Anal. Chem. 1988, 58, 1261-1263
Table I. Gas Chromatographic Analysis for 3-pL Injections of a Standard Methanol Solution relative std dev, compd
mean concn," ng/pL
%
ethylene glycol dimethyl sulfoxide dimethyl sulfone
7.8 10.7
2.3 3.1 2.2 1.9
phenol 2-octanol 1-octanol
8.3 11.3 8.3 8.3
2.0 2.3
"Six replicate injections (n = 6 ) . solvent is more efficient, which reduces the volume of solvent entering the column as a liquid. With this injector configuation, up to 1.0 pL of the methanol solution could be introduced before solute peak splitting occurred. A massive thermal reservoir was fabricated by brazing the small-diameter sheath to the interior of a larger 1.25-cm-0.d. length of stainless-steel tubing. The interior of the outer 1.25-cm-0.d. tubing was then packed tightly with aluminum foil (Figure 1). Figure 2C shows a chromatogram of a 3-pL hot on-column injection of the methanol solution with this sheath installed in the injector block. This massive thermal reservoir results in a highly efficient vaporization of the sample. The solvent peak is narrow with negligible tailing allowing for the analysis of early eluting solutes (ethylene glycol, dimethyl sulfoxide, and dimethyl sulfone) when compared with the conventional hot on-column injection shown in Figure 2A. Solute peak splitting and distortion have been completely eliminated. Up to 4 p L of the methanol solution has been successfully injected by use of this injector configuration. Similar results were obtained using acetone as the solvent. This injector configuration requires a slow injection technique to eliminate sample blowback into the on-column injector. Rapid injection of solution volumes larger than 0.5 pL resulted in an extremely broad, tailing solvent peak indicating that the pressure pulse produced by the rapid injection caused sample blowback into the injector, despite a carrier gas linear velocity of 30 cm/s. Representative results obtained for 3-bL hot on-column injections of a standard methanol solution using the massive
1261
thermal reservoir injector configuration are shown in Table I. Relative standard deviation values range from 1.9 to 3.1% indicating a high degree of precision for this injection technique. Retention times were also highly reproducible (within 0.05 min). The on-column injection technique described allows for the high-resolution capillary gas chromatographic determination of highly polar, water-soluble organic compounds in methanolic extracts. It results in a 10-fold increase in sensitivity when compared with the packed column technique. A detection limit of 20 pg for dimethyl sulfone in methanol is obtained when 3 p L is injected on-column ( S I N = 5, Hall electrolytic conductivity detector/sulfur mode). Sensitivity of this magnitude is essential for the determination of trace quantities of these polar organics in environmental samples. Registry No. Ethylene glycol, 107-21-1; dimethyl sulfoxide, 67-68-5; dimethyl sulfone, 67-71-0; phenol, 108-95-2; 2-octanol, 123-96-6; 1-octanol, 111-87-5.
