Chromatography Forum

Now I'm analyzing liquid chemicals using ICP-MS like H2SO4, NH4OH, H2O2, HF, DHF, BOE etc.
These liquid chemicals are used in manufacturing of semiconductor wafer.
It's very important to analyze impurities in liquid chemicals because these impurities can occur defects during wafer manufacturing.

Sample's initial concentration are as follows;
H2SO4: 96%, NH4OH:40%, HF:49%, DHF:0.49%, H2O2:30%, BOE(consists of HF + NH4F).

To analyze these sample, I prepared these sample with 10~50fold dilution prior to analysis using ICPMS.
Calibration curve also was made at the same time to quantitate metal impurities.

My questions are as follows;
1] As we know, these samples are different matrix for each.
So, do I need to make calibration curve for each different sample analysis?
To save time, I want to make calibration curve with ultrapure water(like DI, without any metal impurity).
Of course, I know ultrapure water is also different matrix with various samples(H2SO4, NH4OH.....).

Is there anybody know how to get sample concentration from different sample matrix fastly?
This is really neck point for me.
If I make calibration curve for each sample, sample running time is too long to get concentration from various samples.

I need really good comments for my questions from experts in the world.

Best regards.
Li

Now I'm analyzing liquid chemicals using ICP-MS like H2SO4, NH4OH, H2O2, HF, DHF, BOE etc.
These liquid chemicals are used in manufacturing of semiconductor wafer.
It's very important to analyze impurities in liquid chemicals because these impurities can occur defects during wafer manufacturing.

Sample's initial concentration are as follows;
H2SO4: 96%, NH4OH:40%, HF:49%, DHF:0.49%, H2O2:30%, BOE(consists of HF + NH4F).

To analyze these sample, I prepared these sample with 10~50fold dilution prior to analysis using ICPMS.
Calibration curve also was made at the same time to quantitate metal impurities.

My questions are as follows;
1] As we know, these samples are different matrix for each.
So, do I need to make calibration curve for each different sample analysis?
To save time, I want to make calibration curve with ultrapure water(like DI, without any metal impurity).
Of course, I know ultrapure water is also different matrix with various samples(H2SO4, NH4OH.....).

Is there anybody know how to get sample concentration from different sample matrix fastly?
This is really neck point for me.
If I make calibration curve for each sample, sample running time is too long to get concentration from various samples.

I need really good comments for my questions from experts in the world.

Best regards.
Li

Acid concentration makes a difference in the response of analytes on ICPMS. I find that if I make standards in 2%HNO3 then run samples in 4%HNO3 I will see my internal standard areas change by sometimes 30% and if you run 4%HNO3 standard then run a sample that is unacidified water, it can be off by 50%.

When we do samples that are concentrated acids we normally do a 50:1 or 100:1 dilution of the sample into 4%HNO3 so that it is as near the acid concentration of our calibration standards as possible.

The more concentrated your acid, the faster you will erode your cones, so you need to dilute as much as possible for that also. Plus you should not run HF above 1% concentration unless you have a PTFE spray chamber and ceramic center tube as it will quickly destroy your system.

You should definitively use the PFA inert kit for HF, it will also allow you to introduce higher H2SO4 concentrations.

Prepare your calibration curse in 2% ultrapure HNO3. Ultrapure acids can be very expensive, maybe you should consider buying a acid purification/destillation system in order to obtain always fresh destilled and pure acid.

Use adequate internal standards to compensate the matrix effects.

Now I'm analyzing liquid chemicals using ICP-MS like H2SO4, NH4OH, H2O2, HF, DHF, BOE etc.
These liquid chemicals are used in manufacturing of semiconductor wafer.
It's very important to analyze impurities in liquid chemicals because these impurities can occur defects during wafer manufacturing.

Sample's initial concentration are as follows;
H2SO4: 96%, NH4OH:40%, HF:49%, DHF:0.49%, H2O2:30%, BOE(consists of HF + NH4F).

To analyze these sample, I prepared these sample with 10~50fold dilution prior to analysis using ICPMS.
Calibration curve also was made at the same time to quantitate metal impurities.

My questions are as follows;
1] As we know, these samples are different matrix for each.
So, do I need to make calibration curve for each different sample analysis?
To save time, I want to make calibration curve with ultrapure water(like DI, without any metal impurity).
Of course, I know ultrapure water is also different matrix with various samples(H2SO4, NH4OH.....).

Is there anybody know how to get sample concentration from different sample matrix fastly?
This is really neck point for me.
If I make calibration curve for each sample, sample running time is too long to get concentration from various samples.

I need really good comments for my questions from experts in the world.

Best regards.
Li

Acid concentration makes a difference in the response of analytes on ICPMS. I find that if I make standards in 2%HNO3 then run samples in 4%HNO3 I will see my internal standard areas change by sometimes 30% and if you run 4%HNO3 standard then run a sample that is unacidified water, it can be off by 50%.

When we do samples that are concentrated acids we normally do a 50:1 or 100:1 dilution of the sample into 4%HNO3 so that it is as near the acid concentration of our calibration standards as possible.

The more concentrated your acid, the faster you will erode your cones, so you need to dilute as much as possible for that also. Plus you should not run HF above 1% concentration unless you have a PTFE spray chamber and ceramic center tube as it will quickly destroy your system.

Thanks for your reply.
It's good for me to understand.

You should use standard addition, especially when you're dealing with these matrices in ICP-MS. It will allow you to back-calculate the apparent concentration. Remember to use the smallest volume change possible (relatively concentrated standards, microliter volumes) so that you don't materially change the matrix.