Agung,

If your filters are "dirty", that is, have on them traces of what you are sampling for, you will have a problem using your results.

At this point, you cannot state that what you observed is not from the filter itself as best as I know.

You can try and wash your filters, or will need to use better filters (clean, ashless, etc, depending on what you are sampling for and use to sample). It is best to test your filters first before using them for sampling if you are not familiar with the filter.

The blank concentration is always and consistently higher than the sample concentration? Is the weight measurement within the limits of error of your balance. In this case, the probability is that this behavior will show nearly half of the time.

Hi @Sandra, thanks for your answer. Actually, only some sample that have lower mass of ions. I'm not sure about limit of error of the instrument itself, never check it. How to do that?

Hi @Oren thanks also for your answer. As your suggestion, I'm testing with another blank filters at the moment. I am testing 6 blank filter. I will let you know how the result are.

Check the manual of the instrument. If it is not available or there is no information there, you can assume the smallest value measurable by the instrument as the instrumental error.

I have repeated the analysis and the result is the same. However, I run two type of blank filter, including field blank filter and fresh blank filter (the filter that just baked in the oven). If I correct the sample concentration by using the average concentration of field blank filter, almost all sample is in negative value. But if I correct sample concentration by using the average concentration of fresh blank filter, the result is better, only some samples have negative value. Is it fine if I correct the concentration by using fresh blank filter data?

No

The field blank is determining because that determines the contamination that is picked up also by the sample filters

I actually over 40 years of working in the area of filter analysis I have never heard of field-blank filters with a higher loading than sample filters. What can happen is that operators touch the filters with their hands so that they contain NaCl especially from sweat

Hi Harry. Thank you for your answer. I am more confident that field blank concentration should be lower than sample concentration.

I am suspicious of two possible reason why I found negative value of concentration after it was corrected by field blank filter. First, the sensitivity of instrument is not good enough for measuring very low concentration of ion, remember the sample with the negative value has a very clean filter (please see attached photo). Second, the method that I use to extract the filter is not appropriate for very low concentration of sample. Please see attached file describe the methods that I used.

Is the reason I mention above make sense? If yes, is there anyway to improve the methods? Because it is impossible to change the instrument.

Thanks for your answer Sandra. So, could I said that the concentration of sample with negative value actually is lower than smallest concentration of sample because of this instrumental error?

Agung

What is the type of filter you use and what ions do you analyse?

What is the lower detection limit of the instrument compared to that of a solution from the leached filters?

And the attached photo: what is that foil at the left?

I used quartz filter and measured K, NH4, Mg, Na, Ca, Cl, SO4, NO2, NO3, F, and PO3.

I don't have information about LoD, but I have tried to estimate it by using calibration curve and blank, and LoD is higher than the lowest concentration of ions.

The foil is alumunium foil that usually be used to cover the filter.

LoD you can and must measure yourself from a blank solution of highly purified distilled water

I was afraid you used glass fibre

and that the foil was taped to the filter

Harry.. do you mean that I measure just high purity water into IC?

No.. We used quartz filter to catch the aerosol..

Your LoD is determined by the performance of your analyser/IC

That you can test best with the best purified water

Then you should test your leaching solution and compare it with blank filters

(To improve the amount of sampled material you might increase the flow rate and duration of the sampling)

Should I use formula:

Average of purified water + 3*SD of purified water

In order to quantify the LoD..?

Apologies I just now saw your question

in principle yes

but in case the "value" of purified water is high you should distill it and/or further clean it

Agung,

Please check the chromatogram peaks in the IC. If there is fluctuation in the injection pump pressure, your column peak might have shifted giving the wrong area hence wrong concentration value at the end. Then you might need to repeat the analysis. In my opinion, if you have prepared the calibration curve using the same DI water by which you are extracting the solution. Then you need not worry much but if your concentrations are very low then you might need to recalibrate your column using lower concentration also based upon your column type.

Hi Yang Lian , apologized, I just read your answer.

If I see the chromatograph, the baseline shift happens to some samples, and it always happens during my analysis. I thought that this shift occurs because of the low concentration of ions in samples. My college who use the same instrument never found this shift because their sample collected in an urban area in which the filter color is dark. Because of this shift, I already repeat the analysis twice, but the shift is always found for every 6-8 samples analysis during the instrument's run. Is there any other option instead of recalibrating the instrument? How about adding filters extracted in DI water?