Decrease in Retention Time?

[TNA]

Dear all,

I've developed an isocratic method for one product containing 2 analytes. One is a basic compound while the other is an acid. I used phosphate buffer 0.025 M pH 3 and acetonitrile for the mobile phase. It seems that this method is quite valid for the product except there was a little effect on retention time of acid analyte. Even though some replicate injections while running system suitability test showed relative standard deviation less than 2%, the retention time of acid compound tends to decrease. On the contrary, I didn't meet similar problem with the basic compound.

I've increased the buffer molarity into two times. This resulted smaller relative standard deviation of some replicate injections but the retention time still tends to decrease slowly. Do you have any ideas to solve that problem?

Thank You

[TNA]

Sorry, at the previous post I was wrong in mentioning pH of phosphate buffer. I used buffer with pH 6.6.

[XL]

What type of column did you use, C18 or other type?

[TNA]

Actually I used a CN column

For addition I've mixed the mobile phase manually in order to reduce any influences variables from LC instrument.

[XL]

Silica based bonded phases usually have at learst 50% underivatized silanol groups on the surface. Above the pKa such as the pH you are using, silanol groups are negatively charged, repeling anions (acids) and attracting cations (bases) via electro-static interaction. Any HPLC column will lose its bonding over time. That's the reson we call them consumables. After loosing bonded phase, more silanol groups are present which repels acids more and should retain based more at the same time. Some columns, such as CN phase you are using, have lower hydrolytic stability than others (e.g. C18). pH6.6 is not a very mild condition for CN phase. Thus what you have seen is not unusually. The solution to it can be 1) a more mild mobile phase, and/or 2) another column type with higher hydrolytic stability. Did you try other column types?

[TNA]

I've tried other column types (C8, C18, Hilic) but all of them didn't show excellent separation. CN was a better choice for isocratic since the analytes have great variation on pKa (acid:4.4 and basic:9.5).

In the previous post (topic: pKa and Buffering Capacity) it was mentioned that using pH in the mid of both pKa region (neutral pH) would give stable retention time therefore I choosed pH 6.6. But it seems that it's not so when we use CN column.

According to your suggestion, what pH should I use with the same column (CN phase)?

Thanks in advance

[XL]

Tina,

I would say pH between 4 and 5.5 is friendly mild condition for silica columns. Acetate is a good buffer for this pH range. But I am not sure how this wil affect your separation. Probably pH5.4 acetate buffer will serve the purpose.

[TNA]

Dear Liu,

I've tried to use ammonium acetate buffer pH 5.4 as your suggestion. So far it seems that its running was successful since the retention times of both analytes are stable and the system suitability was fulfilled. Furthermore I have to perform its method validation. I wish there are no other problems

Thanks a lot for your last suggestion.

[HW Mueller]

Just wonder what conditions you tried on C8, C19, or HILIC before "junking" the idea of using them.

[Uwe Neue]

When you do the method validation, you should do this on a new column. This one has obviously deteriorated already due to the higher pH in which it was used.

[TNA]

Dr. Neue,

Is it absolute that I have to use a new batch CN column for method validation because it was said on the column details that the column is stable under pH condition 1.5-7?

HW Mueller,

For C8 and C18 I've tried using gradient elution since isocratic didn't show an effective and efficient separation. There was a ghost peak appeared exactly at the retention time of acid compound which difficult to be well-separated

Thanks for all

[Uwe Neue]

You saw a change in retention when you used it at pH 7. What makes you think that the column si the same before and after these changes?

[TNA]

Ok Dr. Neue

I'll use a new one for its validation.

Thanks for your suggestion

[HW Mueller]

The main thing about which I wondered was whether you tried optimizing conditions, especially of pH, when you worked with C-8, C-18, HILIC.

[TNA]

Dear HW Mueller,

When I ran method development with those three types column, I used either pH 3.0 and 6.6 for each based on some literatures I found. Do you have any suggestion?

Thanks in advance