Analysis Results for KamLAND PPO Samples




Zelimir Djurcic1, Kevin Lesko2, Andreas Piepke1, Al Smith2 and Bryan Tipton3

1University of Alabama
2Lawrence Berkeley National Laboratory
3California Institute of Technology

2/14/2001
Up-dated 2/22/2001
Up-dated 3/12/2001


Introduction

In order to understand whether the Packard PPO, to be used in KamLAND, is suitable we have analyzed various PPO samples for their content of radioactive impurities using different techniques. For the reactor neutrino measurement we require K, Th and U concentrations of less than 10-10, 10-14 and 10-14 g/g in KamLAND's liquid scintillator. At a a concentration of 1.5 g/l the resulting PPO dilution factor of 520 requires the radio-impurity concentration of above elements not to exceed 5· 10-8 and 5· 10-12 g/g, respectively. A factor 10 better purity would be better to guarantee a negligible PPO contribution. The purification of the PPO by water extraction will remove some of the impurities. The efficiency of our purification is not exactly know at this time. The Borexino group from Munich has measured the efficiency of 3-fold water extraction on PPO by neutron activation analysis (NAA). Their web site gives it as 500, >160 and 70 for K, Th and U, respectively. Assuming that the same purification factors apply to us the PPO, as received, must not contain more than 25· 10-6 g/g K, 800· 10-12 g/g Th and 340· 10-12 g/g U. This should be a cut-off condition. PPO of lesser radio purity should not be used in KamLAND.


Analysis Methods

We used three different methods to determine the K, Th and U content of the PPO:
(A) Low background gamma ray spectroscopy (Counting), performed at Caltech and LBL, measures the concentration of gamma ray emitting isotopes. For the Th and U decay series this method is mainly sensitive to 228Ra and 226Ra. Although the sensitivity is somewhat limited it is the only method able to test a broken equilibrium. The relatively short-lived Ra is only present in a very small chemical concentration. It is un-measurable with the two other methods.

(B) Inductively coupled plasma mass spectroscopy (ICPMS), detects a beam of the isotope of interest. At the concentration levels we are interested in the samples contain a considerable number of target atoms. Coupled with efficient ion sources sensitivities of few ppt can be achieved. The Caltech group has access to such an instrument and used it for the PPO analysis. However, the ICPMS is optimized for aqueous samples which requires chemical treatment of the sample. 235U tracer has been used to demonstrate the overall efficiency of the method. The achievable sensitivity is ultimately limited by the quality of the chemicals used for the preparation of the sample before introduction into the mass spectrometer. This analysis is only sensitive to the long-lived mother isotopes of the decay chains. Our summary also includes data obtained by the Tohoku group in collaboration with a commercial laboratory.

(C) Neutron activation analysis (NAA) relies on nuclear transmutation through capture of thermal neutrons on the meta stable target isotopes into short lived, and hence easy to count, daughter isotopes. The activation is performed in the high flux research reactor of MIT. After activation the samples are shipped to UA, LBL and Caltech for counting with shielded high resolution Ge detectors. The shipping delay limits our sensitivity towards the short-lived activation product of K (42K, T1/2=12.36 h). Sample packing is done in a clean room at UA. No chemical treatment is required before activation. At this point the sensitivity of this technique is limited by the presence of side activities, namely 24Na and 82Br. The Compton background of their respective gamma peaks limits our sensitivity to detect the characteristic gamma lines of the target isotopes. Limits are then inferred by a statistical analysis of the spectra.



Results

Each group has described the details of their analysis in separate documents on the web. In the following we want to summarize all results obtained thus far. Please have a look at the appropriate web site for the analysis details. These sites also report the quantitative analysis results for a variety of stable impurities (side activities):
* UA

* LBL*

* Caltech

The following table summarizes all measurements done with various PPO samples. For ICPMS and NAA, measuring the same quantity, we report the concentration of the radio-isotopes of interest. The group web pages listed above are giving a more detailed summary of the quantitative analysis of various other chemical impurities in the PPO. For direct counting data the measured activities (or limits) have been converted into an equivalent U/Th concentration assuming secular equilibrium. Entries are color-coded according to the definition given above.

