Hello,

in our company is an old discussion: Should we use HS-GC with or without a split? Traditionally we work without a split, however, we have learned in the last weeks that using a split doesn´t reduce sensitivity and the reproducibility of the HS analysis is better. What is the general opinion of the "GC community": Split or no split? Who has experience in both variants of HS-GC?

Regards,

Textor

Look into manual.
It depends on GC and HS construction and analysis parameters.
If you are using capillary column and around 20 ml/min of HS flow - there is no way for all of the sample to go into GC. Where it goes? In some GC instruments split flow during splitless sampling is diverted trough Purge outlet.
I have used HP5980/HP headspace combination. It required split mode to be used all the time.

If you are using a 'standard' installation of a capillary column a split injection may be more useful. If you use certain 'tricks', you can get more sensitivity using a direct injection as well as increased reproducibility.
Every technique has its advantages and disadvantages. You optimize for your application.
Short answer: Use what works best for you.

Rodney George

In most cases, a split injection will give better senstivity, peak shape, and precision when using capillary columns. This is not true for 100% of applications, but as a general rule split will work better.

It depends on the GC inlet and the GC type you are using. If you are using an HP 5890 series II, I would strongly suggest split mode over splitless. On a 6890 the sensitivity is generally good enough to where you can split and still see what you are looking for. I hope this is of some help. You may want to review some application notes to see what others are doing.

I actually just performed an experiment on this. I found that when using a split injection, sensitivity is not lost for early eluting compounds which is consistent with the theory described above and in many books on HS-GC. However, I did notice a marked decrease in the sensitivity for later eluting compounds when the split is used. It is probably best to evaluate the sensitivity using both techniques during the method development stage.

Split discrimination can be non-reproducible as well as variable from one set of GC-HS conditions to another.

I prefer to use direct injection if possible which essentially eliminates this problem.

Your experience with early analytes agrees with mine, but very volatile analyte overload (peak broadening, a problem sometimes with direct injection) can be overcome. Direct injection can give you excellent sensitivity for both volatile and later eluting analytes.

But use the technique which works for your conditions and which is easier for you to implement.

Rodney George

Hi,

I am using a Perkin elmer HS-40. I use a direct linkage from the HS straight into the GC without thus eliminating the option to go splitless so that I could get 100% of all the analyte being tested.

Anyone doing this as well?

thanks

Hi Textor,

I use a 5890 HP / HS with split mode - no problem.


Thanks,

We use the hp7694 to monitor hexanal in snackfoods. We are running in split mode(.2:1 split ratio) and havent had any problems.

(.2:1) Split is that a typo ?? Just curious..

You are right. We are running with a 22.2 column flow and a 4.0 split flow. Split ratio= split flow/column flow which calculates to a .19 split flow rounded to .2 ml/min . I have seen the ratio represented in a ratio format but would probraly be more correct to leave it in its calulated form which would be .2 ml/min.

Thanks,
Chris

Hot or cold on-column will do the trick: you get what you put on!!!!

VEM

The sample loop volume and peak width determine the minimum flow rate of headspace carrier, which is usually higher than the (capillary) column flow, so a split is needed if only for that reason, surely? I use an HP 6890 with 1290 HS sampler.