GC Chromatogram Question

[dlarsonca]

I am relatively new to GC analysis, and have a quick question. Where I work we have an Agilent 6890 GC-PDECD. We are analyzing PCBs in soil. My question is in the chromatagrams the baseline starts at 1800uv then begins to decrease to 600 by the 20minute mark. The solvent injection peak occurs in the first 2 minute. Afetr 20 minutes the baseline stabilizes and the peaks are around 4minutes and higher. Should I be concerned by the initial drop in the baseline or is this normal? Is there someting I can do to stop this?

[zokitano]

Dear dlarsonca,

please give us more information about GC column, temperature programme of the GC oven and so on...

Best regrds

[shaolinseperation]

post a pic as well

[dlarsonca]

The column is DB 608 capillary column, length 30m, ID0.250mm, film 0.25um, operating temperature range 40-280Celsius.

Front split injector set at 150C and the oven final temperature is 250C.

Here is a link of a scan of ones of the chromatograms.

http://www.mediafire.com/?4nwztm5xbdr

[gcguy]

I had a look at your chromatogram, what it might be is contamination in your system somewhere. I would suggest that you hold the entire system at the column max temp for a couple of hours. You seem to have a low injection temp, I would be tempted to change the temp to at least match your oven profiles highest temp. Why do you cool the oven down to 250 at the end of the temp gradient?

GCguy

[dlarsonca]

I will bake the column for a few hours today and see how that helps. If this doesn't solve the problem could it be a damaged column? I have no past history recorded for this instrument so I do not know the age of the column or what the students have been doing to it for the last 8 months.

That is a good question, I will check and see why they cool the oven at the end of the program and maybe adjust the front injector temperature.

I am a little frustrated that for 8 months, prior to me, no one bother to document what they did. My background is primarily ICP analysis and knew how important it is to document what has been done.

Thanks for the suggestions, it is greatly appreciated.

[AICMM]

dlarsonca,

GCguy is right, you're inlet is way too cold. 275 or 280 minimum. PCB's are not going to degrade in the inlet so running it hot should be no problem. Since you are running the inlet too cool, I would get in the inlet and clean everything including solvent rinse everything you can.

One quick way to check, leave the oven isothermal and hot (250) and then turn up the heat on the inlet or switch from split to splitless and watch the baseline.

Best regards.

[AICMM]

dlarsonca,

I finally got a chance to look at your post and have a couple of other comments/questions. First, why are you running in IPA? IPA is going to have some response on the ECD due to the oxygen and it is probably going to have a pretty good expansion volume. Have you thought about switching to something like hexane. Second, are you really running a 1000 ppm standard? If so, what is your split ratio? Is this range linear?

Best regards.

[dlarsonca]

I checked an older version of the method the student based their method on and your right they reduced the front injection temperature. So I fixed that part of their method. Now this Temperature is set to 250C, i will try this temperature first and then increase it as per your suggestion.

IPA. That is a very good question I will ask the student why they picked it over hexane and suggest maybe to change it if her project coordinator agrees.

The student is calibrating from 5-1000ppm. She wants to make sure her samples will be with in calibration range. I suggested that she should do a low curve and a high curve and then analyze the samples.

I will check the split ratio and check linearity.

Thanks for the help

[dlarsonca]

I have cleaned the front injector and increased the temperature. I still get the same result.

The student's project is to use IPA since they are trying to come up with an on-site process that does not involve the use of acetone/hexane/dcm.

The split ratio is 10:1. The curve is linear but the student will be doing a low and a high curve now.

I also baked the column and the detector for 2 hours but this did not solve the problem.We are still getting the same chromatogram problems. Do you think the column could be damaged? When comparing her data this year from what she analyzed last year the peaks or about 1/3 the area.

[AICMM]

dlarsonca,

I don't think column damage because the back of the chromatogram looks good while the front looks bad. Damaged column would yield more garbage at the end and poor chromatography.

Can I suggest that you run a hexane blank to see if the chromatogram looks the same. I would reiterate that the IPA is my most plausible candidate for your baseline issues. If a hexane blank does not give you the same early baseline, I would suggest other candidates for extraction solvent. Another possibility would be to shoot less (and alter your split appropriately) or to start with a much higher initial temp, say 110C so that the polar solvent does not re-condense.

Best regards.