Good afternoon! A time change of 0.3 minutes on HPLC is not critical if there are no critical peak pairs. You are right, changing the pH will affect the retention time. if you want stability (stability is difficult to achieve on HPLC) then you have to prepare different mobile phases and solvents for each concentration. I advise you to keep the ammonium formate concentration and pH stable, also name the same solvent for the analyzed analytes, and when calibrating allow for a difference of +/- 0.5 min maximum.

Quynh Khoa Pham asked, "whether the retention of analytes of interest on the C18 column can be slightly affected by buffer concentration"?

- Yes, retention time may be affected by mobile phase buffer concentration. In fact, the entire process of HPLC method development utilizes this fact, and many others, in the process of developing HPLC methods which selectively retain and elute different compounds for analysis. Buffer concentrations may change peak retention and elution. Also, before concluding that retention times are changing you would need to run the same sample at least five times to determine what the repeatability is for the analysis (STD, mean Rt, Rt window). Any "drifting" of Rt may that the mobile phase is not properly buffered, the system not equilibrated, the temperature is changing and so on (many things). Someone with a basic level of training in HPLC method development will utilize these differences to develop methods that are reliable, stable and reproducible. Please contact someone local to you who has professional training in liquid chromatography for assistance (as students are not expected to have the training needed to do this on their own). *It takes the average scientist at least five years of full-time industrial experience to learn just the 'basics' of HPLC (emphasis on 'basic' as learning method development takes much longer). Request help to end up with a better method that follows good chromatography fundamentals.

Dear Alexandra Knyazeva and William Letter,

Thank you so much for your kind and informative answer.

Regarding ammonium formate, it is truly not a real buffer in HPLC, so different samples loaded onto LC column can cause drifted retention time. However, in my experiment, I used the same neat solutions (reference standards dissolved in a mixture of methanol and water), so I think that we could omit this issue.

In addition, to make sure that the LC column was fully equilibrated, an amount of mobile phase equivalent to 8 column volumes was allowed to pass through the column at the equilibration stage. Furthermore, I also checked the back pressure at the initial and end points of the gradient elution program to assure that they were similar.

For repeatability, the retention time of these compounds was the average of the 6 replicate samples with CV lower than 0.07% under the same concentration level of ammonium formate. Hence, I think that the repeatability of the method is good enough.

However, when I increased the concentration level of ammonium formate, I faced the retention time drifting as I mentioned above. As a researcher, I think that I have to explain this phenomenon instead of ignoring it.

I hope that I may get more valuable suggestions from you.

Best regards.

Quynh Khoa Pham: Based on your comments, you do not have enough formal training in HPLC yet (which takes many years to achieve a basic level. As a student, I doubt you wish to put aside 5+ years, full time, in industry, to just learn the basics of HPLC method development so you can complete this one task; For LC-MS basics you will need 10 more!). I apologize for being harsh, as it is great that you are trying to figure this out on your own, but these facts may be permitting you to misjudge and not understand what you are observing and which parameters may change the results (and why). False conclusions are the result. We can NOT teach someone a complex technique such as liquid chromatography in a few paragraphs here on a web forum (nor should we attempt to). You may not have a valid method of analysis at this point or are observing effects of a specific column and/or you are making basic errors (i.e. such as not dissolving your samples/standards in mobile phase, not water/MeOH; 8-column volumes may or may not be enough to equilibrate any column; ammonium formate is a buffer, but in your case it may be added, if you are using MS detection, only to increase / improve selective ionization, but will impact both the chromatography (peak shape and retention) and/or create source issues too; Also column chemistry is not as written in books as there are far more interactions going on the surface, which may vary depending on the column you are using [and their are thousands of columns which have similar spec's on paper, but are in fact, different]). Please find a local professional in chromatography to assist you with your project. In doing so you will have an opportunity to learn some of the basic principles while moving forward with your project.

• Learning HOW to research questions and problems efficiently is one of the most important skills a scientist can learn. Find and work with others (locally) who have the skills that you do not have so you can always move forward in completing projects. The web is a poor training tool for techniques such as these which are very hands-on and require years of on-site training.
• The reason you observe the Rt change with changes in the mobile phase chemistry are likely attributed to the amount of time the molecules spend interacting with the specific column surface chemistry. This may be the result of MULTIPLE forms of interaction, not one alone. Change the column type and they may go away or increase. The companies that develop the supports are not going to share with you their exact synthesis, coating/bonding procedure so you may never know exactly what types of chemical (or electronic) interactions are the cause, but as long as they are reproducible with a specific method and column, then it may not be worth investigating further (as this is normal when different columns, even of the "same" type are used). *I have been doing this for 30+years, developed ~ 20K + methods for companies worldwide and these are normal findings because the columns used are different (and have different behaviors leading to different results).

Dear William Letter,

Thank you so much for your answer.

I do not mind when you said such things. For my comments, I just showed some of my main points so maybe, you do not catch up my ideas and do not know exactly what I am doing in my own project. For that reason, you suggested what you thought was the best for me. Thank you so much for that. I am here to share what I know and observe during my research with other researchers ONLY. I think that this is also one of the good ways to learn many things from experts like you in my field.

