Chromatography Forum

Goodmorning to all,

I'm a new user of this forum and also a quite newbie chromatography user so please apologize me if I make any mistake or ask for silly question.

Here is my question:

I've got GC (Agilent 6890) with 3 coloumns:

- Capillary Column (can't get the model right now)
- Molecular Sieve Column
- Hayesep Q Column

I would like to condition all the columns; while this information could be easily know for the capillary column since I get the exact model and have a look on what the producer suggest, how can I know the conditioning temperature of the Molecular Sieve and Hayesep Q column? Would you suggest to unmount the column and recondition them separately (maybe in another GC) or just try to find temperature and time that could fit for all the three of them?

Thank you very much!

You could Google the various columns and find out what the maximum allowable temperature is for each column. You would chose the lowest maximum allowable temperature and put the GC at this temperature (or 5 - 10°C lower) with carrier flow (you may increase the flow through the column to speed up the process). Keep them until you have a flat baseline. The best way of course is to individually condition the columns as PLOT columns have a "low" maximum operating temperature. Sometimes it takes a long time before residues are eluted from these columns. So patience is sometimes necessary.

Success,
Gilbert

Why do you think you need to condition the columns ? Are they new ? Are they all installed in one GC ? Do you have a gas analyser?

The easiest way to condition a GC is to leave it cycling over night running your temperature program. Unless you have a specific issue with your separation I would just stick to this.

You could Google the various columns and find out what the maximum allowable temperature is for each column. You would chose the lowest maximum allowable temperature and put the GC at this temperature (or 5 - 10°C lower) with carrier flow (you may increase the flow through the column to speed up the process). Keep them until you have a flat baseline. The best way of course is to individually condition the columns as PLOT columns have a "low" maximum operating temperature. Sometimes it takes a long time before residues are eluted from these columns. So patience is sometimes necessary.

Success,
Gilbert

I've tried to Google a bit and this is what I've got:

1. Porapak Q Column (actually, I thought it was an Hayesep Q, but I just checked and it seems I was wrong) - No information about conditioning - just 250°C Maximum Temperature

2. Restek Rt-Q-Bond 300°C Maximum Temperature - Restek suggest a 4h 200°C isotherm for conditioning

3. Molecular Sieve 5A - Conditioning: 300°C for 3h

Why do you think you need to condition the columns ? Are they new ? Are they all installed in one GC ? Do you have a gas analyser?

The easiest way to condition a GC is to leave it cycling over night running your temperature program. Unless you have a specific issue with your separation I would just stick to this.

No, they're not new but, correct me if I'm wrong, I thought tha conditioning was a standard procedure for "restarting" a GC that has been shut down for a long time (2 years); also, we're not really pleased by our analysis reproducibilty (not sure if could be a column or maybe a detector problem - all suggestions are welcome). Yes, they're all installed on 1 GC. What do you exactly mean by "gas analyser"? Are you talking about GC detector? Sorry if you find this question stupid

Actually, what we will try is to set an isotherm at 200°C (the lowest maximum temperature suggested for conditioning), and try to recondition all the columns together

I've tried to Google a bit and this is what I've got:

1. Porapak Q Column (actually, I thought it was an Hayesep Q, but I just checked and it seems I was wrong) - No information about conditioning - just 250°C Maximum Temperature

2. Restek Rt-Q-Bond 300°C Maximum Temperature - Restek suggest a 4h 200°C isotherm for conditioning

3. Molecular Sieve 5A - Conditioning: 300°C for 3h

Why do you think you need to condition the columns ? Are they new ? Are they all installed in one GC ? Do you have a gas analyser?

The easiest way to condition a GC is to leave it cycling over night running your temperature program. Unless you have a specific issue with your separation I would just stick to this.

No, they're not new but, correct me if I'm wrong, I thought tha conditioning was a standard procedure for "restarting" a GC that has been shut down for a long time (2 years); also, we're not really pleased by our analysis reproducibilty (not sure if could be a column or maybe a detector problem - all suggestions are welcome). Yes, they're all installed on 1 GC. What do you exactly mean by "gas analyser"? Are you talking about GC detector? Sorry if you find this question stupid

Actually, what we will try is to set an isotherm at 200°C (the lowest maximum temperature suggested for conditioning), and try to recondition all the columns together[/quote]


Now we know what columns you have and that they are installed in one GC it is clear you have some sort of permanent gas analyser
If the GC has sat for a couple of years you could well have separation issues some or all of which may be sorted by leaving the GC at 200C for a while however without a flow diagram we do not know if carrier flows through all the columns when left iso thermal.