LITERATURE CITED (1) "NIOSH Manual of Analytical Methods", 2nd ed.; National Institute for Occupational Safety and Health: Cincinnatl, OH, 1977. (2) Grob, K., Jr.; Neukom, H. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1070, 2 , 12-21. (3) Schomburg, G.; Behiau, H.; Dielman, R.; Weeke, F.; Husmann, H. J. ChrOrtWtOgr. 1977, 142, 87-102. (4) Grob, K.; Grob. K., Jr. J. Chromatogr. 1078. 151. 311-320. (5) Schomburg, G.; Husmann, H.; Rittman, R. J. Chromatogr. 1081, 204, 85-96. (6) Grob, K. HRC CC, J . Hlgh Resolut. Chromatogr. Chromatogr. Commun. 1078, 1 263-267. (7) Grob, K., Jr. J. Chromatogr. 1081, 213, 3-14. (8) Grob, K., Jr. J. Chrometogr. 1982, 237, 15-23. (9) Wang, F. S.; Shanfield, H.; Ziatkis, A. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1082, 5 , 562-564. (IO) Ghaoui, L.; Wang, F. S.; Shanfleld. H.; Ziatkis, A. HRC CC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1083, 6 , 497-500. (11) Wang, F. S.; Shanfiekl, H.; Zlatkis, A. HRCCC, J. High Resolut. Chromatogr. Chromatogr. Commun. 1083, 6 , 471-479. (12) Grob, K., Jr.; Mueller. R. T. J. Chromatogr. 1082, 244, 185-196. (13) Yang, F. HRC CC, J . High Resolut. Chromatogr. Chromatogr. Commun. 1083, 6 , 448-450. (14) Hinshaw, J. V., Jr.; Yang, F. HRC CC, J . Hlgh Resolut. Chromatogr. Chromatogr. Commun. 1983. 6 , 554-559. (15) Fehringer, N.; Walters, S. HRC CC, J . High Resolut. Chromatogr. Chromatogr. Commun. 1085, 8 , 98-100. ~
RECEIVED for review September 16,1985. Accepted December 2, 1985. This work was supported by the National Oceanographic and Atmospheric Administration under Cooperative Agreement NA85-WC-H-06134,
Determination of Technetium-99 in Complex Compounds by Short-Time Instrumental Neutron Activation Analysis Wolf Garner* a n d Hartmut Spies Central Institute of Nuclear Research, Rossendorf, P.O. Box 19, 8051 Dresden, German Democratic Republic The importance of technetium-99m radiopharmaceuticals has stimulated investigations with the long-lived nuclide technetium-99 in order to get more knowledge on basic chemistry of this element. In recent years, a great number of technetium-99 complex compounds have been prepared and characterized (1). The composition of these substances, commonly ascertained by elemental analysis of C, H, and N, is ambiguous in cases where the Tc compound contains, e.g., oxygen or, occasionally, heteroatoms such as phosphorus, selenium, arsenic, or halogens. The knowledge of technetium content would give further 0003-2700/86/0358-126 1$01.50/0
important information on metal to ligand ratio and would help to confirm the real formula of the compound. The aim of this work is to obtain fast and rather precise information on the content of technetium in small amounts of sample. Until now, the technetium content has been either determined by conductometric titration of pertechnetate (2) or by liquid scintillation counting (3). Most of the wide variety of other analytical methods, available for the determination of technetium, have the disadvantage of demanding chemical procedures with the samples before or during the determination process (e.g., gravimetry, spectrophotometry, and 0 1986 American Chemical Society
1262
ANALYTICAL CHEMISTRY, VOL. 58, NO. 6, MAY 1988
Table I. Tc Contents of Compounds M[TcOL2]
no.
M
2 3
AsPh4 NEt4
4
NEt4
formula
L
S-CHZ-CHZ-S S-CH-COOCH, S-L-COOCH, S-C-CN
/I
calcd
no. of
Tc content, % found (mean f std dev)
determinations
CzgHzgOS4AsTc C20H36N09S4TC
14.50 14.96
13.89 f 0.28 14.92 f 0.12 14.89 f 0.10
8 44 22
C16H20N50S4TC
18.83
18.79 f 0.14
20
C24H36N50S4TC
15.52
15.63 f 0.11
22
C16H20N60S4TC
18.83
19.02 f 0.13
34
C16HzoN50SzSezTc
15.98
16.13 f 0.29
16
S-C-CN
5
NBu~
6
NEt4
1
S-C-CN
I1 S-C-CN
I semiconductor
dt
gamma
t (tldt
clock time
live time
spectrometer
~
eAt l i t l d t
12 \eAtl(tldl tl effective time window
Flgure 1. Block diagram of y spectrometer involvlng counting loss and decay correction circuitry.
electrochemical methods) (4). The most selective method for determining technetium is the neutron activation analysis, allowing, in addition, a good reproducibility. As far we know, this method has only been used for determining trace amounts and not for precise analysis of macro amounts (5,6). In this paper we describe a technique, based on neutron activation of technetium-99 and developed for the analysis of technetium in pure complexes.
