturation activity for Cr51 The determination of the concentration dependence of the radiation degradation was of paramount importance in choosing the most stable ligand system and in order to do quantitative analyses. Concentrations from x 10-2 mg/ml to 5.3 mg/ml were used. This range was chosen in order to give a signal to noise ratio of at the low concentrations (which corresponds to 1.2 x 10-2 of chelate per 0.5 sample injected) and the maximum concentration was limited by the solubility of the chelate in benzene. The samples were chromatographically analyzed before and after irradiation at each concentration and the fraction of metal chelate remaining was calculated at the average of the peak areas for at least five replicate determinations. 13

PAGE 19

f' 6 A c I 5 I A '\ 1 en I 0-I I :IE
PAGE 20

Typical chromatograms of Cr(hfa) 3 before and after irradia tion are shown in Figures 2 and 3. After irradiation of the low concentration solutions of Cr(hfa) 3, three major peaks were found in the chromatogram. Peak A was identified as hfa by a comparison of retention times. Peaks Band C were trapped for analysis by mass spectrometry and gamma ray spectrometry. The former was identified as Cr(hfa) 3 but peak C was not posi ti vely identified. The mass spectra showed a parent peak at a mass to charge ratio of 644 but no chromium photopeak was found in the pulse height spectra after extended counting periods. The ratio of the peak heights of peaks Band C was found to be a function of concentration as the ratio of B to C increased with increasing concentration. This is the expected resul t and consis tent with the SzilardChalmers theory if the major degradation products resulted from cleavage of the metal-ligand bond. Such an equilibrium should be affected by the presence of an excess of ligand in favor of the recombination reaction and give larger recoveries of the metal chelate. The_resul_ts o the_Cr(hfal_3_s tudy __ are Sh01'LU_iu _._. __ Figure 4. The percent of metal chelate recovered from a benzene solution ranges from 81 to 96% and the recovery is seen to be significantly less in the presence of excess hfa (from 40.5 to 90.0% recovery). These results are seen to be consistent at all concentrations but cannot be explained for this system. A series of Cr(tfa) 3 solutions were also studied in order to determine the ligand effect. These solutions were similarly analyzed chromatographically before and after irradiation. The chromatogram of the Cr(tfa)3 and Nanogram benzene showed only the solvent and chelate peaks and no other peaks were discernable above the baseline noise (3.5 x 10-12 amps). Figure 5 shows the results of this series of experiments. The chelate recoveries ranged from 90 to 98% with increasing metal chelate concentration which is in marked comparison wi th the less stable hfa chelate. The aliquots containing excess ligand were not analyzed because the tails of the excess ligand peaks obscured the remaining chromatogram due to detector poisoning. Savory et al. (81) used sodium hydroxide to hydrolyze the excesS-ligand before analyzing by electron capture. However, he reported recovering only 80 -85% of the chelate. Since at least 90% of the irradiated chelate could be recovered without excess ligand and because of the uncertainty of the hydrolysis effect, the effect of the excess ligand was not measured. 15

PAGE 21

p (C 60 r-l-f/) Q. A :E 4.5 I-m '0 )( I.. I--Z w 3.0 I-0:: 0:: :::> 0 I-0:: 0 I--0 IJJ 1.5 l-I--IJJ C I\ \ \ -J, fa ..... _-----to-_.,J\... ___ -__ J o I J I I I I I I I I o 2 4 6 8 10 12 TIME, min. Figure Chromatogram of a saturated solution of Cr(hfa) 3 in benzene before (dashed line) and after irradiation (solid line) in an integrated thermal neutron flux of 1.4 x 1016 neutrons/cm2-sec 16

PAGE 22

I-' ' o u w a: 100 w r-80 :5 w :r: u ...J
PAGE 23

I-' co If') o )( en IZ ::l o U LLI a: ::.:::: LLI a. 1001 80 60 40 20 8 4 2 oj, ", ", a.....-:.. ,. ... t:'> --6--I!:l -6 [!] [!] ", [!] H \jJ (., e-NONIRRADIATED Cr (tfa)3 IRRADIATED CHELATE CALCULATED CHELATE RECOVERY C LLI a: LLI > o U LLI a: 100 80 ..J LLI :r: 60 u ..J 40 .LLI :E IZ LLI 20 u a: LLI a. II I I I 110 QOI .02.03.04.06 .08 0.1 0.2 0.3 0.4 0.6 0.8 1.0 2 3 4 6 8 10 CONCENTRATION mg/ml Figure 5. Experimental uata for the recovery of ctCtfa)3 following thermal neutron irradiation

PAGE 24

Chromium heptafluorodimethyloctanedione solutions were run under similar conditions and analyzed as before. These solutions Ivere also found to be clean before irradiation within the limits of detection. The recoveries shown in Figure 6, are seen to increase with concentration from 90.0 to 100.0%. The results of the Cr(tfa)3 and Cr(fod)3 studies were particularly pleasing as they indicate that these chelates are stable to irradiation times up to 7 hours and are analytically useful for chromatographic post-irradiation separations in neutron activation analysis. The relative irradiation stabilities of the three chromium beta-diketonates was found to be Cr(fod) 3 Cr(tfa) 3 > Cr(hfa) 3. The amount of Cr(hfa) 3 recovered is more a function of concentration than are the metal chelates of the other ligands. The presence of excess ligand appears to be detrimental to the quantitative recovery of the Cr(hfa) 3. The chromatograms showed no radiolysis peaks to obscure the chromatogram with either Cr(tfa)3 or Cr(fod)s and this study has clearly shown these to be the chelates of choice for chromium analyses. 19

