RE: Analytical method

[Franz Schoenhofer <schoenho@via.at>]

At 12:13 02.05.2000 -0500, you wrote:
>Radsafers:
>
>As some of you may remember, I was interviewed on a local TV show here in
>Miami regarding meetings held here by the Tooth Fairy Project people.  This
>list was very useful in providing me with information for that interview.
>Below is one of the documents which was provided to me on the condition that
>the source not be identified.  It is in the first person and I am not the
>one who generated the information.  Hope it provides some information for
>those of you in the analytical world.
>

It is good to hear for the first time something about the procedure. 


Following is a summary of his procedure:
>He typically analyzes one tooth at a time, sample weight approximately 100mg
>(0.1 gram).  He claims that 24% of the weight is attributed to calcium and
>reports his results in pCi/gram Ca.  
>He dries the sample at 110°C, grinds it, and dissolves it in acid.  The acid
>is then evaporated to dryness.  The residual is redissolved and then a
>calcium carbonate precipitation is performed.  

If only a calcium carbonate precipitation is performed, then some
radionuclides which could interfere (Radium- and lead-isotopes for instance
would not be removed). The usual protocols use further purification steps
like a chromate precipitate etc.

The precipitant is
>centrifuged and separated into a liquid scintillation vial.  He then counts
>the sample via scintillation counting.

The details - which might be decisive - are missing.

>The liquid scintillation counter he uses is a Quantalus Model LKB 1220X. 

This counter (correctly spelled as "Quantulus") is the very top model in
ultra low-level liquid scintillation spectrometry, so instrumentation is
excellent. I had in my former lab four of these counters, also doing Sr-90
analysis among many other analysis, so I think my judgement is justified.


 He
>sets his windows at 400 to 1000 channels (200 to 500 keV).  

The optimum window depends on for instance quenching effects and can be
adjusted so to maximise the figure of merit, using instrument software. 

He said below
>400 would give tritium and C-14 interference.  

This is correct. The man obviously understands LSC.


The efficiency he uses is
>based on a Sr-90 source he received from NIST.  He noted that the Sr-90
>source is certified by NIST with a known dpm activity, and since the Y-90 is
>in equilibrium with the Sr-90 he uses a counter efficiency of 1.8 in order
>to account for both Sr and Y.  

I totally disagree to use a value for counting efficiency without having
determined it experimentally. This is best done with an internal standard,
or with a standard which has been prepared the same way as the sample and
has the same quench. Otherwise the certification of the NIST standard will
not help. Nevertheless the counter efficiency of 1.8 is reasonable and the
error introduced by this will not be very big.

>The backgrounds he normally sees is 2.3 cpm,

This is in good agreement with the background of my Quantulus counters.


>with a count time of 400 minutes per sample.  With this protocol he claims
>that his sensitivity (MDA) is less then 1.0 pCi/g Ca.  

I do not have the data at hand with respect to calcium, but we had usually
a LLD of appr. 50 mBq per sample, using a 500 min count and a basis of
three sigma of the background for LLD definition.



He's currently
>analyzing 20 teeth/week.  The results he has seen so far are between 0 and
>10 pCi/g Ca.  He did note that very few are at the upper range but most of
>them are between 0 and 1 pCi/g Ca.
>He told me that he does not know the location from where the teeth come
>(blind study) and that he is working for a "charitable organization".  
>
>On September 1, 1999, I called Dr. Sharma again with some follow-up
>questions.  Again, he was very cordial and open to discussing his procedure.
>He also told me that he's in his 70's and semi-retired.  He was hired as a
>consultant for REMS Inc. (Radiological Environmental Monitoring
>Specialists).
>I wanted to find out more about his counter so I asked him again about the
>window setting that he uses, i.e., 400 to 1000 channels.  He did not know
>what the channel per keV was, 

He cannot know, because it is hardly possible to calibrate channels versus
keV unless you have a well defined constant quench. I know that a
competitor of the Quantulus always refers to keV, but this is not correct.
It is pulse height which is recorded and sorted into the channels and pulse
height depends among other factors heavily on quench. Moreover the
Quantulus has built in multi-channel analysers, through which you record
the pulse height spectra. These spectra can be used to optimize the window
and to detect possible interferences of other radionuclides, especially
alphas, because they would show up as relatively narrow alpha-peaks on top
of the very broad beta spectra. 



but he said that while in the logarithmic
>scale he gets two distinct peaks, one for strontium at approximately channel
>500 and another for yttrium between channels 600-700.  When he sets his
>window from 0 to 1000 channels his background increases to 3.4 cpm.  By
>using 400 to 1000 channel window his background drops to 2.3 cpm, thus
>giving him a lower sensitivity. 
>As stated earlier, he forms a calcium carbonate precipitation.  He then
>repeats this process by washing with weaker dilutions of sodium carbonate
>until he gets a white calcium precipitant.  This is dissolved in pH of 2
>then diluted up to 2.0 ml with 2N acid.  He removes 0.1 ml for Atomic
>Absorption analysis to determine the the calcium content (between 24 and
>27%) and also to determine chemical yield of the analysis from strontium
>nitrate carrier.  The remaining 1.9 ml is mixed with 11.9 ml Ultima Gold AB
>(he said he was not certain of the Ultima Gold volume).   He claims that he
>eliminates all interferences (such as, K-40) because of the many calcium
>carbonate precipitations performed.

This is correct.



>He does not participate in any cross-check program, but he said he would
>like too.  

This seems to be a critical point to me. So it is clear that the laboratory
is not certified and there is no independent control over the quality of
analysis. I mentioned above that the man obviously knows well, what he is
doing and is experienced, but in order to claim the influence of
hypothetical Sr-90 emissions on the health of children by the Tooth Fairy
Project - not to repeat all the other pitfalls - one should use a certified
laboratory. 


He asked me with whom we participate  and asked if I could send
>him information so he can get into a program.  He also asked me something
>very interesting.  He asked if I would be interested in doing split samples
>with him.  He gets a number of molars that he cuts in half for his analysis.
>He would be very eager to send the other half to me for an interlaboratory
>comparison.  

This is a very positive aspect of the person involved that he wants to
control his results, but no data are in any way legal binding, if the
laboratory is not certified, for which participation in intercomparison
exercises is compulsary.

This reminds me on my own philosophy for measurements: I have always
refused to do blind measurements, to simply deliver data and having others
evaluate the results. Even if the laboratory would be certified, the
evaluation of data is obviously done by amateurs, who have started this
project because they want a certain result to be "confirmed". I do not
think that any certification could prevent this.


This is a first comment in a hurry, because I will be away from Vienna the
next days and it is quite late. 

Franz


Franz Schoenhofer
Habicherg. 31/7
A-1160 Vienna
Austria
Tel.: +43-1-495 53 08
Fax.: same number
mobile phone: +43-664-338 0 333
e-mail: schoenho@via.at


Office:
Hofrat Dr. Franz Schoenhofer
Federal Ministry of Agriculture, Forestry, Environment and Water Management
Radiation Protection Department (BMLFUW I/8 U)
Radetzkystr. 2
A-1031 Vienna
AUSTRIA

phone: -43-1-71172-4458
fax: -43-1-7122331

************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html