Ghost Peaks -- HELP!

[mhof]

I run a Hewlitt-Packard 6890 GC with a 5793 MS. I see signigicant peaks in my solvent runs (we use ACS grade hexane as a solvent). I have cleaned the inlet, and the ion source, replaced the inlet liner, and base seal, and do not see these peaks when I run the method without making an injection. The air and water check for the system comes up clean (18/69 = 0.31, 28/69 = 1.12, 32/69 = 0.22, 44/69 = 0.19).

If I concentrate the hexane under a stream of nitrogen, and then inject it, the peak areas increase significantly, but I purchased a new lot of hexane, and the problem didn't go away.

I'm at a loss. Where could the contaminants be coming from?

[Peter Apps]

My money is on the hexane. Try another solvent, or buy (beg or borrow) hexane from another supplier.

Second possibility is the glassware, what happens if you rinse with hexane, discard the hexane and re-fill.

You are not using plastic pipettes by any chance ?

Peter

[chromatographer1]

Good calls, Peter.

And let's not forget to check the nitrogen. Is there any hosing or piping involved? Or any pumps or traps involved in supplying the nitrogen? What happens when you concentrate using another gas?

best wishes,

Rod

[Rick]

Your split vent line is possibly dirty...if you inject some µl over liner capacity, you will have backstream...check liner capacity
You can check your syringe, change your liner, clean or change your split vent line, cut 1 meter of your column but before, check if problem is not your solvent..(try ethyle acetate for example)

[DR]

My vote is also for new solvent. ACS grade is not generally suitable for GC work (use chromatography grade).

[mhof]

First of all, I'd like to thank you for your responses. I'm so grateful that you have taken a moment to offer assistance.

So...

I cleaned the split vent line, and replaced the trap. I changed the carrier gas, as well as the gas traps and lines, thoroughly cleaned the ALS syringe, and finally, I ran a sample of GC grade hexane (from a new supplier) in rinsed and baked glassware.

The results with the GC hexane? Still very messy, although the contaminant peaks are markedly different than those from the HPLC grade hexane run. Concentration of the GC grade hexane under a stream of nitrogen results in massive amplification of the contaminant peaks. I ran a sample of the HPLC grade hexane that I had tried earlier, and there is also no change with it. The nitrogen I used to concentrate the sample is UHP grade, and because concentrated samples show amplified peaks, and not new contaminants, I don't believe that it is a source of the problem.

As before, only a blank run (no injection) looks clean. I'm stumped.

[Bruce Hamilton]

Are you using n-hexane, rather than hexane fraction?. The latter will have more higher boiling components. The specification to check is the residue on evaporation, or non-volatile matter, it should be as low as possible. You can warm the blank vial and sonicate to see if more of the rubbish appears.

Ensure that the syringe, sample vials, etc are fully cleaned by rinsing with a little of the solvent first. I would condition the instrument, and then perform multiple injections and runs ( about 6 ) to ascertain if there is any downward trend. I'd leave the instrument for an hour or two at initial temperature and repeat the injections.

If the peaks remain the same size, consider changing your hexane grade again, looking for the lowest residue, and ensuring the minimum number of transfers are via clean glassware. I pour straight from the winchester into the autosampler vials when I'm searching for solutions.

If the peaks increase, start looking at your inlet, column, and gas supplies.

Good luck,

Bruce Hamilton

[Victor]

I did not reply to this question before because I did not understand what ACS grade hexane means? However, it seems this is some sort of HPLC grade hexane. This is a complete no-no for GC; as far as the HPLC is concerned, it does not matter at all if the hexane is contaminated with higher alkanes, because they are UV transparent and all have similar elution strength in HPLC to hexane.

There are many grades of hexane. There is hexane which is especially made for trace analysis involving GC e.g. for analysis of pesticides, where the hexane used for extraction is concentrated during the process maybe hundreds of times. so it needs to be highly pure. Are you sure, when you say "GC hexane" that this is what you are using? Also, you did not mention whether you had cleaned out the injector, but I am still suspicious of the hexane (as Bruce seems to be), because your peaks get bigger when you concentrate it.

[Bruce Hamilton]

I did not reply to this question before because I did not understand what ACS grade hexane means? However, it seems this is some sort of HPLC grade hexane. This is a complete no-no for GC; as far as the HPLC is concerned, it does not matter at all if the hexane is contaminated with higher alkanes, because they are UV transparent and all have similar elution strength in HPLC to hexane.