PAGE 25

It) 0 )( en Jz :> 0 u N 0 oCt LLJ a:: oCt oCt LLJ a.. 800 600 400 I}// ./ 1j1 20 I --e-:---6""'i -Cllt--Jf' NONIRRADIATED CrUod)3 IRRADIATED CHELATE CALCULATED CHELATE RECOVERY 0 20 I 01 I I I I i I I I I I I !, I I C' 0.01 .02 .03 .04 .06.08 0.1 0.2 0.3 0.4 0.6 0.8 1.0 2:3 4 6 8 10 CONCENTRATION Figure 6. Experimental data for the recovery irradiation OF Cdfod)3 v mg I mi of Cr fod) 3 following thermal neutron 0 w a: LLJ > 0 U LLJ a: w t-oCt -.J W ::I: U -.J
PAGE 26

PART THE QUANTITATIVE GAS CHROMATOGRAPHIC SEPARATION AND ANALYSIS OF RADIOACTIVE, VOLATILE METAL BETA-DIKETONATES It has been shown that in order to obtain reproducible chromatographic peak areas with the metal chelates, several injections of the chelate solution must be made to "condition" the coulmn before quantitative measurements are made (38). This suggests that the first few injections leave chelate in the instrument, and only after these chelate adsorption sites are filled can a sample be eluted quantitatively. Gainer and Ponta (58) have that radioactive chromium ions in solution will exchange with chelated chromium. Exchange reactions might also take place between the chelate in the vapor phase and the chelate held on active sites in the chromatographic column. Under these conditions, the mass of chelate may be quantitatively recovered, but there is no assurance that the activity eluted is representative of the In order to keep these reactions to a mlnlmum, glass and Teflon systems have been used. It has been further shown that metal chelates will react with metal columns, and it is suspected that metal injection ports will aid and even catalyze the decomposition. Treatment of the chromatographic system with silanizing agents has also been used in practice to decrea.se on-column adsorption. Radiotracer techniques have been known for a long time and applied to metal chelate synthesis and extraction studies (39). However, there has been no application of the technique to the determination of metal chelate residues in gas chromatography. This determination of metal chelate residues is a necessary step in the activation analysis-gas chromatographic technique. 1. Experimental The gas chromatograph used in these studies was a Varian Model 1200-1 equipped with a 1/4 inch injection port, a hydrogen flame ionization detector and a linear temperature programmer. The columns were 1/8 inch Teflon (Alpha type, TFT-250/ll), 0.095 inch I.D. and 0.016 inch in wall thickness The seals at each end of the column were made by forcing the 21

PAGE 27

column inside a constriction in a Pyrex tube. The injection port linear was a 3 mm O.D. Pyrex tube tapered to fit inside the Teflon column. All of the columns and their operating conditions are given in Table 4. Liquid samples were injected with a Hamilton 10 syringe and a solid sampler was used to inject the sealed ampules (64). Samples to be injected using solid sampler were sealed in quartz tubes 2 cm x 2 mm. Both ends were sealed with a methane-oxygen torch. After filling, the tube was placed inside a 1/8 inch copper tube which was placed in a dry-ice-acetone bath. This prevented decomposition of the chelate by the heat from the torch. Solid sample injection has many advantages. There is no solvent to interfere with trapping the effluent or obscure the chelate peaks when using a standard gas chromatographic detector. The maximum amount of chelate which can be placed in the solid sampler without overloading the column would require too much solvent for a satisfactory injection. In addition, the sealed quartz tube which contains the sample during irradiation can be placed directly in the so] i d sampl ing device wLthont any of the handling or open vessels required for the more conventional liquid injection. Several sampling techniques have been developed for gas chromatography. The type which uses a plunger to break a quartz tube containing the sample is a logical choice for the gas chromatography of irradiated metal chelates. In order to study the effect of sample size, a series of injections from 1 10 of a benzene solution of 32 mg/ml of Cr(tfa)3 were runon column 6. The H2 flow rate on the flame ionization detector was 46 ml/min and the O 2 flow rate was 135 ml/min. All peak areas were measured with a digital integration Model CRS-IOO (Infortroriics, Inc.). 2. Quantitative Column Elution To evaluate the effectiveness of silanization of the gas chromatographic system, a series of 1 injections of a solution with 29 mg/ml of Cr(tfa)3 in benzene were injected into column 5. The study was repeated with a Cr(fod)3 solution of the same concentration on column 7. After 4 to 6 chelate injections, 10 of DMCS was injected to silanize the system. 22

PAGE 28

TABLE 4 CHROMATOGRAPHIC COLUMNS CONDITIONS He Flow Column Literat. Length Liquid Solid Rate, Temperature, C No. Ref. ft Phase Support m1/min IP. Col. Det. Che1ates 1 39 4 5% SE-52 60/80 mesh I 50 170 80 Be(tfa)2 Gas Chrom. Z 2 2 20% High 100/120 meshl 75 170 150 Cr (hfa) 3 vacuum AW, DMCS grease Chrom. P 3 4 5% QF-1 80/100 mesh 60 120 100 150 Cr (hfa) 3 Chrom. W N 4 4 5% QF-1 80/100 mesh 30 150 100 150 Cr (hfa) 3 tN Chrom. W 5 4 5% QF-1 80/100 mesh 60 160 135 160 Cr(tfa)3 Chrom. W 6 4 5% QF-1 80/100 mesh 30 170 150 170 Cr(tfa)3 Chrom. W 7 4 5% QF-1 80/100 mesh 60 170 150 170 Cr(fod) 3 Chrom. W 8 4 5% QF-1 80/100 mesh 30 200 175 Cr(fod)3 Chrom. W 9 3 15% QF-1 80/100 mesh 20 165 100 165 Mn(hfa) 2 AW, DMCS Chrom. W

