Chromatography Forum

Hi guys,
Currently I am doing some analysis of residual acetone and ethyl acetate in water. However, the RSD of standard injection is very poor for those two solvents.
The conditions I am using are:
column I use is DB624,
injection volume is 2 uL (this is a registered volume, can not change). Split flow is 75ml/min.
Injector:200
Detector:300

Glass liner straight.

Can anyone suggest the reason?

Your main problem should be reproductibility of the injection.

You should try a liner with constriction,for example a laminer liner or a liner with glass wool.
However water will ruin the deactivation of the glass woll....but will ensure a better evaporation process,and as better reproductibility.

What is your oven program,and column dimensions?

Bye

This is probably an example of different hardware performing differently with the same parameters.

Is the 'written in stone' injection volume a government requirement or self imposed?

2µL seems a very large injection if the method was developed on a capillary column to begin with. Why do a split if you need to inject so much on column? Why not inject 1/20th as much directly instead of a split injection which will immediately raise your RSD to a large fraction of 1%?

Are you using a megabore ID column?

You mentioned the split flow but not the carrier flow, do you know it?

Please give additional information. All of us here would like to help.

You should get RSD values of less than 1% unless your present parameters of your method are causing it to exceed 1%.

Of course a mis-installed column can cause the problem you describe.

I've done splitless injections on a similar column (megabore cap) - the only way to get 2µL on was to do it Very slowly. In otherwords, use an autosampler programmed to deliver the injection over the course of several seconds. This will minimize the flashing and sample losses associated with it (not to mention minimization of back-flash contamination of gas lines etc.).

Thank you very much for your suggestions.
The column we use is DB-624M 30MX0.32MM (the wide bore should be larger than this right?)
The inlet liner I use is the deactivate straight liner with deactivated glass wool. However, we found with the glass wool it seems the peak shape is very bad and as well, can not got the system suitability criterial which actually is 5% for 6 std inj.
The flow is ~3 with helium as carrier gas. The reason why I cannot change the parameters such as split ratio is that the method is registered.

If need more details pls let me know.
Thanks again!
sg

You are injecting 2µL of acetone and ethyl acetate in water, splitting the sample 25:1, with an injector port temperature of 200°C. I bet you are doing a fast injection with an autosampler, a burst injection which occurs in a few hundredths of a second.

You are seeing worse when you have glass wool in the injector.

You have all the characteristics of an injector that is overloaded with vapor flash, where you are backflashing upstream into your pneumatics, and where you are overloading the design of your GC.

Was your registered method originally used with a packed column?

Whoever designed this method overcame the problems by luck or by the good design of their equipment. Have you changed the film thickness of the column?

Parameters which I think are not optimal.

200°C is too hot for the injector. 25:1 split ratio for a 2µL injection of water is not an optimum injection in my opinion, and a 5% RSD for injection reproducibility is not very good either. 2% is generally the manufacturer's criterion.

I can only add that I hope the column is ok. All that boiling hot water doesn't do a column any good over months of use.

Good luck. I hope you can get this method to work.

I agree with chromatographer 1. In fact if you have a packed column injector or your capillary injector will accept a packed column I might even suggest that this analysis might be better performed on a packed column like Chromsorb 101. You could then inject the aqueous sample directly on to the packing and you would not have the problems of transfer in the splitting process. The packing is immune to water and will last for thousands of injections. Even with such a column I would not recommend such a large injection volume because of the expansion of water on vaporization.

Dear all,
Thank you very much for your reply. I feel what you said are reasonable for the water injection. However, the issue is that the guys don't want change the method since basically all the parameters are registered. So what we did is to try those things not registered.

Actually as u mentioned since the water expands too much, the backflash is serious if high volume of water is injected. So what we tried is to use pulse split injection mode combined with a Siltek laminar cup liner which according to supplier can tolerate up to 5 micro-liter injection. The results are very good. The peak shape and RSD improved significantly.

So I think for water injection (>2miro-liter), the important thing is to control the backflash is can not change the split ratio.

Thank you for your response. Many never tell us the "end of the story".

I would still recommend that you revalidate the method or replace the method with one with more appropriate parameters.

One never knows when the hardware will no longer be available for a short or long term basis.

A method using reasonable parameters that can be easily duplicated in another lab is the goal of any analytical method used in business.

Glad you were able to overcome.