Tags:
1. Mar 11, 2017

### muonneutrino91

I am computing limit of detection of zinc of an electrolyte. Using standard additions method, I have obtained positive x intercepts for 6/8 samples all very close to zero.
Positive x-intercepts imply final concentration values to be negative. if x-intercept taken to be zero, the concentration is zero likewise.
I have read that one can just take the modulus as the concentration value determined, but I am sure that this is only on the assumption that the x-intercept is negative.

Is my problem then that the amount of zinc in my electrolyte to begin with is less than the smallest quantifiable amount? Even so, I am not sure how to tackle this

y = mx + c ;
y = signal (nA), x = concentration (ppm);

Attempt has been made taking modulus of everything, but as I say positive this isn't right thing to do. Only other possibility can foresee is taking intercepts to be zero, giving a zero concentration in the electrolyte to begin with.

2. Mar 11, 2017

### John Park

If I understand you, yes you have a problem. Do the positive x-intercepts look "real" (greater than the probable errors) and are your lines reasonably straight in those cases? If they are, you can't say much about those concentrations.

You didn't say how you were measuring the zinc concentrations; if the positive intercepts are outside the likely errors, I'd wonder about some sort of interference in the analysis--it looks as though something is removing zinc from the solution, or preventing it from being measured, or is producing some other effect that looks like that.

3. Mar 13, 2017

### muonneutrino91

thank you for the response John, very much appreciated.
the lines are all reasonably straight. In those cases of positive intercept, they have values of order 10ppb at most with axis being ppm, so they are very close to zero. Likewise, those with negative intercepts are also very close, but yield ppm LODs.
the electrolyte is 1M KCl for which zinc has a very good response vs SCE. Using diff pulse ASV with hdme.
measurement is 3x standard addition of 10ul Zn (1000ppm) to 20ml of KCl. I can see from the initial KCl measures that there is a very very small amount of metals in it.
could my amount in the standard additions possibly be too much, causing too steep a gradient?

4. Mar 13, 2017

### John Park

If you could work with a tenth, or less, of your present additions, I think it would be trying. Unfortunately I haven't been in a lab for ages, so I can't really guess at anything that might cause problems with your analytical methods, but are you assuming the zinc is completely Zn++? Could there be any impurities that would complex with it or chelate it?

5. Mar 14, 2017

### Staff: Mentor

How is the LOD defined and how is it calculated? I have a feeling you are sidetracked by an irrelevant fact. In real world no experimental curve goes exactly through zero, unless forced to do so.