PAGE 29

<, TABLE 4 -CONTlfUED He Flow Column Literat. Length Liquid Solid Rate Temperature, C No. Ref. ft Phase m1/min IP. Col. Det. Che1ates 10 64 7.5 10% SE-30 60/80 mesh I 205 170 Fe(fod)3 Gas Chromo Z 11 2 20% High 100/120 meshl 75 170 150 Cu(hfa) 2 vacuum AW, DMCS grease Chrom. P 12 83 2 5% SE-30 /200 meshl 60 170 130 Cu(tfa)2 Chromo W 13 1.5 20% SE-30 100/120 meshl 30 150 150 Cu(hfa) 2 N AW, DMCS .j::>. Chromo P 14 82 6 10% SE-30 80/100 mesh 30 230 170 Gd(fod) 3 Chromo W 15 82 6 10% SE-30 80/100 mesh 30 230 170 Lu(fod) 3 Chromo W

PAGE 30

In order to evaluate the elution of irradiated chelates, solid samples of Cr(hfa) 3 and Cu(hfa)2 were irradiated for 4 hours in the UFTR at x 1012 neutrons/cm2-sec and injected into columns 2 and 11, respectively. One mg solid samples of Mn(hfa) 2 were irradiated 15 min at the same flux and run on column 9. After the chelates eluted, the solid sampler, charcoal trap, and column inlet, column, and outlet system were counted. The quartz fragments in the injection port were counted, the injection port rinsed with acetone, and the solution counted. Similar system counting was done after irradiated solutions had been injected: Cr(hfa)3 in column 4, Cr(tfa)3 in column 6 and Cr(fod)3 in column 8. In order to separate the effects of adsorption and on column reaction from the radiolysis decomposition of the metal chelates, pure radioactive chelate solutions were prepared. These were injected into their respective columns (see Table 4) in a series of 5 injections, 1 min apart. The series of injections totaled 50 and 100 After the carrier gas had been allowed to flow after the last injection for 3 times the retention time of the chelate, the system was dismantled and the components counted. The packing was cQuIlted5n a 60 x 15 mIll Pyrex petri dish. 3. Counting System All radioactivity measurements were made using two 3 x 3 inch NaI(Tl) detectors connected to the multichannel pulse height analyzer system previously described. A block diagram of the counting system is,shown in Figure 7. The three different counting geometries shown in Figure 8 were studied and they are described in detail elsewhere (34). The exponential dilution flask exhibited sever adsorption properties and was difficult to heat uniformly, and consequently, will not be considered here. The spiral coil flow-through geometry consisted of a 57 inch piece of 1/8 inch O.D. #316 'stainless steel tubing with a 0.005 inch wall (Superior Tube Company). It was wound in a spiral 2 7/8 inch O.D. x 1 1/8 inch I.D. The inlet and outlet arms each consisted of 6 1/2 inches of tubing. The charcoal traps were made from 5 inch pieces of tubing packed with 1 inch of 40 50 mesh charcoal. The packing began at the center and extended toward the outlet. Some of the charcoal traps were made from the same stock as the spiral. Others were 1/8 inch O.D. Teflon lined aluminum tubing or 3 mm O.D. Pyrex. 2S

PAGE 31

N 0\ 'l / i J I No I 0 Nol I I "-,/ 1 HIGH VOLTAGE I DISTRIBUTION P:ANEL I,. HIGH VOLTAG:E I POWER SUPPLY i LINEAR RC REGULATED LINEAR RC AMPLI F I ER POWER SUPPL. Y AM FIER i I SUMMING I i I AMPLI FI ER j i I I i L ___________ __ I I TELETYPE I I I I INTEGRATOR : 400 CHANNEl. I !POINT PLOTTER I PULSEHEIGHT ANALYZER! I DEAD TIME I I J I i I Figure 7. Block diagram of the NarCTl) and counting system used for radiochromatography

PAGE 32

INJECTION --PORT HE CARRIER GAS GAS CHROMATOGRAPH COLUMN 1----v.D.C. 01 D 0: 0 I \ I THROUGH D '" / I LiQ N2, TRAP Nol DO DO DO DD i1U., II II II c:=sc II II II 10 Figure 8. Counting geometries studies for measuring the radiochromatographic effluent activity 27

PAGE 33

The reversible counting geometry consisted of a 6 inch piece of Teflon tubing packed with 2.5 inches of 20% SE-30 on Chromosorb P, (AW DMCS). 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 geometry. A He purge line was connected upstream to help flush the counting geometry. The sample counting in geometries connected to the system was done in a cave beside the gas chromatograph. A heated Teflon tube passed through the oven wall into the cave. The cave had 8 inches of lead between the detectors and the gas chromatograph, 4 inches of lead on the side, top, and back, and no lead on the bottom and front. The inside dimensions were 24 x 20 x 8 inches. Sample counting, where the traps were removed from the gas chromatograph system, was done in a lead cave 40 x 22 x 12 inches with 4 inches of, lead on 5 sides and 2 inches on the bottom. The response of the NaI(TI) detector as a function of was obtained by passing a 0.2 inchlong byO .061 inch diameter--piece-ofi1l.oleculaT sieve containing CS-137across the surface of the can containing the detector. Two perpendicular passes were made. The time sequence store mode of the analyzer was used to obtain a plot of relative response versus position. Three different modes of counting radioactive metal chelates were used. Where two or more radionuclides were present, a pulse height spectrum was taken and the areas under the photopeaks integrated using a PDP-8/L laboratory computer (Digital Equipment Corp.). When only one radionuclide was present, a single channel analyzer was used and the entire spectrum above a fe1.v kev was measured. When radiochromatograms were made, the time sequence store mode of the analyzer 1.vas used. The flow through spiral counting geometry "\'las used at 1700 C with a continuous He carrier gas flow of 30 ml/ min to obtain radiochromatograms of Cu(hfa) 2 -Cr(hfa) 3 mixtures using column 2. The time sequence store mode of the analyzer was used 'vith a dwell time of 0.02 min/channeL The stop-flow counting geometry was operated in a flow-through mode at a constant temperature of 1150 C with a sorbent identical to column 13 inside. The geometry was held at 600 C to trap effluents, such as the Cu(hfa)2, and then heated to 1500 C to desorb the complex. The procedure for repetitive use of the reversible stop-flow counting system was: 28

