Tungsten (W) as a homologue of seaborgium (Sg) is used to develop a reduction experiment for Sg. Previous off line studies [1] with carrier free amounts of W showed the possibility to reduce W(VI) to W(III) with solid Al at increased temperature (80 ^circC). While W(III) runs through an anion exchange column in diluted HCl/HF as a cation, W(VI) forms anionic oxyfluoro complexes and sticks to the resin. In off line experiments, 4 M HCl/0.01 M HF was used to strip the column to ensure that all activity was recovered.

For the on-line experiments with solid reducing agents, the head of ARCA was modified (see figure). W (mainly ¹⁷⁰W) produced on-line at the Philips Cyclotron of the Paul Scherrer Institut or at the UNILAC at Darmstadt, respectively, was transported to ARCA with a He/KCl gas jet system. After a collection time of 120 s, the KCl spot was dissolved in 333 ml of a mixture of 0.1 M HCl and 0.1 M HF. It was eluted under varied conditions (see below) and monitored by g-spectroscopy.

Fig.1: Modified ARCA head as used in exp. 3 and 4. The Ti head in exp. 1 and 2 contained no PEEK inlays.

To simulate a Sg experiment (which has to be a 1-step experiment due to the short half live) no strip fraction was taken. Instead, reference values were obtained using `empty' reducing columns in which the dead volume was diminished by a piece of PE capillary.

For each experiment are given 4 values: Elution through a) the the empty head (showing the absorption on the head material), b) through the reducing agent (RA) only (showing absorption in both head and RA), c) through AIX without RA (showing the efficiency of the W(VI) separation), d) through both RA and AIX. Using these values (or a known K_d instead of c), respectively), one can calculate the overall yield of W(III) (last column of the table).

Exp.#	head material	reducing agent	resin	% eluted	% reduced
1a	Ti	none	none	39	-
1b	Ti	Al	none	17	-
1c	Ti	none	AIX	5 *	-
1d	Ti	Al	AIX	11 *	10 *
2a	Ti	none	none	64	-
2b	Ti	Al	none	41	-
2c	Ti	none	AIX	7	-
2d	Ti	Al	AIX	31	30
3a	PEEK	none	none	100	-
3b	PEEK	Al	none	85	-
3c	PEEK	none	AIX	10 *	-
3d	PEEK	Al	AIX	67 *	64 *
4a	PEEK	none	none	100	-
4b	PEEK	Zn	none	98	-
4c	PEEK	none	AIX	13	-
4d	PEEK	Zn	AIX	<5	0

Table 1: Yields of eluted W under different conditions. All yields are given relative to a direct catch from the gas jet (=100%). Asterisks indicate that the values were not measured but calculated using known K_d values.

Experiment 1 showed that W absorption on the Ti surface is quite strong, thus resulting in overall reduction yields too low for a Sg experiment. For experiment 2, the HF concentration was risen to 0.3 M, resulting in weaker absorption but in rapid dissolution of the reducing agent Al. Using a PEEK head in experiments 3, the absorption in the head was reduced to a minimum, but the rapid dissolution of Al was still unsatisfying. The weaker RA Zn (Exp.4) showed almost no dissolution, but also no reducing effects. This is remarkable, because the potential difference for the reduction of W(VI) with Zn indicates that the reduction is thermodynamically possible (dE₀<-0.5 V). Obviously, it is kinetically hindered. The lower yield in experiment 4d) with respect to 4c) indicates even a stabilisation of the 6+ state, but the difference is insignificant taking values from different runs as the error range.

The results show, that a reduction of on-line produced W is possible at least to an overall yield of 64%. This would be sufficient for a Sg reduction experiment. The aim of forthcoming experiments is to find a finally useful RA. Possibilties are either varying the particle size of Al (preventing rapid dissolution) or choosing another RA (for example Ti, whose redox potential in HF solutions is between those of Al and Zn).

References
[1] E. Strub et al., Institut für Kernchemie der Universität Mainz, Jahresbericht 1997, p.12