Analyte RT difference between Agilent 6890 and 7890

[Terry]

Hello,

I am on the receiving side of a GC residual solvent method transfer. The analytes are Acetone and Ethanol.

The RT for Acetong and EtOH in the orignating lab are 3.0 min and 5.2 min respectively, using Agilent 6980 GC + PE HS40XL headspace sampler. I use exactly the same chromatogrphic conditions in my lab, using Agilent 7890 GC + G1888 HS. The RT are 4.1 min and 8.2 min.

Any idea on what causes the difference in RT? Thanks in advance.

Regards,

Terry

[Peter Apps]

Hi Terry

Can you do injections of headspace using a syringe directly into the GC inlets ? - you need to find out whether it is the GC alone that is the cause of the discrepancy, or the headspacer.

There are any number of possibilities involving gas flows, pressures, and restrictions, that will require some detailed troubleshooting. Another one that is much easier to check is that the phase ratio or the stationary phase in the columns is different.

If memory serves this is the third thread on retention time shifts when methods are transferred from 6890s to 7890s, so this might be quite common, and Agilent might know about it and be able to offer advice.

Peter

[Terry]

Dear Peter,

Thanks very much for your input. I will make a trial run using the liquid sampler tommorrow, and see what happens.

Regarding the GC parameter (pressure, temperature, flow rate) and column (manufacture, part No.), actually they are exactly the same.

It is holiday season here in China, Agilent service people will not be available until Feb 09. Could you share with me some key words for searching the previous two threads on this issue?

Cheers,

Terry

[Peter Apps]

Hi Terry

Remember that you must inject headspace vapours, not liquid, or you will get huge peaks that obscure the retention time shifts. This is probably going to be easier with a manual injection than fiddling with the needle penetration on an autosampler.

Key words: try "retention" in the GC and hyphentated forums, and then refine it from there. Most thread titles are not search friendly unfortunately, and if I recall correctly neither thread came to a satisfactory conclusion, one of them was very recent, in the past couple of weeks I think.

Peter

[R13]

In our lab we also had some differences in RT using 6890 and 7890 with the same type of the headspace sampler.

To add to what was already said there are two pints:
- is the column the same piece on both systems or You use each own column? Age and column type/producer might influence the result.
- headspace samplers you are using have completely different internal construction: G1888 is loop filling type, PE HS40XL is pressure swing type. The way the gas management systems operate in these systems might result in slightly different carrier gas speeds. Try to measure methane gas samples on both systems and compare the retention times.

[Terry]

Dear Peter,

Then I will follow your suggestion of manual injection.

I have found the thread of interest, thanks.


Dear R13,

The originating lab is in US, and I am in China.

1. the column in my lab is brand new AT-Wax column,
2. I will try to make a manual injection of the headspace sample by a syringe, and see the difference.

Regards,

Terry

[Consumer Products Guy]

Lot-to-lot differences in phase thickness?

When we do test method transfers, we don't detail how closely retention times match up; peak shape, resolution, results are of more importance.

[chromatographer1]

There are many differences between ANY two GCs which can account for the timing differences.

Flows must be checked MANUALLY to confirm their rates. Temperatures likewise must be manually checked.

I assume you are using the same detector type. You are not comparing a FID with a MSD. The MSD will be faster with the same headpressure.

You might even have a delay in your computer method starting the integration.

Do you have any unretained peaks to compare? Put some natural gas in your headspace vials and compare RT of methane.

There are some variances in column ID from column to column but they usually are major. It is more likely you have a difference in column lengths than ID.

THERE is a physical difference and you just have to find out where it is.

Good luck,

Rod

[Mr.Buffer]

I had both system in my lab 6890+PE and 7890+G1888 . I am thinking you are using your GC at Constant Pressure mode. Since PE HS uses constant column head pressure mode. If so, there should be no problem on the paper.
Actually , I never had same retention time with any peak with using same column.Difference gets bigger at late eluting peak in both systems.
In my opinion problem is all about both company use slightly different pressure control mechanism or so.
But I never had real answer , this is just a thought.

Best regards.

[aldehyde]

Yeah as others have said you have differences in the actual configuration of the instrument, and then you have the lot-to-lot type differences with columns and other materials your sample is interacting with.

If their column is older/shorter peaks would be retained less strongly and take less time to travel through the column, causing their RTs to be shorter. You'll need to compare all that if you want to track it down.

[Locutus_nor]

PE HS40XL and Agillent G1888 uses different injection tecniques
(balanced pressure vs. loop injection)

The HS40XL sampler probably uses high pressure injection, were the pressure is elevated during injection.
If the system is set up with a direct connection between the transferline and the column, this will cause increased flow and shorter retention times.