PAGE 34

1. Start carrier gas, apply cool air to geometry 2. Inject sample 3. Wait 2.5 3.0 min, stop carrier gas flow 4. Count sample 10 min 5. Heat geometry and allow purge gas to flow for 15 min 6. Cool geometry, stop purge gas, and make a 10 min background count 7. Restart carrier gas and trap the next component. Irreversible charcoal traps were used to obtain integral radiochromatograms. The metallic traps were cooled with dry ice, but the 3 mm Pyrex traps required no cooling. The integral chromatograms were obtained by stopping the carrier gas flow and counting the accumulated activity in the trap. Cu(hfa)z -Cr(hfa)3mixtures were injected as solids and counted using this technique. 4. Chelation Studies and H(od) ____ Research) were all redistilled prior to use. For H(tfa), the 1070 C fraction was collected. The CU(hfa)z and Cr(hfa) 3 were used as purchased (Pierce Chemical Company). The Mallinckrodt Nanograde benzene was dried over a l3X molecular sieve. The Vitreoseal quartz for the ampules was obtained from the Thermal American Fused Quartz Company. Cr51C13 in 1 N HCl, 0.02 mg/ml, 3.7 mCi/ml (Union Carbide), Fe59C13 in-l N HCl, 2 mCi in 0.83 ml(New England Nuclear Corp.) and Be7Clz in 0.5 N HCl, 2 mCi in 0.5 ml (New" England Nuclear) were used as tracers in these experiments. The atmospheric sublimation apparatus consisted of an 8 cm x 6 mm O.D. Pyrex tube bent into a "U" shape with a connecting tube on each end. The apparatus was heated in the gas chromatographic oven. The He flO1'l through the tube and the heat applied by the oven were sufficient to sublime the chelates and move them from the bottom of the "U" through a short piece of Teflon tubing to a 3 mm O.D. Pyrex trap outside the oven. This trap was maintained at room temperature. The vacuum sublimation apparatus was a standard design ''lith a 15 mm O.D. Pyrex pot and an 8 mm O.D. cold finger. It was connected to a mechanical forepump. 29

PAGE 35

Mixtures of 2 mg of solid Cr(hfa) 3 and 1 mg of solid Cu(hfa) 2 were sealed in 2 mm x 2 em quartz tubes and irradiated for 4 hours at 1012 neutrons/cm2-sec in the University of Florida Training Reactor (UFTR). Solid 1 mg samples of Mn(hfa) 2 were sealed and irradiated under the same conditions for 15 min. The Cr(hfa)3 and CU(hfa) 2 mixture was injected into columns 2 and 11 (see Table 4), and the Mn(hfa)2 was injected into column 9. No annealing treatment was used on either sample. The chromatographic column was then dismantled and each component counted. Cr(tfa)3, Cr(fod)3, and Cr(hfa) 3 were dissolved in Nanograde benzene, irradiated for 14 hours at 6.5 x lOll neutrons/cm2-sec in the Georgia Technology Research Reactor (GTRR), and injected without further treatment into columns 5, 7, and 3, respectivelyo Each of these columns, charcoal traps, and injection port liners were similarly counted. Thermal annealing of irradiation damage in Cr(tfa)3 was studied by irradiating two benzene solutions: one containing 30.2 mg Cr (tfa) 3 and 2. 7 mg H (tfa) (10 % exces s) in 1 ml of benzene, and the other contained 30.4 mg Cr(tfa) 3 and L7mg H(tfa) CIOOt_S!xce;;sJ_iI1 J_mlQf 14 hours at 6.20 x lOll neutrons/cm2-sec in the GIRR.Part-of the sample with 10% excess ligand was injected into column 5 without treatment. The rest was annealed (heated) 9 hours at 1000 C and then injected. The sample with 100% excess ligand was annealed 9 hours at 1000 C in its sealed irradiation vial and injected. All three columns were dismantled and counted. Hot Cr(tfa)3 was by synthesizing the chelate using Cr spiked with Crs as the starting material. Forty mg CrC13 6 H 20 was spiked cmd dissolved ,in 1.5 ml H 20 in a 5 ml round bottom flask. Three hundred mg of urea and 86 mg (10% excess) of H(tfa) were added, and the mixture refluxed at 100 1150 C for 7 hours. The insoluble product was filtered, washed with H 20, and air dried. The product was dissolved in ether and placed in the atmospheric sublimation apparatus. After the ether had evaporated, the temperature was raised to 1500 C for 15 min. Forty-seven mg of purple product was recovered from the trap under a flow rate of 10 ml/min. Radioactive Cr(fod)3 was also prepared by synthesizing the chelate using chromium spiked "\vi th CrSl as the starting material. Fifty-two and one-half mg Cr(N03)3 9 H 2 0 was spiked and dissolved in 0.5 ml absolute ethanol in a 10 ml round bottom flask. One hundred thirty-four and two tenths mg (10% excess) H(fod) dissolved in 1 ml absolute ethanol 30

