DRO high recoveries but no peaks

[Gugu]

Hi

Okay my head is heating up. I calibrated the DRO method (intergrated method) using Diesel for my standards, with values of 10, 20, 50, 100 and 200 ppm respectively. My calibration curve came out very nice and straight with a coeff. det. of 0.988 (I know it can better).

Now, I ran the same standards as samples and recovered half less than the actual quantity. I then carried on to dilute my 10 ppm standard to make it a 1 ppm and I recovered 78.856 ppm instead but there are no peaks on my chromatogram. I prepared another standard of the same concentration and it still yielded high results. I then took the same standard and ran it again but included a divisor of 50 in the acquisition to see what it would be and I got approximately 1.5 ppm.

I am really getting confused now. Earlier this morning I took a 2000ug/mlL Restek standard and injected 10uL into a 10 mL volumetric flask and yielded 3267 mg/L and the peaks were clear but with the diesel standard that yields high results there are no peaks at all. The baseline is not quite as high but is a bit curved (like a smile) at the beginning and end of the integration.

Please help.

Kind regards
Gugu

[Hornet]

It seems you are very confused about this analysis.
Which method are you using? EPA, ISO, ASTM, etc? What type of extraction are you doing?
I'm assuming you are using a GC/FID as instrument, the fact that your baseline "smiles" makes me think you didn't even use baseline subtraction, which leads you to a poor linearity (no 0.988 is not good), poor recoveries which aren't really recoveries but chromatograms with no sense which leads to wrong integrations.

Assuming your instrument is set and ok:
- no leaks
- Correct column and ferrule since you are reaching high temperatures with this method
- good splitless liner if using splitless
- new septa

Before you start to inject standard solution and samples you have to make a baseline.
The baseline is a chromatogram profile aquired when the instrument is conditioned and clear, just by injecting solvent.
Depending on your software, this chromatogram has to be subtracted to any other chromatogram you'll aquire that day.

To be a good baseline it has to be low of signal, no strange peaks, when the bleed starts to raise the signal it must stabilize when you reach the final, highest isothermal stage of your temperature program. It must be flat in the last part.

[MCAbraha]

I'm not sure your setup.. but I used to analyze DRO via GC/FID and would note two things;

As stated before:
1) Baseline subtraction from a "Solvent Blank" analyzed after any CCV/ICV which would subtract from any QC/Sample area integrated before reporting. This is likely your issue since any sort of baseline integrated will give you some sort of area value.

Expanding upon this, our GC/FID had an ENORMOUS solvent front.. which was likely because of Dichloromethane. We did manual integrations for DRO so it was very important to get a good "zoom" on the entire chromtogram and to preform the manual integration from the same start/end points. The "height" (for lack of a better term) in which you integrate the chromtogram from, on the Y-axis, made an enormous different in recovery values.

2) I'm not sure what you are using to inject.. but our autosampler had issues whenever we would install a new syringe. The syringe would need to run through a decent amount of injections to "break in" and then would be more consistent with injections.

Also, we used a vial with a thicker PTFE septa and if you were to screw the caps on too tightly, it would pressurize the septa and make it harder for the needle to inject and draw from the sample vial.

Needless to say we've moved on from DRO so it has been a year or two since I've worked with them. I did not enjoy working with them much... lol

[James_Ball]

When I was doing DRO we never used baseline subtraction(didn't have it on the old versions of Chemstation) but we did always do a manual integration because most software seems to do a terrible job of finding the proper baseline for this type of sample.

Currently I am doing GRO by GC/MS and have similar problems with needing to manually integrate the baseline because the integrator always tries to integrate above the baseline.

You also need a column that gives very low bleed at the highest temperature of the oven program to make low concentration standards report correctly.

[Gugu]

Hi guys

I here what you guys are saying. I think I can relate a lot to what James_Bell has shared. I am also using a GC/MS and whilst going through each step of my method setup with a colleague, it turned out that the integration is the main problem and the column is also not doing a great job at separating oils from the diesel range organics which was indicated by a hump that appears within the first 5 mins of the analysis resulting in an overestimation of the results.

I will run a high standard and clearly label my peaks just so I can be able to properly adjust the integration and then will go further on to remove some quant ions (specifically 41 and 43) , which also seem to be adding a whole lot of noise and interfering with peak separation.

I hope this will work, what else can be suggested?

Regards
Gugu

[MCAbraha]

Agreed, James. We eventually moved away from the baseline subtraction and just went with normal integrations but eventually dropped the method altogether. But, yes, we ALWAYS had to manually integrate. The starting point for your integration makes a world of difference in your response value.

I always felt the baseline subtraction was the best way (We used Empower 2 or 3) because I would notice on Method Blanks that any sort of wavi-ness in baseline or hump/interferences would often cause a false positive. We would run the 50PPM Diesel standard and a solvent blank and then integrate the baseline and that value would be subtracted from the subsequent analyses within the CCV bracket.

Best of luck!

[Gugu]

Hi

Well I decreased my integration window and removed those 41 and 43 quant. ions. I am nolonger getting the oils range organics eluting simultaneously with the diesel range organics which is one problem solved. My peaks are also very clear now and more prominent. Though, I am still picking up higher than expected readings. I also noticed something quite unfamiliar when changing my inlet liner, it has water in it. What could be the course of that and would it however have an effect on this whole problem I am facing?

Gugu