Greetings, Fatimah Dinan Qonitan

Long answer short - I do not believe you can convert the calculated Wt% (or At% for that matter) into concentration.

Attached picture shows the results obtained using FP Method. This method relies on software specific assumptions that require tuning. Even then, the calculated Wt% are calculated assuming that the only elements present in the samples are the ones being detected by the spectrometer. It is quite suitable for qual-quant analysis or for determination of the composition of some solid samples, but not for a more complex analysis. To my understanding, there is no way to use the given Wt%, unless you can find a way to calculate the mass ratio of non-carbon to carbon pollutants in the sample.

In your case, some external calibration may be required. Off the top of my head, I would suggest creating a number of samples containing known amounts of particulates (both relative and absolute. It is best if the total volume of the particulates is the same, as to make sure that the analysed volume stays the same). You then can measure the XRay Fluorescence of the samples, but take note not of the calculated Wt%, but of the second column, marked Net. My best guess, given that I am unfamiliar with the output of your particular programme, is that the Net column shows the total measured signal for a given analytical line.

You can then find the dependance between the signal for a given element and its relative amount in the sample, which can be used to determine that amount in the sample of unknown compositon. Given the mass of the sample you can then calculate the mass of the element, and given the volume of air filtered in the paper, you can calculate the concentration in ng/m^3.

Another approach would be to add a small known amount of an element to the studied sample and use the change in the Net column, calculate the amounts of the element - Standard Addition Method.

Dear Dr. Fatimah Dinan Qonitan , yes I agree with you in that the the organic and elemental carbon compounds was excluded as the The XRF technique has proved to be a powerful tool for only the elemental analysis and also in agreement with you in that the results of elemental pollution analysis using XRF, are usually reported in ng/m³ or μg/m3. However, the XRF analysis results you attached, seemed to me that the percentages of elements are far from the common pollution units. In any case, if you are interested in converting the units from %weight to concentration units please follow the coming simplified steps:

1-Assume 1 gram of the sample as a basis.

2- Calculate the weight of each element = (1 gm sample x weight % of element)/100

3-Now calculate the volume of each element in cm3 unit as dividing the weight obtained from step 2 by the density of that element in gm/cm3 units.

4-Apply the calculation in step 3 on all element shown on the table you attached, then added in sum to get the total volume of all elements in cm3 units

5-Divided each element weight calculated in step 2 by the total volume obtained from step 4 to get its concentration in gm/cm 3 units units

6- Finally, covert the units from gm/cm3 to ng/m3 or μg/m3

Example:

weight of Na = (1 gm x 15.669%)/100=0.15669 gm

volume = 0.15669 gm / density of Na= o.15669 gm / 0.97 gm.cm-3 =0.1614 cm3

the total volume of all elements I calculated roughly = 0.791 cm3

Then ,the concentration of Na = 0.15669 gm/0.791 cm3 = 0.198 gm/cm3

Best regards

Dear Maxim Bakhmetev ,

Thank you for your detailed and insightful response—I truly appreciate it. You’ve perfectly captured my confusion, and your explanation of the assumptions behind the FP method and the limitations of using Wt% for direct concentration calculations has clarified a lot.

Since you seem familiar with the software, do you think it has the capability to estimate the percentage of undetected elements? Additionally, I do have data on the carbon mass, and I’m wondering—could we reasonably assume that, aside from the detected elements, the remaining portion of the particulate mass consists entirely of carbon? In other words, could we approximate that Carbon Components + Metals Detected by XRF = 100% of the Particulate Mass?

Your suggestion regarding external calibration and using the Net column instead of Wt% makes a lot of sense. The Standard Addition Method also seems like a viable approach, though obtaining a Standard Reference Material for particulate matter might be a challenge for us. I’ll explore these methods further and see how they can be applied to my dataset.

Thanks again for your time and for sharing your expertise—I truly appreciate it!

Best regards, Dinan

Dear Dr. Mohammed Hussein j. H. Al'Atia

Thank you for your response and for outlining the calculation steps. I really appreciate your explanation, though I believe there might still be some discrepancies in applying this approach directly to atmospheric particulate matter samples because of the total %Weight of elements in a particulate matter samples couldn't be 100%. I’ll review your method further and see how it aligns with my dataset.

Best regards,

Dinan

Dear Fatimah Dinan Qonitan

I believe the method doesn't allow for estimation of undetected elements, the best you can do is to add the possible amounts into the calculations if your software allows for changes in measurement methods (in Shimadzu software they are called analytical groups, but it may vary between manufacturers), but that would require you to know all the necessary data for the assumption.

As for assuming the rest of the mass to be carbon - only if it makes sense for the object you are analysing. I would assume that some ligther elements (oxygen, nitrogen) might be present, but if they are not - it seems to be a reasonable assumption to make.

Best regards

Respected Fatimah!

HOPE YOU SEE THIS!!

I am currently facing the same concern. I am working on PM and I am using XRF to analyze the elements but I need to know the elemental concentrations but XRF gives results in Wt%, and I have not come across any article or research paper that has shown the solution to this, and they always report the values in ng/m3.

Can you help me out? Were you able to come to a solution for this?

I will be very grateful to you