PAGE 36

was added. After heating for 1 hour at 800 C, the mixture was transferred to the atmospheric sublimation apparatus. Heating for 1 hour at 1400 C with 10 ml/min He flow transferred 6.5 mg chelate to the trap. A solution containing 43.5 mg Lu(fod)3 and 42.1 mg H(fod) in 1.5 ml benzene was prepared. One ml of this was irradiated for 14 hours at 5 x lOll neutrons/cm2-sec in the GTRR. It was annealed 10 hours at 1000 C. The sample was vacuum sublimed at 1150 C for 15 min. Thirty-three and onehalf mg p"ale yellow chelate was obtained. The same procedure was applied to Gd(fod)"3' except that the sublimation was for 45 min at 1500 C. Radioactive Fe(fod) 3 was synthesized from FeC1 3 spiked with Fe59 dissolved in ethanol. H(fod) was added and immediately after mixing, the deep red chelate was formed. It was extracted into 2.5 ml hexane and sublimed at atmospheric pressure at 1400 C. Six mg of deep red chelate was obtained. The pale red crystals were discarded. Radioactive Be(tfa)2 was synthesized from BeS04 4 H 2 0 spiked with Bel. Forty-one mg BeS04 4 H 2 0 and 90.0 mg NaAc were dissolved in 1.5 ml H20. One hundred mg (50% excess) H(tfa) dissolved in absolute ethanol was added. The white product was extracted into benzene and vacuum sublimed at 650 for 15 min. Thirty-four and a half mg chelate was obtained. The Cu(tfa)2 was synthesized according to the method of Berg and Truemper (84). A saturated benzene solution with excess solid chelate, 65.5 mg total was irradiated for 7.5 hours at 1012 neutrons/cm -sec in the UFTR. The sample was annealed at 1150 C for 7 hours and vacuum sublimed at 1150 C. 5. Results and Discussion Both solid and liquid samples were evaluated in this study. The solid sampler's large sample size (1-2 mg) was ideal from the standpoint of counting, especially when using the flow-through type geometries. However, the activity measurements made on the solid sampler and quartz fragments in the injection port showed large amounts of radioactive residue, most of which was due to SzilardChalmers decomposition. There were visible chelate residues in the sampler barrel and among the quartz fragments, indicating ei ther that the injection port 'vas not hot enough, or that the solid sampling technique needed to be improved. This contributed to the injection port activity to a small extent. 31

PAGE 37

The length of the sealed sample tubes which were broken by the solid sampler was found to be critical. If the tubes were too long, large amounts of glass (or quartz) fragments prevented all the sample from leaving the sampler barrel. If the sample tubes were too short, they were not broken at all, and were forced upward,breaking the injection port liner which surrounded the solid sampler barrel. For 0.060 inch O.D. melting point capillaries, the best lengths were found to be between 0.406 inches and 0.469 inches. For liquid injections of 5 or less, the total sample activity was so low that long counting times were requi red. lVhen a total of 10 0 of Cu (tfa) 2, inj ecte d 5 at a time (in order not to overload the column) 7 x lOG cpm were measured in the trap. If all factors affecting recovery are linear, a 1 injection should have at least 7 x 104 cpm acti vi ty\vhi ch is roughly 70 times b ackgrounc1, The results of the injected sample size study aye shown in Figure 9. The benzene Cr(tfa) 3 chromatograms gave one peak for benzene and two chelate peaks. The second chelate peak area was only 10% of the first and it was necessary to integrate both chelate peaks together. From _these results i tis __ seen_that the 10 sample deyiatE;s __ significantly fyom linearity and is therefore considered as the upper limit of liquid sample injections. In terms of eluting the chelates quantitatively, the silanization did not produce any measurable improvement. Even after 70 of DMCS has been injected, the chelate peak areas were the same as those observed before any DMCS had been injected. For all of the solid injections the elution efficiency was poor. Eighty ninty% of the acti vi ty was found in the injection port, the solid sampler, and the column inlet. Both the quartz fragments and the acetone rinse were radioactive. Most of the activity was removed from the injection port by this procedure. The only time any radioactive chelate was measured was when the sample activity was so high that the 5% which was eluted 1vas hot enough to be counted. The color and appearance of the solid Cr(hfa) 3, Cu(hfa)2, and Mn(hfa) 2 did not change during irradiation. Since only 2 atoms in 109 of the chelate undergo the (n,y) reaction during a 15 hour irradiation at 1012 neutrons/cm2 -sec, only a small fraction of the sample is effected by the Szilard-Chalmers reaction. Thus, even though these solids may have been essentially unaffected by irradiation, the chelation of the radioactive atoms was by no means quantitative. This was shown by the results of the gas chromatograph residue measurements. When solid Cr(hfa) 3 and CU(hfa) 2 were 32

PAGE 38

t.no 10 )( .. e a::
PAGE 39

injected into columns 2 and 11, respectively, large amounts of activity were found in the solid sampler, the injection port, and the head of the column. Smaller amounts of activity were found at the middle of the column and at the outlet. When a charcoal counting trap was used for the above experiment, approximately 50% of the radioactive sample '.'Jas found in the charcoal trap. Undoubtedly, there was some annealing in the reactor and/or the injection port. After the irradiated solid was injected, no trace of Mn 56 was found beyond the inj ection port. There was apparently insufficient annealing to rechelate any Ivin 56. The Cr(tfa)3, Cr(fod)35 and Cr(hfa) 3 solutions behaved the same way as the solids. Very small amounts of activity were found in the counting geometry and large amounts stayed in the injection port. The use of a Teflon injection port liner showed no improvement over quartz. The injection port retention was not due to the liner material, since hot chelate passed through the rest of the Teflon tube satisfactorily. results of the Cr (tfa) 3 annealing experiment are shown in Table S. The-residUe sLudy1\TaS used-to evaluate the purity of the radioactive chelate. The figures from an untreated sample similar to this one are included for comparison. It can be seen that heating at 1000 C for 9 hours in conjunction lvith irradiation with a 100% excess amount of ligand is quite effective in minimizing the amount of unchelated radioactive metal. When only 10% excess ligand was used the results were much better than when no treatment was used. Still, the results were poor. Heating caused some improvement, but only with 100% excess ligand were the results optimum. The solutions which were irradiated and not annealed were equally unsatisfactory. Here, there lvas no significant solid residue in the injection port as most of the metal chelate was eluted through the column. Only the small number of radioactive nucleii apparently underwent radiolysis and were consequently not in a form sui table for chromatographic elution. Only after annealing can irradiated chelate solutions be used for gas chromatography (see Table 5) The solution of Cr(tfa)3 which was irradiated with 100% excess ligand and then annealed gave excellent chromatographic results. The sublimation step used in the residue study for purifying the irradiated mixture is apparently not necessary for efficient elution. Table 6 lists the results of the residue studies of the chelates. 34

