Carryover

[chfoo]

I am developing a gradient method separating Dihydrocodeine tartrate, Sodium Benzoate, degradants and excipeints on Kinetex C18 100x4.6 2.6um. Diluent used is 5% ACN. HPLC systemand degradation products- Agilent 1100. I am getting carryover of up to 0.2%. A diluent (blank) injection straight after many runs of stds and samples may contain minimum carryover while a second and third diluent injections may contain a much higher carryover. After several blank injections, the carryover eventually disappear. The carryover is less significant for Sodium Benzoate.

I have injected a blank, running the gradient condition 4 times i.e. the instrument method was repeated 4x in one run. Carryover was seen in the 1st gradeint but not in the 2nd to 4th gradient. This showed that the problem was not in the column.

I have also put an injector wash into the instrument method but did not solve the carryover problem.

Please help.

Thanks

[danko]

What liquid do you use for injector wash?
You may need to utilize a stronger solvent (higher ACN concentration) in this situation.

Best regards

[Gerhard Kratz]

It could be that your carryover comes from an so called "extra column effect", mostly it is the fitting on the column top. Please check that your connection to the column is dead volume free. Good luck

[chfoo]

The injector wash is the diluent, 5% ACN. This is used as the the actives are very soluble in water. I will try stronger solvent.

I have also tried washing the injector and the injector seat before starting a run but no luck.

Thanks Danko and Gerhard

[carls]

Are you running an injector program or using any built-in injector "optimization"?

If yes, set to standard injection mode and re-test. If this solves the problem you were not allowing sufficient time for all sample to exit the loop prior to moving the valve to bypass (this is primarily used to reduce gradient delay). To solve this problem insert a wait command with enough time to allow 5x the injection volume to be flushed thru the loop prior to going to bypass. Also, make sure the leak channel tubing is not plugged. Remember the inject valve should be moved back to mainpass with sufficeint time before the next injection to to ensure initial mobile phase completely fills the sample loop (100uL?).

Also keep in mind (If I remember correctly) the inject valve not automatically return to mainpass (loop in-line/inject) when the injector wash is done. Therefore, the wash solution will remain in the loop and be injected with the next sample injection.

[chfoo]

The autosampler used is agilent G1329B. I should correct that I am using needle wash in the instrument method (not injector wash). As I understand, this feature is an external needle wash. The mobile phase goes through the needle throughout the run. Can some one enlighten me on this?

[Bruce Hamilton]

You haven't said what the injection volume is, and whether you have the same problem with other analyses on the instrument. It's possible the rotor is worn. I would conduct a series of experiments to understand the problem.

Ensure that your standard compounds are readily soluble in the initial mobile phase.

Inject a high concentration of a mixed standard once. then inject a blank.
Repeat, but insert a full syringe volume ( 100 ul? ) injection of mobile phase between the sample and blank. Check the carryover for each of the compounds in the standrd. If the carryover disappears, it's probable that the contamination is from autosampler holdup. Could be several causes.

Inject an increasing volume of the standard, does the carryover increase or remain constant?. Inject the same quantity of standard, but varying the dilution and injection volume. does the carryover increase or reamin constant.

[carls]

The autosampler used is agilent G1329B. I should correct that I am using needle wash in the instrument method (not injector wash). As I understand, this feature is an external needle wash. The mobile phase goes through the needle throughout the run. Can some one enlighten me on this?

If you are not using an injector program but are doing a standard injection then, yes, the sample loop stays in the flow path and is rinsed with mobile phase for the duration of the run.

Carryover when using the standard injection mode can be caused by sample adhering to the outside of the needle and thus contaminating the blank solution and/or the needle seat. The amount of carryover possible by this source typically will not be seen with UV-vis detection but it is a possibility. I typically build an injector program which draws 4x the sample volume of an appropriate wash solvent and ejects it thru the needle seat. I repeat this wash twice bewteen runs. Inject a standard 3 times followed by 3 blanks. Repeat this process 3 times and if this solves the problem yb the third run then the standard needle wash option may be enough.

If the needle+seat wash does not solve the problem then you'll likely need to rebuild your injection valve (new rotor and maybe a new stator). You can replace parts one at a time if you want to isolate the source, however, if the stator is badly worn/scored then it could damage the new rotor. The stators cost 2-3x as much as a rotor.