Sample Method Institution K [g/g] Th [g/g] U [g/g]
Packard PPO lot 9 NAA Caltech (40± 12)· 10-12 (400± 150)· 10-12
Packard PPO lot 9 scintillator NAA UA/Caltech <100· 10-12 <70· 10-12
Packard PPO lot 9 ICPMS Caltech <500· 10-12 <100· 10-12
Packard PPO lot 9 ICPMS Tohoku 100· 10-9 <100· 10-12 <100· 10-12
Packard PPO lot 9 scintillator ICPMS Caltech <250· 10-12 <125· 10-12
Packard PPO lot 9 scintillator ICPMS Tohoku <7000· 10-9 <4000· 10-12 <133· 10-12

Sample Method Institution K [g/g] Th [g/g] U [g/g]
Packard PPO lot 21-0113 NAA UA (-0.8± 1.4)· 10-6 (240± 100)· 10-12 (-3000± 1200)· 10-12
Packard PPO lot 21-0113 (bag 1) ICPMS Caltech (300± 100)· 10-12 <400· 10-12
Packard PPO lot 21-0113 (bag 1) ICPMS Tohoku 84· 10-9 <100· 10-12 <100· 10-12
Packard PPO lot 21-0113 (bag 1) ICPMS Tohoku 141· 10-9 <200· 10-12 <100· 10-12
Packard PPO lot 21-0113 (bag 1) Counting Caltech <4000· 10-9 <6000· 10-12 <2000· 10-12
Packard PPO lot 21-0113 (bag 2) Counting LBL (-2400± 700)· 10-9 (-800± 1000)· 10-12(early)
(-1500± 1000)· 10-12(late) 		(3800± 2000)· 10-12 (early)
(800± 280)· 10-12 (late)
Packard PPO lot 21-0113 (bag 3) Counting LBL (-1000± 700)· 10-9 (700± 1100)· 10-12(early)
(-1900± 900)· 10-12(late) 		(5800± 2400)· 10-12 (early)
(500± 300)· 10-12 (late)

Sample Method Institution K [g/g] Th [g/g] U [g/g]
Packard PPO lot 21-1011 (purified) NAA UA (-87± 110)· 10-12 (960± 1300)· 10-12
Packard PPO lot 21-1011 (purified) NAA Caltech <1430· 10-9 <1.2· 10-9 <92· 10-12
Packard PPO lot 21-1011 (purified) ICPMS Caltech (70± 30)· 10-12
<200· 10-12 		(50± 30)· 10-12
Packard PPO lot 21-1011 (purified) ICPMS Tohoku 270· 10-9 <200· 10-12 <100· 10-12
Packard PPO lot 21-1011 (purified) Counting Caltech <1500· 10-9 <3000· 10-12 <1000· 10-12

Sample Method Institution K [g/g] Th [g/g] U [g/g]
Packard PPO lot 21-1012 (unpurified) NAA UA (1700± 1000)· 10-9 (125± 110)· 10-12 (-275± 205)· 10-12
Packard PPO lot 21-1012 (unpurified) ICPMS Tohoku 130· 10-9 <200· 10-12 <100· 10-12


Sample Method Institution K [g/g] Th [g/g] U [g/g]
Dojindo PPO NAA Caltech <400· 10-9 (51± 47)· 10-12 <100· 10-12
Dojindo PPO ICPMS Caltech <100· 10-12 <100· 10-12
Dojindo PPO ICPMS Tohoku 67· 10-9 <200· 10-12 <100· 10-12
Aldrich PPO ICPMS Caltech (500± 100)· 10-12 (500± 100)· 10-12
Acros PPO ICPMS Caltech (200± 100)· 10-12 (800± 100)· 10-12
Merck PPO ICPMS Caltech (100± 100)· 10-12 (200± 100)· 10-12
Alfa Aesar PPO ICPMS Caltech (800± 100)· 10-12 (900± 100)· 10-12


Discussion

First we want to point out that above data replaces analysis results reported in January 2001. In this earlier analysis we reported Th/U concentrations of 3 and 15 ppb, respectively. In case of the NAA these readings were due to a small amount of quartz wool used to secure the PPO in the irradiation vial. Our re-analysis reported above identified and eliminated this problem.
We thus conclude the following:
None of the tested Packard PPO samples shows radio-isotope concentrations above the minimal cut-off condition.

Our data shows that Packard and Dojindo PPO offer the best radio-purity among the tested brands.

Thus far our sensitivity is still marginal. Counting of the activated samples continues. Due to the long live time of 233Pa (T1/2=27.0 d), compared to the dominant side activity 82Br (T1/2=35.3 h), the Th-sensitivity should improve with more counting time time. It is interesting to note that the purified Packard PPO shows a factor 200 larger Br content than the regular PPO. The purification procedure must therefore introduce Br into the PPO.
We plan another round of NAA analysis to establish a quantitative measurement of K, Th and U in KamLAND PPO. We plan to achieve this by (a) ashing of the PPO to reduce the Br interference and (b) perhaps TBP extraction of 233Pa and 239Np into TBP after activation.
It would be useful to obtain a 1 kg sample of the production Packard PPO. LBL will melt this sample to bring it into a more compact geometry. A long counting run at the Oroville underground counting facility will then establish the best possible measurements for the Ra-content of our PPO.
This page is maintained by A. Piepke
Last update March 12, 2001