PAGE 40

" TABLE 5 EFFECT OF POST-IRRADIATION ANNEALING OF Cr(tfa)3 IN BENZENE Activity Recovered, 9.: 0 of Total Annealing None None 9 hrs, 100DC 9 hrs, 100C Excess Ligand 0% 10% 10% 100% Injection port liner 89.7 11.0 2.80 0.88 Column paCking} 3. 8 35.8 3.0 1. 43 Column tubing 0.25 0.7 0.38 Charcoal trap 7.0 52.9 93.5 97.30 NOTE: Chromatographic analysis run on Column 5 (Table 4) 0 3S

PAGE 41

,0 TABLE 6 PERCENT RECOVERY OF METAL CHELATES IN CHROMATOGRAPHIC SYSTEM Residual Activity:, % of Total Column Injection Column Teflon Chelate No. Port Liner Packing Tubing Cr (tfa) 3 5 0.35 1.36 0.38 Cr (tfa) 3 a 5 11. 06 35.81 0.25 Cr(tfa) 3 b 5 2.80 2.97 O. 70 Cr(tfa) 3 c 5 0.88 1. 43 0.38 Cr(fod) 3 7 0.35 15.7 0.75 ------------Be(tfa)2 1 0.07 2.7 3.-b Fe (fod) 3 10 0.30 23.0 2.48 Cu(tfa)2 12 1. 30 23.0 3.5 Gd(fod) 3 14 1.0 35 1.3 LuCfod) 3 15 0.51 7.0 0.43 aIrradiated with 10% excess ligand bIrradiated with 10% excess ligand and annealed cIrradiated with 100% excess ligand and annealed 36 Trap 98.16 52.87 93.50 97.30 83.0 2 -74.3 72.2 63 92.1

PAGE 42

The Cr(tfa)3 results were quite good. The poorer Cr(fod)3 results may have been due to a higher column temperature which would have contributed to decomposition on the column packing, since a large portion of the activity was found there. The Be(tfa)2 results were similar to those of the Cr(tfa)3. The third column, however, appeared to hold a large portion of the activity. There was no explanation for this. It is not a problem with sample purity because the injection port did not contain any more activity than it did in the other two runs. The Gd(fod)3 data was less accurate than the other experiments because the sample had decayed through 3 halflives considerably before it was counted. This meant that a mixture of daughters and other Gd isotopes were present, the half-life of which was unknown, and therefore no decay correction could be made. The results are accurate to the extent that they show nearly 1/3 of the total activity on the column packing. This is different from the Lu(fod)3 result, even though they are both rare earths and both us e diden ti This--is--s imi laFto the -_ .. Cr(tfa)3 -Cr(fod)3 results. The Fe(fod)3 also left a large residue on the packing material. Some was also present on the column tubing, which would indicate that the chelate and liquid phase did interact and that accurate quantitative results will be difficult to realize. The Cu(tfa)2 behaved similarly. The larger amounts of residue in the column is due to the low operating temperature for that particular experiment. In summary, the Cr(tfa) 3, Be (tfa) 2, and Lu(fod) 3 'vi th the columns which !Vere used wi th them seem to be good choices for gas chromatography of radioactive metals. The rest of the chelates seem to be restricted by the nature of the column packing. A better choice of chromatographic conditions, then, will probably make activation analysis work with these chelates feasible also. Further, it is seen that reaction irradiation of the metal chelates requires postirradiation annealing of both solid and liquid samples. The relative response of the Na(Tl) crystal as a function of sample position is an important consideration in the design of counting geometries. Consecutive samples need to be held in regions of the same response if reproducible results are to be obtained. Figure 10 shows the plot of response versus position for the 3 x 3 inch NaI detectors used in this study. The region where the response decreased by no more than 10% from the maximum was found 37

PAGE 43

VI O:l l&J en z 0 Q. en w a: w :> ...J w a::: '" I-t. OUTSIDE OF CRYSTAL I I 5 10 15 20 25 130 35 40 I DISTANCE FROM CENTER Of CRYSTAlm MMe I 1:' Figure 10, Response of the NaT (Tl) detector as :;t function of position on the crystal face, Points denoted by circles represent the I response of the fctcing side of the detector face

PAGE 44

:: Figure 11. Cu(hfa)2 activity counted in a stainlesslsteel flow-through counting geometry, after separation from a Cn(hfn):,-Cr(hfn) 3 mixture

PAGE 45

tl = /0 min. 3.9 Kev/Chan. Cr 51 0.322 Mev ..J ILl ;z ;z :r I I 10K u 0:: I Cu64 w I a.. .;::.. O.SII M.\!. C/) t-Z :::> 0 u CU64 ------\".. 'e""""" __ Figure 12. Multichannel pulse-height spectrum taken a;fter thegas chromatographic separation of Crs1roJlolVCd IlY the elutioll of CU64 PHOTON I. 34 Mev. ENERGY, M IV.

