Analysis of U/Th spiked Sample UA30, activated at MIT 3/21/2001 22:26 for 5 hours

Zelimir Djurcic and Andreas Piepke
University of Alabama
(Status 4/26/2001)

Two of the Packard PPO samples (batch 21-634) have been spiked by Bryan Tipton at Caltech with organic U and Th compounds. These samples CIT30 and UA30 have been activated at the MIT reactor together with the other samples. UA30 has been activated for 5 hours ending 3/21/2001 22:26 (CST). It was used to determine the retention of ²³³Pa and ²³⁹Np in ion exchange extraction. We used small (2 ml) Eichrom TRU Resin ion exchange columns, supplied by David Glasgow of Oak Ridge National Laboratory. David had worked out a procedure to dissolve Np in 8 normal HNO₃. He then demonstrated, using alpha counting, that the Np is retained with more than 95% efficiency in these columns. In our experiments, described below, we followed the procedures outlined by David.
The ²³³Pa and ²³⁹Np activities were determined by means of gamma ray counting after receipt of the samples. All activities reported are corrected to 3/24/2001 10:28 (CST), the time of our first measurement. According to Bryan Tipton sample UA30 contained:

0.438 g of Packard PPO
2.7 µg of Th
6.4 µg of U

The following activities were determined for sample UA30 before it was subjected to any chemical treatment. This measurement can hence serve as the normalization of the efficiencies to be determined. All sample vials were externally cleaned by etching after receipt.The sample vials were packed in secondary containment bags to avoid direct handling at the reactor.

²³³Pa: 1943 ±25^stat ±194^syst Bq
²³⁹Np: 22750 ±37^stat ±2275^syst Bq

On 3/25/2001 we extracted the Pa/Np activity from the sample using the following procedure:
for the extraction of the PPO, contained in a sealed plastic container (which was cut open to perform the extraction), we used 15 ml 8 molar HNO₃, pre-heated to 68-75 deg Celsius. At these temperatures the PPO melts and is miscible with the acid. The mixture was kept for 30 min at this temperature on a hot plate. The mixture was then allowed to cool down. As soon as the temperature of the mixture goes below the PPO melting point it precipitates. The liquid was then poured into the ion exchange column, which had been activated with 10 ml of high purity 8 molar HNO₃. The remaining PPO slurry was collected and counted to determine the effectiveness of the first acid wash. It was determined (by counting) that approximately 550 Bq of ²³³Pa (or 28%) and 2500 Bq (or 10%) of ²³⁹Np was still left. A second extraction was then performed.
A total of 20 ml of acid was poured through the ion exchange column. This acid was collected in small plastic vials and the amount of ²³³Pa and ²³⁹Np contained in the discarded acid was again determined by counting to provide a consistency check of the extraction efficiency.
In a last step the column was flushed with de-ionized water in an attempt to elute the target activity from the column. David had found quantitative transfer of Np into the water. However, this was not observed here. After three washes we found only (in the following only statistical errors are quoted. We estimate the systematic error to be on the order of ±10%) 224±9.2 Bq of ²³³Pa and 10135±44 Bq of ²³⁹Np in the water.
Counting of the ion exchange column alone yielded 1260±5 Bq of ²³³Pa and 13800±30 Bq of ²³⁹Np in the columns. Counting of the columns together with the water (as done for the PPO sample) yielded 1440±13 Bq of ²³³Pa and 22640±27 Bq of ²³⁹Np. Note that the activities measured in the water and in the columns separately add up reasonably well to the numbers obtained counting water and columns together.

This finding of a high retention efficiency is confirmed by the activities measured for the discarded acid after it went through the column. This data gives us a direct cross check on the inefficiency of the column itself. We measured 20±23 Bq of ²³³Pa and 629±27 Bq of ²³⁹Np in the acid plus 352±14 Bq of ²³³Pa and 3130±26 Bq of ²³⁹Np in the PPO left after the second acid digestion.

Using above data we determine the following efficiencies and in inefficiencies for the extraction by ion exchange itself:

Efficiency:

²³³Pa: 0.74 ±0.01^stat ±0.11^syst
²³⁹Np: 0.995 ±0.002^stat ±0.14^syst
Inefficiency:

²³³Pa: 0.19 ±0.014^stat ±0.027^syst
²³⁹Np: 0.165 ±0.0017^stat ±0.027^syst

The measurements of the initial activity and the inefficiency involved the same counting geometry. We hence consider the inefficiency to be the better quantity to deal with. The Spread of above efficiencies are understood on the basis of the systematic errors quoted. These are mainly caused by the fact that the counting geometries do slightly vary from run to run.

As we filled the PPO-acid mixture into the column without prior filtering we removed the PPO residue left in the neck of the column. All data shown above was obtained with this residue present in the neck. However, as it was removed for the real PPO sample we need to quantify its contribution to the inefficiency. Counting of the residue (removed) alone yielded: 11.8±1.3 Bq of ²³³Pa and 142.4±12.6 Bq of ²³⁹Np. Before removal of the PPO residue we found the following activities left in the column (without the wash water): 1260±6.1 Bq of ²³³Pa and 15140±40 Bq of ²³⁹Np. Counting the column after removal of the PPO yielded 1186±14 Bq of ²³³Pa and 14090±200 Bq of ²³⁹Np. The retention efficiency is derived from the ratio of column activities after PPO removal divided by that determined before removal, it is: 0.94±0.01 for ²³³Pa and 0.93±0.01 for ²³⁹Np. The inefficiency as determined from the measurement of the discarded PPO is: 0.01±0.001 for both ²³³Pa and ²³⁹Np.
As identical counting geometries could be used in the measurement of the efficiency 0.93 ±0.01^stat ±0.13^syst as will be used to quantify the removal of the PPO residue from the column.

Retention of ²³³Pa and ²³⁹Np in ion exchange:

In view of the similar inefficiencies measured for the ion exchange itself we will use one unified efficiency for both species. Based on the analysis presented here we derive the following efficiency: 0.75±0.15

This page is maintained by A. Piepke
Last update April 27, 2001