EXPERIMENTAL SECTION Substances. Tetrabutylammonium technetate(VI1) (1) (C16H3,N04Tc,Tc = 24.42%) as a reliable reference substance was prepared from ammonium technetate(VI1) (Amersham) and tetrabutylammonium bromide. Before use, the substance was recrystallized from methanol/ether giving white needles, mp 241-242 OC. Compounds 2-7 with the general formula M[TcOL2]have been prepared by reaction of technetium(V) gluconate with dithiole ligands as described in ref 7 (Table I). Preparation and analytical data of compound 8 were published in ref 8. Compounds 9-12 were obtained in our recent investigations by the reaction of technetium(V) gluconate with the corresponding ligand. The formulas of 11 and 12 were derived from appropriate data of elemental analysis: Anal. Calcd for 11 (C2,H,,NzO3S2Tc): C, 54.73; H, 3.36; N, 4.91; S, 11.24. Found: C, 54.37; H, 3.31; N, 4.75; S, 11.30. Anal. Calcd for 12 (ClBHl1N,O3S2Tc):C, 45.56; H, 4.04; N, 5.90. Found C, 45.88; H, 3.74; N, 4.66. The formulas of 9 and 10 were proposed on the basis of the Tc value only and have not yet been confirmed by further data. Irradiation and Counting. The reaction
was used. Small solid samples (1-2 mg) were inserted into double polyethylene covers. Freshly prepared tetrabutylammonium technetate(VI1) was used as reference. For irradiation, samples and standards were placed close together within polyethylene vials designed for use in the pneumatic tube systems of the Rossendorf Research Reactor. Preferably, the irradiations were carried out
in the thermal column. This led to moderate sample activities after 20 s at about 10l2 cm-2s-l. The cooling time was 20-30 s. The Ge(Li) y spectrometer had to be run at high and variable pulse rates up to 70000 counts/s, in order to obtain good statistics. The device was equipped with an in-line half-life compensation IHC (9) and a sophisticated counting loss corrector (CLC) (IO), delivering reliable results up to rates of more than 100000 counts/s. Much attention had to be paid to the precise adjustment of this specific instrumentation: The preset digits of the half-life may not differ from the true value by more than 0.5% at a maximum. The CLC is additionally fed with the fast event signal, derived from a trigger level immediately above electronic noise, and with the “busy”signal from a bipolar base line inspector (BLI) observing the output of the main amplifier. The CLC even accounts for nonobservable pulse pile up (two pulses within 200 ns); width and position of the BLI’s voltage window have to be empirically set with respect to the foreseen peak evaluation method. It should be noted that procedures for precise counting at variable rates above 30000 counts/s are rarely discussed in the literature. Not less than 20 000 counts were accumulated in each of the two ‘ q c peaks during 30 s. In order to exclude systematic errors, the sequence of sample and standard was systematically exchanged in successive runs and the group mean values were checked by comparative statistics. The total time consumption of a single Tc determiation was about 6 min. After this period the sample activity is low enough to restart the procedure.
RESULTS AND DISCUSSION Analytical accuracy and reproducibility are demonstrated in Figure 2 and Table I. A very good agreement between expected and found Tc contents has been observed in the case of the “reliable” substances. The mean standard deviation of a single determination turns out to be about 5% (relative). This reflects the present stage of the method. Counting statistics, counting loss, and decay correction are not responsible for this 5% level. In the course of a few hours, up to 50 determinations can easily be carried out, and the uncertainty of the mean value can be restricted within the limits of less than 1% (relative), if necessary. The knowledge of approximate Tc values gives rapid information on composition of Tc complexes with chelate ligands. Figure 3 shows levels of calculated Tc contents of neutral complexes with different Tc/ligand ratios in dependence of molecular weight of the ligand. The presence of coligands, e.g., oxygen or chlorine, which are often found in Tc complexes, has also been considered. It can be concluded from Figure 3, that even a single Tc value may help to assign a special Tc complex to the 1:1, 1:2, or 1:3 Tc/ligand ratio series. This is evident with ligands of a molecular weight exceeding 100, where the different levels become separated from each other. From the known standard deviation and the “forbidden” gaps in Figure 3, the necessary number of determinations can be concluded for each separation problem.
