Memory effect

[LCFan]

Dear all,

I inject a standard solution (isocratic method). I get a peak with an area of 1113953.

Afterwards, I make 6 injections of mobile phase and I get the following peak areas:

11629 (S/N > 10)
1807 (S/N about 3)
1019 (S/N about 2)
1033 (S/N about 2)
953 (S/N about 2)
732 (S/N about 2)

In % of standard peak area:

1.04 %
0.16 %
0.09 %
0.09 %
0.09 %
0.07 %

Now, I dilute my standard by factor 5 and inject once again in the same sequence: standard / 6 x mobPhase. The following areas are obtained:

Standard: 226765
Mob Phase 1: 6767 (S/N about 10): 3.0% of standard
Mob Phase 2: 1180 (S/N about 3): 0.52 % of standard
Mob Phase 3: 735 (S/N about 2): 0.32 % of standard
Mob Phase 4: 578 (S/N about 2): 0.25 % of standard
Mob Phase 5: 384 (S/N about 2): 0.17 % of standard
Mob Phase 6: 400 (S/N about 2): 0.18 % of standard

Independent on the concentration I have injected: the second injection of mobile phase leads to an area being roughly between 400 and 1200, which correspinds to an area near the Limit of detection. But the peak does not completely disappear.

My boss means that the substance might stick on the column and with each new run, a small portion is eluted.

In my opinion, this is not the case due to the fact that I do not observe any peak broadening. For me, I should check the injection system.

Any opinions are highly appreciated.

Thank you in advance for any reply.

Florian

[danko]

I think you’re dealing with a good old day’s carry-over.
You’ll need to revise the injector washing procedure.

Best Regards

[bunnahabhain]

I see this effect on my Agilent 1100, too. Unfortunately I was not able to get the gameboy controller working together with Masslynx 4.1, so I cannot set up a needle wash program and must live with it.
If you have an Agilent injector, try the needle wash program. Make sure that the wash vial has a solvent at least as strong as your mobile phase, and remove the septum from the cap. Sometimes the carryover sticks to the septum, and as soon as the needle is withdrawn from the wash vial, it is contaminated again.
Good luck
Jörg

PS.: I know I can make a zero-volume injection from a wash-vial after each "real" injection for washing purpose, but I hate these sequences where only every second line contains a measured value. I'ts a PITA

[Gerhard Kratz]

I also think it is the injection system, it is in my opinion nearly 100% sure that it is not from the column. But it is a good opportunity to get a PO singed by your boss to purchase a new column. Take that opportunity. Just in case. Good Luck.

[danko]

Ohhhh………..bosses know so much!
Is there anything bosses don’t have an opinion on?

[push]

So,
Which wash program do you recommend?

[LCFan]

Hello,

thanks a lot for all your replies.

I will now start to establish a suitable wash program. I'll have a look on solubility data in the literature (solvent, pH) and will use it as a strong wash solution in my Waters Acquity Sample Manager.

And also the weak wash solution will be optimized. In addition, I will have to decide whether I shall inject in the full loop mode (don't like it as I have to change the loop for each method with different inj volumes) or in the partzial loop mode (with needle overfill). If I apply the partial loop mode and my initial composition of the gradient is 20% methanol: what would you recommend as weak wash solution?

Regards

Florian

[danko]

Weak wash would be 15 - 20% methanol.
Best regards

[LCFan]

15% Methanol applied as weak wash

Strong wash evaluated based on solubility data of analyte:
80 % methanol, pH 8

Mobile phase part A (20% methanol, pH 2.5 (H3PO4)) injected
=> analyte peak obtained (area: about 40'000)

Variation of injection volume (standard PLNO mode): 10, 15, 20 uL mobile phase part A
=> analyte peak obtained for each of these injections (area: always about 40'000)

Wash volumes varied from 500 to 2500 L for each wash solution (steps of 500), 10 uL injections of mobile phase part A:
=> analyte peak obtained for each of these injections (area: always about 40'000)

Peak shape ok in all cases, no peak broadening

Any suggestions for further trials?

Regards

Florian

[danko]

Florian, why these different pH values? At which pH is your analyte most soluable?
Maybe 15% methanol without pH control isn't the optimal weak wash liquid. Nor the strong wash for tha matter (?)
Please explain the rationales behind these pH choices and it might help finding a solution.

Best regards

[LCFan]

Hello,

the analyte is an API of a tablet, which dissolution is tested by UV. Within the development of the dissolution procedure, the solubility of the analyte has been verified by varying the pH:

pH Solubility [mg/mL]
3.0 0.065
4.1 0.100
4.5 0.200
6.8 0.130
8.0 0.370

In the European Pharmacopeia, the analyte is dissolved in methanol for several tests. Thus, I prepared methanol/water mixtures with 0, 20, 40, 60 and 80% methanol and adjusted to pH 8. Fastest solubility obtained for 80% methanol/pH 8 => chosen as strong wash solution.

Monitoring wavelength: 240 nm
Column: Kromasil C4

The analyte is a base (pKa of 11); I perform chromatography at pH 2.5, where I have a good suppression of the generation of anionic silanols: peak shape is very good (and going up with the pH to 4 or higher, the peak begins to tail). I adjust the pH of the mobile phase with phosphoric acid due to its low UV cutoff as I have a second analyte with no chromophor that I have to monitor in the same run at 215 nm.

Regards

Florian

[danko]

Try to adjust the weak wash' pH to 2.5 or better yet utalize mobile phase A as a weak wash.

Best Regards