PAGE 46

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 flexibi Ii ty, indefini te counting time, freedom from cri tical stop flow timing, and complete reversibility. Figure 13 shows a 400 channel radio chromatogram ich illustrates the reversibility of the geometry. A solid Cu(hfa)2 sample was injected into column 13. The activity suddenly rises as the radioactive plug enters the geometry. Since flow is not stopped the tail of the chromatographic peak accumulates and the count rate slowly increases. At channel 364, heat was applied to the 60 geometry and the trapped chelate started to move. 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. 42

PAGE 47

;:. J-l 5.0 i II t W i i ow i 'T few, rt> 0 ......... )( .. ...J W Z Z J: u 2.5 0:: W c.. 2.0 (f) ...... z :J 0 u 1.5 1.0 0.5 0 0 1.0 200 TIME 04 .b 400 5 7. I AFT R Figure.; 13. lzacliochrornatogram illustrating the counting of Cu(hfa) 2. in the revers Ie counting geometry

PAGE 48

ACKNOWLEDGEMENTS The diligence and dedication of the members of this research group during the period of this grant has made this work possible. They are the ones who are responsible for the productivity of this project. Particularly, the efforts of Dr. R. W. McCoy F. T. Varcoe, Mr. T. R. Booher and Miss J. Durham should be recognized. The collaborative efforts and stimulating discussions of the entire group are sincerely appreciated. 44

PAGE 49

LITERATURE CITED 1. Fishman, M. J., Robinson, B. P., and Midgett, M. R., Anal. Chern., 39, 261R (1967). 2. Ok un D. A. ,eta1., J. Water Pollution Control Federation, 587(1965) 3. Christman, R. F., et a1., J. Water Pollution Control Federation, 4. Kitano, Y., and Nasu, Y., Bunsek, Kagaku, Shinpo Sosetsu, 1964, 140R, C.A., Ql, 12873c. (1965). 5. Monnier, D., Arch. Sci. (Geneva),.!l, 273 (1965). C.A., 64, 4253b (1966). 6. Landstrom, 0., and Wenner, C. G., Aktiebolaget Atornenergi, Stockholm AE-204 (1965). C.A., 12361g (1966). 7. Yule, H. P., Anal. Chern., 818 (1966). 8. Leddicotte, G. W., and Moeller, D. W., U.S. At. Energy Comm., CF-61-5-118 (1961). 9. Ross, W. J., Anal. Chern., 1114 (1964). 10. Albert, P., Pure App1. Chern., 1, 111 (1960). 11. Brownlee, J. L., U.S. AEC Rept. TID-6311 (1960). 12. Kusaka, Y., and Meinke, W. W., Nat1. Acad. Sci., Nat1. Res. Council, NucL Sci. Sere NAS-NS-3104 (J,961). 13. Ruch, R. R., Tera, F., and Morrison, G. H., Anal. Chern., 1565 (1965). 14. Brune, D.,and Landstroern, 0., Radiochirn. Acta, 228 (1966) 15. Srna1es, A. A., Mapper, D., and Fouche, K.F., Geochirn. Cosrnochirn. Acta, 31, 673 (1967). 16 Thompson, B A., and LaFleur, P D., Anal. Chern. ,41, 852 (1969). 17. Hahn, K. J., Turna, D. J., and Quaife, M. A., Anal. Chern., 1169 (1967). 45

PAGE 50

18. Hahn, K. J., Tuma, D. J., and Sullivan, J. L., Anal. Chern., i.Q., 974 (1968). 19. Girardi, F., and Sabbioni, E., J. Radioana1. Chern., !, 169 (1968). 20. Shipman, W. H., Anal. 1175 (1966). 21. Bigliocca, C., Girardi, F., Pauly, J., Sabbioni, E., Meloni, S., and Provasoli A., Anal. Chern. '-39, 1634 (1967). -22. Patek, P., and Sorantin, H., Anal. Chern. 1458 .(1967). 23. Houtmann, J. P. W., Proc. Intern. Conf. on Activation Techniques in the Life Sciences, lAEA, Vienna, p. 171 (1967). 24. Bark, L. S., Duncan, G., and Graham, R. J. T., Analyst, 347 (1967). --------2-5 Te-r-a: F., and-M0-T-r-i-s--en G. H., :An al. Chern., (1966) 26. Eisner, U., Rottschafer, J. M., Berlandi, F. J., and Mark, H. B., Anal. Chern., 39, 1466 (1967). 27. Tang, C. W., and Ma1etskos, C. J., Science, 167, 52 (1970) 28. Samsahl, K., Wester, P.O., and Landstrom, 0., Anal. Chern., i.Q., 181 (1968). 29. G., Diebolt, J., Junod, E., and Laverlochere, J., Proc. Intern. Conf. Modern Trends in Activation Analysis, College Station, Texas, 1965. 30. Girardi, F., Merlini, M., Pauly, J., and Pietra, R., Proc. Symp. Radiochemical Methods of Analysis, Vol. 2, Salzburg, 1964, p. 3. 31. Moiseev, V. V., Kusnetsov, R. A., and Kalinin, A. J., Proc.lntern. Conf. Modern Trehds in Activation Analysis, College Station, Texas, 1965. 32. Morrison, G. H., Gerard, J. T., Travesi, A., Aurie, R. L., Peterson, S. F., and Potter, N. M., Anal. Ch em., i.!., 16 33 ( 19 6 9) 46