I would like to make something clearer. When I said that "ammonium formate is not truly a real buffer", I said that based on the buffer definition, ammonium formate alone without the addition of weak acid (such as formic acid) cannot be considered as a real buffer even though ammonium formate alone can resist a MINOR change in pH. Ammonium formate is just one of the mobile phase additives I am using to develop an LCMS method for the same purposes as you mentioned above before validating the best method that meets all the requirements of my project. During my method development, I observed this phenomenon not only happening for sulpiride and diclofenac but also for several of my model compounds when concentration level of ammonium formate increased. Hence, I was so curious about that and decided to share it with people on research gate along with my own explanations and hope to learn much more from them even though I know the difference is not critical (as Alexandra Knyazeva mentioned above) and I can keep going.

For sample solvent, we know the mobile phase is a good option for dissolving samples loaded onto the LC column as you suggested. However, for practical reasons, it is not always the best option. In my case, I already tested a variety of sample solvents and a mixture of methanol and water in a determined ratio which I mentioned in my previous comment is suitable for my reversed phase LC-MS method (peak shape and retention) and solubility of my model compounds.

For the re-equilibration step, I also checked different column volumes from 4 to 10 column volumes, back-pressure at the initial and end points of the gradient program, peak shape and the repeatability of retention times of different model compounds from non-polar to polar compounds (not very polar compounds because I used the reversed phase LC column not HILIC), I found that 8 to 10 column volumes showed the same data so 8 column volume was selected. If this is not the right way to do that, may you please suggest me a better way or references (book or publication)?

As to the reversed phase LC column chemistry, I also checked two different columns including conventional C18 and polar RP-LC column (we do not have different LC column types in our lab such as biphenyl or phenyl-hexyl column to check), both of the columns showed the same phenomenon. I know that if a different LC column type is used, the phenomenon may go away thanks to its special stationary phase. However, for our lab condition, I do not have this option to check this phenomenon so I have to make full use of what I have now to do that.

Anyway, for my question, I am curious about the phenomenon so I hope to listen to ideas from everyone. Besides, as you suggested, if they are reproducible with a specific column and method, we do not need to go so far.

Thank you so much again for your advice.

Best regards.

I appreciate the additional information.

Regarding my point about column chemistry variations for the 'same type of column' (example, C18. I was not suggesting different support types, C8, C4 etc or even modes). You wrote; " I also checked two different columns ". TWO columns. What you do not know yet is that if you tried 12 different "C18" columns of identical dimensions and particle size from different brands, groups etc, then you would probably observe 9-12 different results, including different retention time for peaks and different responses to changes in the mobile phase composition (such as changes to the composition or additive level). This is normal because their are many variations in the column chemistry used, even from the same make/model/type of C18 column which change the results. Even the packing process used for each HPLC column may change the results seen (again, this is normal, but we do our best to try to minimize variation within a batch of similar columns). You tried only two columns, not even the same types, so any conclusions gained from that should not be relied on as the sample set is far too small. Without any buffer or control, variations are normal.

Back to your original question (and I am done with this thread)... You observed a small (tiny) difference in Rt from small changes you made to the mobile phase composition. No worries, this is normal because no two columns are the same and each may interact to a different degree (or none at all) when you make small changes to the mobile phase composition. Try 20 different C18 columns and you may find some show no change at all, some do. Pick and choose the one that you want for the application (this is what we do in analytical laboratories). The reason for the observed change (change, not "drift") is due to multiple interactions of the solute on the surface of the support. Small changes in salt conc. may change Rt. Hard to say if it is electronic or ionization (pH) in nature, but it is so small, practically, it does not matter. All chromatographers are familiar with this (no need to 'explain it'). Run enough samples and you will see it. In fact, the one thing that they might question is why you would use a mobile phase of just ammonium formate without an acid (e.g.formic). Why not use a buffer? The method appears incomplete. For LC-MS applications, most would benefit from such a solution to promote ionization and maybe change the separation factor too (depends on samples). *Stability and reproducibility of methods are key goals in HPLC, esp when transferring it to another lab where it should easily be reproducible.

Although there has already been a long discussion on the issue, in which I agree with the allegations of colleagues above, I would like to add two more probable causes. First, if the mobile phase is not buffered, increasing salt concentration changes pH, which inevitably affects retention. Second, increasing the salt concentration increases the ionic strength of the medium, which also interferes with dissociation phenomena.

Hope this helps

I note from the Merck Index that ammonium formate is deliquescent. This could mean that its concentration in different batches of eluent is not constant. Have you allowed for this?

Yes

It i interesting to read the discussion. I just add another possible reason: considering the pKa of formic acid around 3.8 and pKb for ammonia is 4.8, ammonia formate is an acidic salt. if you increase the concentration of ammonia formate, the pH of the buffer will drop down. Therefore, the basic compound will move faster and acidic compound will move slower.

Dear researcher:

Through the many experiments that I conducted in HPLC, it became clear to me that the buffer concentration greatly affects the retention time, so I advise you either to change the buffer and use another buffer that gives you the desired pH (according to the methods of preparing buffers) or to use another mobile phase that does not It needs a buffer (which is the best)