Counting Results for Acid used to soak the welded LSU 1.85 MBq AmBe Source




Zelimir Djurcic, Doug Leonard and Andreas Piepke
University of Alabama
(Status 3/19/2004)




Procedure
In a series of acid soak tests, conducted in 2002, LSU's 1.85 MBq AmBe calibration source was found to be either externally contaminated or leaking. Based on these findings the LSU group decided to to seal the source into a welded stainless steel container. This work was done at LSU. After welding the the source was subjected to a pressure test which showed no evidence for leakage. In order to perform a more sensitive analysis the source was soaked at LSU for 31 days (from December 15, 2003 to Jan 15, 2004) in 0.1 molar HNO3. The soak acid was shipped to UA (received January 17, 2004) where it was counted using our low background Ge detector.


Results
The soak acid was transferred into a 60 ml PE counting bottle. The sample was received in a glass bottle. These cannot be directly counted, as glass contains too many radio impurities for a low background measurement. The PE bottle was placed in front of the round face of our Ge detector. In this location the thickness of absorbing materials is minimal giving us the best counting efficiency for the low energy 59 keV gamma radiation emitted in the 241Am decays. The counting efficiency was experimentally determined using a calibrated dilute 241Am solution in the same geometry. Note that the efficiency calibration was made using a glass bottle, eg, having a higher density and thus absorption of the gamma radiation. In our analysis we thus under estimated the counting efficiency, compared the less dense PE bottle ePE > eg.
Let R be the measured counting rate, and b the gamma branching ratio:
AAm = R / (b * ePE) < R / (b * eg)
The quoted results is therefore somewhat conservative.

The soak acid was counted for 5.63 days. Figure 1 shows the resulting energy spectrum, in comparison to background data (measuring time 21.77 days), taken in September 2003. Figure 1 shows the energy range around 59 keV where the 241Am peak would be expected.
Figure 1: Energy spectrum of acid compared to background.

We do not observe a peak at 59 keV in either the acid nor the background data.
In order to quantify the possible 241Am activity present in the acid we fix the peak position and width to their instrumental values and then perform a Chi-square analysis. Figure 2 shows the 241Am peak as determined 2/23/2004 immediately after the acid counting had been concluded.
Figure 2: Energy spectrum of dilute 241Am source.

Peak position and width were now fixed and a three peak structure fitted to the acid data. The result of this fit is displayed in figure 3.
Figure 3: MINUIT fit to acid counting data.

As the counting data showed no evidence for the presence of 241Am we derive a limit of 190 mBq at 90% C.L. in the soak acid.
Based on a solubility analysis, performed in fall 2002, we found that the solubility of 241Am in 0.1 M HNO3 is 160 times larger than in liquid scintillator. We can thus conclude that a scintillator exposure of this source, performed under the same conditions as the acid soak, would result in an 241Am transfer of <1.2 mBq into the scintillator.




This page is maintained by A. Piepke
Last update March 19, 2004