2,4-dinitrophenol and other little princesses

[MichaelVW]

Howdy,

I'm running 8270 by 6890/5975, using the average response factor for my cal curve. My list is about 100 compounds long, and 98 or 99 of them cailbrate pretty nicely. 2,4-dinitrophenol always seems to show an increasing response factor with concentration, even when the column and consumables are all new. My response factors are all above the reccommended minimum. Anyone else see this or know why it happens?

The method says that 2,4-dinitrophenol is (along with a few others) subject to "erratic chromatographic behavior" (whatever that means) especially if there is high boiling material contaminating the system. Maybe this is what they're getting at.

The Chemstation software gives me the option to use a linear or quadratic regression rather than using the average response factor... I guess I could just do that?

[sav]

I guess I could just do that?

Yes, the quadratic regression must help.

[Steve Reimer]

I think by now you have seen what erratic chromatographic behavior is.
Nonlinear response is common for compounds that react with the system in some way. Especially near the detection limit.

[Yama001]

I love the title of this post .

The phenols and anilines are pretty active and are often absorbed by the metal surfaces in the inlet (sometime the glass as well, chromatography suppliers do a great bisiness on the myths we dream up to handle this). The method often requires horribly abusing the instrument with samples that should be much more carefully handled.

A quadratic can work, although it is often abused. A big problem comes up when you process a low level QC sample (LFB for example). The curve often will not work for values near the reporting limit. This is why some labs resort to pressure pulse injections to minimize the time spent in the inlet, though this has drawbacks.

Verifying the math on a quadratic curve gets a little sticky as well, particularly for and internal standard method where the reponse ratios are plotted verse the mass ratios.

[MichaelVW]

Hmm, okay. Well I'm glad I'm not crazy; I half expected to hear "what? Everyone gets a beautiful linear curve for these compounds, you must just really suck at installing septa!" or something. One of the more difficult things about being a newbie at this is determining what's realistically achieveable.

[whdees]

2 small pieces of advice; don't use silanized quartz wool in your liner. Even the best introduces active sites. I use cyclosplitters myself. They are quite effective and versatile. And, gold inlet seals are inert at room temperature, but hot gold can be reactive. Use silcosteel seals... and yes, 2,4-dinitrophenol is a big pain in the rear, as are the dinitrotoluene twins...

good luck!

[James_Ball]

2 small pieces of advice; don't use silanized quartz wool in your liner. Even the best introduces active sites. I use cyclosplitters myself. They are quite effective and versatile. And, gold inlet seals are inert at room temperature, but hot gold can be reactive. Use silcosteel seals... and yes, 2,4-dinitrophenol is a big pain in the rear, as are the dinitrotoluene twins...

good luck!

The only way I got glass wool to work was to put it in the liner first, then silanize it with the liner before use. That way you get all the places where the wool broke when you inserted it.

Cyclosplitters work good, double gooseneck are the best for the most reactive compounds.

You can sometimes get away with using a 5:1 split ratio to help with breakdown if you can still get your required sensitivity.

Also when doing splittless injections, be sure your split vent line is clean. Gas phase chemistry gets pretty wild when you think of how fast those molecules are moving and how difficult it is to contain them all within the inside of the glass liner during the splittless period of the injection. They will travel all the way to the split line and back before entering the column and a dirty split line can cause breakdown. On the 6890 the split line is the copper one running from the injection port to the split vent trap at the back of the instrument, I now just replace that short piece of copper tubing instead of trying to clean it, works great.

[tangaloomaflyer]

Yep, I've had the same experience with James with the split line. Replacing it is easy and works wonders.

[Yama001]

I was just telling the poor kids who got stuck with cleaning and silanizing the liners to put the wool in first. All I get is strange looks, as they back away slowly .

[mckrause]

Sky liners from Restek work wonders.

The compound responses are not linear - nitro groups play hell with positive ionization (think lots and lots and lots of negative ions). Linear as all get out in neg ion CI. Using quadratic forced zero fit works quite well, and forcing zero anchors the curve and avoids the vagaries of quadratic fits near the MDL.

Gotta keep the head of your column clean and well trimmed. Any high boilers in your system are going to give you fits.