ANALYTICAL CHEMISTRY, VOL. 58, NO. 6, MAY 1986 Table 11. Some Further Compounds, the Composition of Which Is Formulated on the Basis of
no.
ML
ligand monothiodibenzoylmethane
8 9 10
N-salicylidene-2-hydroxyaniline
11 12
N-salicylidene-2-mercaptoaniline N-salicylidenecysteamine
6-mercaptopurine
mean f
std dev
Tc content, % calcd
240 152 213
11.54f 0.08 22.11 f 0.14 20.04 f 0.19
12.12 22.79 18.39
229
17.31 f 0.17 21.30 f 0.14
17.35 20.86
18.99 181
1263
Tc Determination for formula no. of determinations
TcL~ TcOOHL~ TcOHLz TcLZ TcOHL~ TcOHL~
4 8
14
16 16
t
i'l
.=5
0 Tc amtent
t
1%1
-found
ir
15
ML
ca'culated compound 3
13
16
Figure 3. Calculated Tc contents of complexes with different Tc/ligand ratios in dependence of molecular weight M, of the ligand.
Beyond this, the complete knowledge of the composition of any prepared Tc compound may be desirable. In this respect the determination of the metal atom appears to be an important contribution. Our attempts to estimate oxygen and heteroatoms, such as chlorine, selenium, or arsenic, have already been performed by neutron activation analysis and comparable analytical precision for these elements seems promising also.
15 16 Tc content 1%1
Registry No. 1, 16385-58-3; 2, 70177-06-9;3, 74679-89-3;4, 70317-72-5;5, 73465-06-2;6, 77759-64-9;7, 77786-44-8;8, 95860-73-4; 9,99267-43-3;11, 99248-87-0;12, 99248-88-1; "Tc, 14133-76-7.
r
LITERATURE CITED
1,04
1,oa
112
C lnormalizedl Figure 2. Observed distributions of found Tc contents for all analytical results normalized to unity.
The application of this method to a series of hitherto unknown Tc coordination compounds is shown in Table 11. The composition of these complexes was formulated on the basis of T c values only and was confirmed for compounds 8,11,and 12 by subsequent determination of the ligand elements (8). It is evident that reasonable formulations can be proposed for some of these compounds on the basis of the measured Tc contents.
(1) Deutsch, E.; Libson, K.; Jurisson, S . ; Linday, L. F. frog. Inorg. Chem. 1983, 30, 75-139. (2) Mazzochin, 0. A.; Mazzi, U.; Portanova, R.; Traverso, 0. J . Inorg. Nucl. Chem. j974, 36, 3783-3788. (3) Biagini Cingi, M.; Clemente, D. A.; Magon, L.; Mazzi, U. Inorg. Chim. Acta 1975, 13, 47-59. (4) Schwochau, K. Top. C u r . Chem. 1981, 96, 109-147. (5) Foti, S., Delucchi, E., Akamlan. V . Anal. Chlm. Acta 1972, 60, 261-267. Houdek, F.; Obrusnik, I.; Svoboda, K. Radiochem. Radioanal. Lett. 1979, 3 9 , 343-352. Spies, H.; Johannsen, B. Inorg. Chim. Acta 1981, 48, 255-258. Spies, H.; Abram, U.; Uhlemann, E.; Ludwig, E. Inorg. Chim. Acta 1985, 109, L3. Gorner, W., Peters, D., Zschau, J. Nucl. Instrum. Methods 1972, 98, 371-372. (IO) G0mer.W.; Richter, K. H. Proceedlngs of the 10th International Symposlum on Nuclear Electronics, Dresden, April 1980, pp 198-203.
RECEIVED for review March 29,1985. Resubmitted September 3, 1985. Accepted September 3,1985.
