LOD - related question

Aalves

Hi everyone,
I was wondering if someone could help me on how to interpret the LOD, some of the analyses have the LOD field as "inf" while others have the data as negative value. For the negative values, I understand that means the data was reported below LOD but how about the 'inf" ones?

Thank you so much

amzipkin

Aalves I have had a similar problem in the past and again recently. My understanding is that this error occurs when the calculated external sensitivity for a particular channel and selection is negative. If you go to the Selections tab at the top of the Data module and then open up the Information for the specific selections that have "inf" instead of an LOD concentration in the Results, you should see that those selections have negative sensitivities for the channels with the "inf" error". Unfortunately, I do not know how to resolve this problem. Changing the baseline modeling from Spline_Autosmooth to one of the other options might help, but I haven't found it to do so consistently. @Bence Do you have recommendations?

Bence

Hi Aalves and amzipkin

Very sorry for the delay in getting back to this. "Inf" LODS are where there is no sensitivity for the element in question. This comes about when there is no signal for reference materials for the channel in question.

We'll look at adding an alert that this is the case, but it boils down to the LOD equation and count rates.

If you have any questions, please let us know.

Kind regards,
Bence

Khadra

Hi @Bence ,

As you can see in the attached file, the LOD values are displaying as "inf" for all elements across all my analyses, and I'm not sure what the issue could be.

Do you have recommendations?

Thanks in advance!

Khadra

Bence

Hi Khadra,

In this case, it would suggest that you have sensitivity for these channels of 0 (according to the Longerich et al equation, but similarly for Howell and Pettke LODS):

LOD = 3 x sd(b) x sqrt(1/bkpoints + 1/samppoints) / S

where S is the sensitivity.

It is explained in more detail here: https://iolite.xyz/docs/user_interface/user_interface.html#detection-limits

My guess is that you don't have Na values for your ref mats in the example you posted?

Kind regards,
Bence

Khadra

Hi @Bence

Unfortunately, the issue occurs across all my channels, and I'm unsure of its origin. Would it be possible for me to send you my Iolite file so you can help identify the source of the problem?

Kind regards

Khadra

Bence

Yes, please send it to our support email address.

Kind regards,
Bence

Khadra

@Bence I’ve sent the Iolite file. Thank you so much for your help.

Kind regards,

Khadra

Bence

Hi Khadra,

Sorry, I missed a key piece of information in your original post and confused things a bit. The channels you're looking at there are the _CPS channels. These are the baseline subtracted counts values for each channel and each selection. These do not have LOD values as only concentration (i.e. _ppm) channels have a valid LOD value. iolite is leaving an inf value there because it can't calculate the LOD, but really it should leave these blank, so we'll change that. But in the meantime, you can ignore these 'inf' values for the _CPS channels, and focus on the _ppm channels instead.

I hope that helps.

Kind regards,
Bence

Khadra

Hi @Bence

Thank you so much for the clarification, It's helped me a lot!

Kind regards,

Khadra