Peak areas inconsistent, but often following a pattern..

[sirpatty79]

Hello again guys and girls,

I am having more peak area RSD problems. If I make successive injections of the same sample, my areas will often go up a little (2-5%) each time in a definite pattern.

I am trying to determine the concentration of an organophosphonate ester (bp 180-200 deg. C) in the 1mg/mL range in solutions of H2O, EtOH, DCM, and hexanes. The problem has occured with all solvents.

I have replaced the glass wool, tried different columns of the same type, cleaned my liner with acetone, DCM, etc. I have also tried decreasing my split ratio as well as cutting and replacing the ferrules on both columns.

If I remember correctly, I read something not long ago about reproducibility problems due to analyte collecting in the split vent and bleeding back out gradually. I'm thinking that either this is the problem or that there are active sites somewhere (liner or column) that are affecting elution amounts of my analyte.

One question that I have is -- what precisely does activity mean in this context? Just highly polar species or what? I ask because I think that injecting something guaranteed to not cause activity problems and observing the results would be a good troubleshooting step. Something along the lines of tetradecane maybe?

The area for 1mg/mL of my analyte also changes drastically when I use water compared to, say, DCM. Water gives areas around 30000, while DCM gives over 80000. We must deal with high concentrations and found our old split ratio (100:1) caused reproducibility problems.....though I suppose they could have been caused by the issue(s) described above.

I just wanted to bounce this off of you guys to see if anyone has had this experience and am super appreciative of any help you can give. Sorry for the novel and thank you.

Conditions/gear are as follows:

Shimadzu GC-17A chromatograph w/ JW Scientific DB-5 column (30m, .25 mm, .25 um). I am using a plunger-in-barrel 10uL syringe and making manual .5 uL injections.

Injector: 250 C; 120 kPa; helium carrier w/ total flow rate 31 mL/min; column flow rate 1.28 mL/min; avg. linear velocity 32.1 cm/sec; split ratio of 20:1. My FID is at 300 degrees w/ manufacturer-recommended flow rates.

Temp. Program: 100 C for 1 min, then a 40 deg. C/min. ramp up to 280 deg. C, then hold for 1 minute. My analyte elutes out just after 3.9 minutes.

[Don_Hilton]

Oganophosphate esters - it's like the kinds of problems the pesticides people have. Your first few injections condition the column. (You leave analyte bound up with active sites - and the active sites bound up with analyte.) A quick and simple description of activity: Activity is some kind of chemistry that goes on in the column that causes some of your compund to be retained on the column or in the inlet and to be lost from the analysis. And, the ugly thing is that the compund eventually comes free from the active site (not that you will ever see it), leaving the active site able to eat more analytes.

Some ideas for you:

If you are planning to measure the organophosphates in a complex matrix, make the standards in matrix as well. This way with lots of things for the active sites to "eat," the analyte will survive and you will get good analysis.

Plan B: If you are trying to analyze a relatively clean sample, then you can add somethign like a polyhydroxy compound to your standards and samples that will tie up active sites. This can improve RSD's. You may want to check the literature on this one. I know it's around - but It's been a while. Again, this is a trick I learned from pesticides folks.

Plan C: Take a very high level standard and make several injections before starting to work. This will tie up active sites. If you need to take this strategy, watch out on the analysis of low level samples. As the active sites come clear, your sensitivity will suffer. You will need to be sure that you understand the rate at which you need to recondition the column to ensure adequate reproducability.

And if all else fails, quantitate with 13C labeled standards for each compund. Activity will eat 12C and 13C compounds equally.

[freemab]

Where to start? I can't tell from the info. you provide whether you're doing things just fine or are making errors. I am not familiar with your GC, having "grown up" on HP/Agilent GCs, so some of what I say may not apply.

First of all, I've never heard of contamination from the split vent, but contamination from the carrier gas supply tube to the inlet is well known. What happens is you inject your sample, it vaporizes so quickly that the carrier flow doesn't carry it all away downstream. Instead, you get a pressure spike that puffs sample into the carrier gas inlet line. This is NOT common when using a split injection, but with a split of only 20:1, it may be feasible. However, an excessive pressure spike can mess up your split ratio, and this would be especially apparent when changing solvents (consistent with your observation of area vs solvent). A pressure spike, especially with a slow injection, can blow the solvent up past the piston of the syringe. Agilent GCs use a very rapid injection to avoid this. (BTW, don't forget to replace your septum as needed, and to check for leaks at the detector & head of the column. My motto, "A leak in GC can cause ANY symptom."

To diagnose this, determine the volume of your inlet liner (above any severe restrictions, like the little cup in a "J-tube" liner, but excluding simple glass wool). Now calculate the vapor volume of your injected solvent: PV=nRT. Water expands more than organic solvents because of it's small size, so this effect is more pronounced with water. With NO split, your vapor volume must not exceed the volume of the inlet liner. With a split, it may exceed the liner, and it's a judgment call by how much. With 100:1 split, I don't think I've ever seen this problem.

If you conclude this is your problem, you can fix it by using a larger-volume inlet liner, using a smaller injection, using a greater split ratio, using a lower inlet temperature, using a heavier solvent, using a liner that does not have severe restriction (a packed J-tube being the worst). (I recommend a liner of ~1 mL volume containing a small plug of silanized glass wool. This has the effect of moderating the pressure pulse on injection.)

In particular, I'd like to mention the inlet temperature issue. Some chromatographers operate under the misimpression that the inlet temperature must exceed the BP of the solvent. What is actually necessary is that the inlet temperature be high enough to vaporize the solvent and the solutes. Often a fairly low inlet temperature will work better than a higher one.

A couple other considerations: A 10-uL syringe may not be the best choice for a 1/2 uL injection. DCM can be a difficult solvent to work with. When pipetting DCM with an Eppendorf-type pipettor, it is necessary first to fill the air space in the pipettor with DCM vapor. If it were possible to do this for you GC syringe using some sophisticated injector program, it might be worth a shot. I concluded that any solvent more volatile than methanol is not suitable for quantitative GC; but that conclusion is subject to change should better equipment allow it.

An internal standard might help, if the problem is simply consistency in the amount of sample being injected.