The more info. you give us the less "stupid" the questions and the more useful the answers will be

Sorry, I?ll try to exaplin better:

I've got this Agilent 6890 GC equipped with these columns:
1. Porapak Q
2. Restek Rt-Q-Bond
3. Molecular Sieve 5A

and two detectors:
1. TCD
2. FID

It also has 3 valves:
1. Gas Sampling Valve
2. Switching Valve
3. Gas Sampling Valve

Here (http://it.tinypic.com/r/152dpc9/8) I will attach the complete scheme (unfortunately, it's just a poor quality scan).

Actually, I've been able to set the valve in order to let the carrier flow in all the three columns while their out is unplugged. Then, I set the oven to 200°C for 4 hrs.

Tomorrow, I'll try some analyses to see if I'll be able to get some better results

Ok so, I'm still here asking for advice

I've conditioned the columns as said in previous posts and started doing some standard analyses.

1st Step I've analysed a gaseous mix containing N2, CO2 and H2 (12 injection) and I've calculated a reproducibility factor (expressed as [MAXvalue - MINvalue]/AVGvalue) of 1.52% for CO2 and 2.1% for N2 (Hydrogen is missing because I'm not interested in quantifying it).

2nd Step I've analysed a gaseous mix containing N2, CH4 and He (15 injection) and I've calculated a reproducibility factor (expressed as [MAXvalue - MINvalue]/AVGvalue) of 7.1% for N2 and 5.1% for CH4 (TCD) and 1.9% for CH4 (FID).

Now, I'm not really satisfied with the second set of analyses and I'm looking for issue that could cause this bad reproducibility. First of all, I've noticed that inlet flows (either back and front one) are not so costant (i.e. they fluctuate of about +/- 0.5 mL) but, if I have a look in the Column 1 and 2 screen, it seems that the flows are stable; could this be a problem and possibly lead to a poor reproducibility?

First have you leak checked the system ?
What are the concentrations in the standards. The only reason I can come up with for a greatly different reproducability between the TCD and FID results on the same samples is an integration error

First have you leak checked the system ?
What are the concentrations in the standards. The only reason I can come up with for a greatly different reproducability between the TCD and FID results on the same samples is an integration error

If by "leack check" you mean tightening all the fittings outside and inside the instrument, the answer is yes!
If you mean doing the leak check instrument test, unfortunately the instrument doesn't seem to have one.

1st Step 70% H2, 23% CO2, 7% N2
2nd Step 89% He, 10% CH4, 1% N2

Just consider that FID and TCD have different injector, columns and gas sampling valve (see previous posts), so I suppose I can't exclude that the problem is not in the integration

If by "leack check" you mean tightening all the fittings outside and inside the instrument, the answer is yes!

A proper leak check requires the use of a hand-held electronic leak check instrument which is used to check all connections, valves, etc. inside and outside the gas chromatograph.

If by "leack check" you mean tightening all the fittings outside and inside the instrument, the answer is yes!

A proper leak check requires the use of a hand-held electronic leak check instrument which is used to check all connections, valves, etc. inside and outside the gas chromatograph.

Oh well, unfortunately I don't have one :S

Go old school
Set your GC to constant flow. Block the end of each column, End of TCD for Pkd, end of capillary on FID side. Close off any split vents or septum purge. The GC will ramp up the pressure in the column to the head pressure at the GC inlet. If your column system is sealed the flow display of your GC will drop to zero. If you have a leak it won't

Something we do often I don't think anyone has mentioned is baking the rest of the GC. We run our inlets at 125C and bump them up to 200C for bake. Also with the detector we run at 200C, but bake at 300C every few weeks. This is a good idea with the valve box as well, just be careful because the can have lower max temps.
Another possibility might be cleaning! Have you look in the injection port to see if there is any junk in there? We tend to have problems with the spindle in the valves getting gunked up after awhile.
These solutions may not apply as much to these super inert components, but might be a good idea nonetheless!

Thank you every one for the answers! Actually, I've just had a new problem :S

It seems that when the analysis reach the purge time, the inlet pressure drops immediately (from 1.07 to, about 0.85) and never returns to set-point! In a few minute, the GC will shutdown. What could be the problem? Is it possible that I should have to modify the purge time?