Bad shaped peaks

[KK]

Dear All,
my problem is bad shaped solvent peaks. The method is as follows:

Temperature of injector 140 d;
Temperature of column 35 d;
Split ratio 10:1
Column Supelco SPB-624 30 0.53 3
Carrier gas is nitrogen. Linear velocity is 25 cm/sec.
Standard solution: solution of methanol, ethanol, aceton, i-propanol, dichlormethan, ethylacetate, 0.02% of each solvent in water.
Injection volume 0.5 ul.

Sometimes peak shape is rather well, but mostly peaks are very broad and asymmetrical. Please, advise, what is my problem.

[img]

[DR]

Check for leaks at either end of the column (pay special attention to the inlet). Also, if you haven't done so lately, replace your septum.

[zokitano]

Dear KK,

Did you try to make oven temperature programme instead of using isothermal separation at 35C?

Using a ramping temperature programme can sharpen the peaks and solve your problem. (all solvents have different boiling points, take them in to consideration when making the oven temperature programme)

Best regards

[Peter Apps]

The problem is the injection of water to a hot inlet - this causes pressure and flow pulses, and you probably have some solvent effects from water entering the column. The solution to the problem is not going to be simple because the composotion and concentration of the sample do not leave much room to change parameters. The best way to do this analysis would be by equilibrium headspace.

Try reducing the inlet temperature and prgramming the column temperature.

What instrument are you working on, what kind of liner do you have in the inlet ? Please post a chromatogram - the instructions are in a sticky at the top of the LC page.

Peter

[AICMM]

KK,

I agree with Peter that you should seriously consider headspace at the concentrations you are working with. I would also like to know what liner since water has a very large expansion volume. I would also suggest that you consider increasing your split even more.

Finallly, I would venture that your methanol, ethanol and IPA are coming out right on top of the water peak on the 624 under these conditions but you would not be able to tell if you are using an FID. I would also venture a guess that your ethyl acetate and dcm have fairly decent peak shape since these are later?? With the column flooded with so much water, good separations will be difficult.

Best regards.

[KK]

Thanks for replies. I use Shimadzu GC2010 with FID. This is my chromatogram.

[Peter Apps]

Hi KK

Thsi looks like classice absorbtive tailing, and all the peaks are affected so it may not be silanols and siloxanes on the column. When last did you clean or change your inlet liner ?. If the liner is clean you could try cutting a meter off the beginning of the column. Ids there any packing or glass wool in the inlet ?

Peter

[KK]

I've cleaned linear before analysis. I am using Supelco untreated glass wool.

As a matter of fact, I am trying to do OVI analysis. Sample is 10 % solution of pharmaceutical substances in water. I was told, that one colleague of mine is doing this analysis in this way during three years using only one column.

[Bruce Hamilton]

That tailing may also be due to incorrect gas flows in detector or the injector, or to exposure of column to oxygen at temperature.

Check all gas purifiers, flows and pressures and, if they are OK, test the column using the manufacturer's test chromatogram method to see if your column has died..

I developed an OVI method with injection of a sample that only dissolved in water, and it was really, really sensitive to activity in the instrument. The instrument was left on with carrier flowing and detector hot, but a new manager turned it off unless needed - to save electricity.

When the analysis was required 6 months later we got very variable injections and tailing peaks, We replaced the insert, gas purifiers, and cleaned the solenoid valve and lines, injector ( but didn't passivate ) and used a new column, and still couldn't match the earlier consistent performance when injecting aqueous samples, but it was excellent for organics. The method was transferred to a headspace technique.

Bruce Hamilton

[shaolinseperation]

just had a similar problem.

our problem was the gas trap that was filtering oxygen and other junk out of the helium reached saturation point and started letting oxygen etc in.

the oxygen that made it in destoryed the coating and resulted in severe tailing like what you are seeing in your chromatogram.

so yeah, is it possible that oxygen is getting in somehow? a leak perhaps?