PAGE 51

33. Morrison, G. H., Gerard, J. T., Kashuba, A. T., Gangadharm, E. V., Rothenberg, A. M., Potter, N. M., and Miller, G. B., Science," 167, 505 (1970). 34. Cram, S. Advances in Chromatography Vol. 9, J. C. Giddings and R.A. Keller, eds., Marcel Dekker, New York, 1970, p. 243. 35. Cram, S. P., and Brownlee, J. L., J. Gas Chromatog., 353 (1967). 36. Cram, S. and Brownlee, J. L., J. Gas 305 (1968). 37. Cram, S. P., and Brownlee, J. L., J. Gas Chromatog., 313 (1968). 38. Sievers, R. E., and Moshier, R. W., Gas Chromatography of Metal Chelates, Pergamon Press, New York, 1965. 39. Ross, W. D., and Sievers, R. E., Talanta, 87 (1968). 40 .-Ross-, W. D. ,-Sie-vers, R.-E-. ,-and -Whee-ler ,--G. Chern., ]2, 598 (1965). 41. Ross, W. D., and Sievers, R. E., Anal. Chern. Q, 1109 (1969) 42. Scribner, W. G., Smith, B. H., Moshier, R. W., and Sievers, R. E., J. Org. Chern., 1696 (1970). 43. Booher, T. R., M.S. Thesis, University of Florida, Gainesville, Florida, 1969. 44. Shankar, J., India At. Energy Comm., BARC-348, Bhabha At. Resch. Centre, 1968. 45. Lin, T. K., and Yeh, S. J., J. Nucl. Sci. Technol. ( To ky 0), l, 2 89 ( 19 6 6) 46. Torko, J., Energ. Atomtech, 200 (1967). 47. Nath, A., and Nesmeyanov, Radiokhimiya, 125 (1963). 48. Lazzarini, E., J. Inorg. Nucl. Chern., 7 (1967). 49. Gainer, I., and Ponta, A., Rev. Roum. Phys., li, 645 (1968). SO. Gainer, I., and Ponta, A., Rev. Roum. Phys., 13, 887 (1968). 47

PAGE 52

51. Machado, J., Machado, R. M., Vargas, J., Chem. Effects of Nuc1. Transformations, Proc. Vienna, 1964,. 195 (1965). 52. Thomas, V. G., Nath, A., and Shankar, J., USAEC Tech. Rept., NP-12330, p. 415 (1962). 53. Thomas, V. G., Nath, A., and Shankar, J., USAEC Tech. Rept., NP-12330, p. 407 (1962). 54. Shankar, J. Venkateswar1u, K. S., and Lo1, M., J. Inorg. Nuc!. Chem. (1966). 55. Shankar, J., Nath, A., and Thomas, V. G., J. Inorg. Nuc1. Chem. ,30, 1361 (1968). 56. Thomas, V. G., Indian J. 1,247 (1969). 57. Rao, K. A., and Nath, A., Radiochimica Acta, 162 (1966). 58. Gainer, I., and Ponta, A., Rev. Roum. Chim., 13, 401 (1968). 59. Sievers, R. E., Moshier, R. W., and Morris, M. L., Inorg. Chem., 1, 966 (1962). 60. Sievers, R. E., Ponder, B. W., Moshier, R. W., and Morris, M. L., Inorg. Chem., 693 (1963). 61. Ross, W. D., Anal. Chem. 1596 (1963). 62. Ross, W. D., and Wheeler, G., Anal. Chem., 36, 266 (1964). 63. Hill, R. D.; and Gesser, H., J. Gas Chromatog., 1, 10 (1963). 64. Sievers, R. E., Connolly, J. W., and Ross, W. D., J. Gas Chromatog., 241 (1967). 65. Baetzer, A. M., Damron, C., and Budaez, V., A.M.A. Arch. Indust. Health, 136 (1959). 66. Brieger, H., Ztschr. F. Exper. Path. V. Therap. ,21, 393 (1920). -67. Goldman, M., and Karotkin, R. H., Am. ,189, 400 (1935). 68. Major, R. H., Bull. John Hopkins Hosp., 56 (1922). 48

PAGE 53

69. Mancuso, T. F., Industrial Med. and Surgery; "20, 393 (1951). 70. Morris, G. E., A.M.A. Archivis of Dermato1ogy,"2!.", 612 (1958). 71. Pascale, L. R., Wa1dstein, S. S., Engbring, G., Dubin, A., and Szanto, "P. B., J.A.M.A.,"149,1385 (1952) 72. Curran, G. L., J. BioI. 765 (1954). 73. Mertz, W., Roginski, E. E., and Reba, R. C., Amer. J. PhysioI. ,209, 489 (1965). 74. Mertz, W., and Schwarz, K., Measurements of Exocrine and Endocrine Functions of the Pancreas, J. B. Lippincott Company, Philadelphia, Penn., pp. 123-237 (1961). 75. Schroeder, H. A., J. Nutr., 439 (1966). --------7.fJ S B a1 as sa, J. J., and Tip t on I. H. J. Chron. Dis., 941 (1962). 77. Schwarz, K., and Mertz, W., Arch. Biochem. Biophys., 292 (1959). 78. G1insmann, W. H., Feldman, G. J., and Mertz, W., Science, 152, 1243 (1966). 79. G1insmann, W. H., and Mertz, W., Metabolism, 510 (1966). 80. Levine, R. A., Doisy, R. J., and Streeten, D. H. P., Diabetes, 15, 539 (1966). 81. Savory, J., Mushak, P., and Sunderman, F. W., J. Chrornatog. Sci., 2, 674 (1969). 82. Springer, C. S., Meek, D. W., and Sievers," R. E. Inorg. Chern., ., 1105 (1967). 83. Tanaka, N., Shono, T., and Shinra, K., Nippon Kagaku Zasshi, 669 (1968). 84. Berg, E. W., and Truernper, J. T., J. Phys. Chern., 64,482 (1960). 49