ACS stands for American Chemical Society, and referes to their " Reagent Chemicals " handbook, published about every 5 years, which lists the specifications for analytical grade laboratory chemicals. However, as with most such specifciations, they are an agreed compromise.

Hexane petroleum fractions and n-Hexane come in different grades, eg the ACS n-hexane grade has a residue limit of <= 10 mg/kg, whereas the Merck catalog lists n-hexane grades for GC and Trace Organic applications with <= 4.6 mg/kg residue limit.

GC grade specifications tend to focus on the effect on ECD response, HPLC grades on UV response, and the impurities can be different.

However, because hexane is such a good solvent for non polar and partially-polar molecules, you also have to minimise transfers and contact with polymers. You really hace to watch out for leachable material from vials, volumetric dispensers, filters, caps, etc. It always pays to have fresh bootle of solvent from a known supplier when problem solving.

Bruce Hamilton

[chromatographer1]

trouble-shooting suggestion.

Try using pentane or isopentane as your solvent and concentrate it and see if you are still having problems.

This will eliminate everything but the hexane as the cause of your problems. Perhaps you will be able to substitute another solvent for the hexane as well.

It sounds like you are making progress.

best wishes,

Rod

[Peter Apps]

A further thought - what kind of autosampler vials and caps are you using ? This is not likely to be the sole cause of the problem because you see different contaminants with different hexane, but when all the obvious causes have been eliminated it has to be something unexpected and obscure.

What are your target analytes ? Any clue as to how big the contaminant peaks are in relation to the analytes ?, if you jack up the sensitivity on any instrument you see a messy baseline, but it might not be a practical problem. By default MS Chemstation normalises the TIC to the biggest peak on the chromatogram, so it will blow up even an insignificant blip to a full scale peak if there is nothing else on the chromatogram.

Maybe you could post a chromatogram - the how to is at the top of the LC page

Peter

[mhof]

I have inserted two sample chromatograms: the first is the ACS grade hexane, the second is the GC grade hexane. The contaminant peaks are at least as large as the compounds we target (plant/insect produced volatile compounds). I have added a standard to the hexane to make sure the machine's sensitivity was normal, and the resultant peak was of normal size/shape.

The Epavorative Residue in the GC grade hexane is rated at <0.001%; the ACS grade is rated at <0.0002%.





Also, I rinsed and re-rinsed my sample vials, and have run numerous samples of each hexane (dispensed directly from the original bottle to the autosampler vial and used no cap). Even with several bottles of each hexane, the results are quite repeatable... There is no consistent downward or upward trend in the peak sizes.

Using the exact same temperature/flow programs with the ACS grade hexane showed no contaminants earlier this year.

[Bruce Hamilton]

Are these normal hexane or hexane fraction?.

Neither profile is what I would expect from normal hexane, however the GC grade profile does look somewhat like what I would expect from a petroleum fraction of hexane. As you see no downward trend, and different profiles for the solvents, I would still place the contamination in the hexane.

If product from the same supplier worked previously, you need to find out what has changed. Given your trauma, I would now ensure your solvents have not been contaminated during any upstream transfer, so I would evaporate down a couple of hundred ml using a clean rovap, and measure the residue to ensure it does comply with specification.
I'd also send the chromatograms to the hexane suppliers technical section for comment. Can you identify any of the major peaks?.

As suggested by Rod, I would also try some other solvents. if n-pentane is too volatile for your purposes, try suitable grades of cyclohexane, n-heptane, or iso-octane ( 2,2,4-trimethylpentane).

There used to be some special grades of C5 and C6 solvents with very low residues used for arson investigations, and it may pay to purchase a small quatity of one of those for reference purposes. Alternatively, look for the organic trace analysis grades of normal hydrocarbons from a major supplier.

I hope this helps,

Bruce Hamilton

[Schmitty]

Are you running in SIM or SCAN mode? What is your amu range if in SCAN mode. Can you run some of these peaks through a library? This is an MS afterall...

[mhof]

I found an old bottle of HPLC grade hexane and ran a sample through the system. The contaminants are absent... I spent the rest of today running the new bottle versus the old, and it's very clear that the new hexane is the problem, even though it is the same product. I guess I need to contact the supplier now, and see what's going on.

By the way, the contaminants id (NIST library) as cyclopentanols; cyclopentanes are listed on the bottle as a contaminant in the hexane (although it is supposed to be in very low levels). I just couldn't believe that all the bottles we have would be bad. I assumed that there must be some accumulation in the GC that I was unable to find.

Thanks again for your help.