Hello!
Can anybody tell me how can I determine if a column has a void volume without oppening it?
Thanks.

Balto,

Void volumes cause peak broadening everywhere, and at the column head, broadening and even worse.
Sometimes the peaks trail and rarely start to split.

Run the quality sample for your column, calculate the number of plates, (using the manufacturer's method of measurement and calculation,) and compare that with the manufacturer's specification. If the system is normal and plates are decreased more than 10% you could have a void volume. Compare the shape of the peak with the quality control run that came with the column. Look for trailing, trailing will indicate a void.

Jim

You can inject Uracil or xyleen to determ the void time, this are unretained peaks at the front of you chromatogram.

A.Nonymous,

I think you mixed up two different problems. Just have a look a the answer from Jim...By injecting an unretained substance (on C18 material) like Uracil you can determine the void volume of your hplc system. You will need it for some parameters performing a system suitability test.

Balto;

An almost sure indication your column has developed a void is when the signals show a shoulder on the side or are very clearly split. This can be easier to see in well retained peaks, and it usually shows in all the compound shown in the separation.

The reduction in plate numbers (previous answers)can also be caused by many other factors (age, tubing, temperature, solvent composition etc).

Normally it is not recommended to try to repair the void by filling it with a packing similar to the one originally in the column. This requires some experience and caution. The best thing to do is simply to replace the column.

Shoulders do not always mean voids. Sometimes (rarely) a deposit of some insoluble material on the column inlet frit will cause a similar symptom.

Benjamin

Thanks to everybody.
I īve just done the test, and iīve lost a lot of plates...
And I have strong band tayling an d bend broadening...
What I donīt know is how it could have made a void volume with just a few runs...

You can try to use this column in reversed way, this may increase the lifetime of you column if it works.

What I usually see is what Benjamin (May 6) notes, and the shoulders are on the FRONT side of the peaks. These shoulders will slowly turn into split peaks as the void gets bigger. The other diagnostic observations I use are:

1. System previously worked fine. Current system must be the SAME, right down to the tubing lengths, pre-column hardware, etc. Changing these can introduce voids! Has the system been subjected to high pressures?

2. Reversing the column(s) solves the problem, but perhaps only temporarily. This is probably the definitive test. Be sure to take proper precautions to prevent back-washing particles into the system. With 1-2 cm guard columns, reversing the column may only solve the problem for a few hours. The packing is clearly moving back and forth inside the column. We had a batch of poorly packed guard columns that drove us nuts!

3. If the method uses a gradient (almost all of ours do), the problem is worse with early peaks, and disappears at longer RT's.

4. Reducing the injection volume, but keeping the analyte concentration constant, has minimal or no effect.

A problem that mimics a void is the injection solvent being more polar than the mobile phase at injection time. As an extreme example, we do a lot of work using ethyl acetate (solubility in water is <10%) as the injection solvent into a 10-90% ACN/water gradient. We are limited to analyzing 'later' peaks, and a 10 uL injection volume. If we need to analyze 'early' peaks, we must switch the solvent to MeOH. In extreme cases, on a very, very low dead volume system, even 5 uL MeOH injections of highly soluble materials caused peak shape problems, including peak splitting.

This problem was solved by adding some pre-column dead volume--approx. 100 uL--to allow the injection solvent to mix with the mobile phase. The analytes are re-focused on the head of the column before the gradient is started, and the peaks are beautiful!

Slight correction, Scott: the last problem occurs when the sample solvent is more non-polar than the mobile phase (in RPLC).