buffers

[adamdrugchemist]

I've searched through the forums and haven't found an answer to my question as of yet so here goes.

At my laboratory we currently use a 25mM potassium phosphate buffer, adjusted to a pH of 3 with phosphoric acid. We would like to go away from such a salty buffer as it is destroying most of our stainless steel parts due to leaky pumps, seals, etc.

We use Agilent 1100 HPLC-DADs (there are a few 1200s around but the majority are 1100s). We collect signals at 210, 233 and 250nm so the type of buffer is key to our analysis. As an FYI we are a forensics lab that deals with the analysis of miligram amounts of ilicit drugs.

I've tried ammonium formate and ammonium acetate buffers (both 10mM conentrations and pHs of 3.5 and 4.5 respectivly) I've heard that both of these buffers are great as volatile buffers and are used widely in the world of LCMS. However, both of these buffers have a UV-cutoff that is not acceptable to our standards. We essentially need a buffer that can work around pH 3, be volatile, and have a UV-cutoff of less than 190nm.

After all of that background, here is my question. Would a buffer of ammonium phosphate, with a pH of around 3, be volatile enough so that upon leaks we should not see salt crashing out and corroding our systems?

I've done a lot of searching online and the only hits that I come up with pertaining to ammonium phosphate have to do with protein analysis at pH 7.

Any help would be much appreciated.
Thanks!

[Noser222]

Ammonium Phosphate will be fine as the buffer, and it should be a little better for the problem you mention.
As for being "volatile enough"...it's more hygroscopic so instead of dried up salt you might see thick liquid residue.

[danko]

...…..a UV-cutoff of less than 190nm.

I think there is only a few persons that can fulfill this particular wish. One of them is Santa Claus

Seriously I’m not sure you’ll find such a buffer

Best Regards

[grzesiek]

"We essentially need a buffer that can work around pH 3, be volatile, and have a UV-cutoff of less than 190nm. " - I want it too

you can use phosphoric acid instead as an alternative or any other acid that fulfills your list

[JA]

you can use phosphoric acid instead as an alternative or any other acid that fulfills your list

Is it possible to reproducibly prepare, by weight or volume, a buffer that is resistant to changes in pH without using a salt of the acid's conjugate base? I didn't think it was so.

I was about to make a point that, for our Fluka bottles at least, the phosphoric acid content (stated as 85-90%) changes with each batch and how this might confer some variability to the mobile phase preparation. Then again I have seen a few methods in our lab which utilise something like a 0.1% solution and as they presumably satisfied the typical validation parameters (I'm unsure if robustness w.r.t. acid content was included) I wonder if I'm making a point about nothing?..

[Bryan Evans]

At my laboratory we currently use a 25mM potassium phosphate buffer, adjusted to a pH of 3 with phosphoric acid. We would like to go away from such a salty buffer as it is destroying most of our stainless steel parts due to leaky pumps, seals, etc.

Keep your system pressure well below the 200 Bar ceiling.
That should help with leaks (& instrument downtime).

[Consumer Products Guy]

HPLC purists will probably roll over in their graves, but we RARELY use real HPLC buffers here, even for our in-house validated cGMP and GLP test procedures. I suggest using aqueous phase of simply dilute acetic acid or dilute phosphoric acid, with no corresponding salt. We don't flush out the system between uses, and don't keep our pumps running (they shut off when the sequence is completed); plus, we don't need to membrane-filter any buffers. True, we don't have a "true buffer" system, but the pH stays steady and avoids all the issues you describe. We've got about five 1050 systems and equal number 1100 systems, have done this type of HPLC for almost 30 years.

[Camisotro]

I was about to make a point that, for our Fluka bottles at least, the phosphoric acid content (stated as 85-90%) changes with each batch and how this might confer some variability to the mobile phase preparation. Then again I have seen a few methods in our lab which utilise something like a 0.1% solution and as they presumably satisfied the typical validation parameters (I'm unsure if robustness w.r.t. acid content was included) I wonder if I'm making a point about nothing?..

Well, have you tested how sensitive the retention times are to the difference between pH=2, 2.5, 3? If any of your acidic analytes have a pKa close to that region then you will see a lot of variability... if their pKas are, let's say, 4 and up, then maybe it won't make so much of a difference.

[JA]

I'm sorry to hijack the thread, but in response to ctroster:

I know that where we have a method defined at a precise pH it will have been assessed for robustness at +/- 0.1 units. I think is fair enough for a weights-and-volumes buffer preparation, but I do not now how I would justify it for cases where the mobile phase is titrated to a particular meter value - I'm referring to variance in the meter calibration and measurement which is never assessed; once the meter 'passes' it's daily 2- or 3-point calibration it is essentially treated as being spot on

I would therefore ask what pH window other users validate on for the two aforementioned techniques of buffer preparation?

[kerri]

What about ammonium citrate/citric acid buffer? (pkas: ~3.2, 4.8, 5.2)

Anyone ever try these in the mass spec?

edit: I don't think that the UV cutoff of citrate is good in this case though.

[tom jupille]

I would therefore ask what pH window other users validate on for the two aforementioned techniques of buffer preparation?

The USP by default allows pH to be adjusted by +/- 0.2 units if necessary to meet system suitability, which implies that you can measure more accurately than that.

There is actually no "global" answer, because in practice, it depends on the robustness of your method with respect to pH. If you are well away from the pKa's of your analytes, then even a significant pH error (on the order of 1 unit) may not have any effect on the results. On the other hand, a non-robust method may have problems with 0.05 unit error.

[HW Mueller]

If one reflects the statements of CPG and Tom one may conclude that a low pH mobile phase is a robust system if one chromatographs weak acids.

Now we have been through buffer discussions for years, still this low pH (or very high pH) matter is still unresolved here, as a matter of fact, some recent buffer calcs showed me that I should be unhappy about an old contribution of myself. I am thinking of starting a separate chain on this. It is clear, though, that at a pH near 2 or below one can neither have a classical buffer nor need one. A consequence of arithmetic.

Ja,
do I understand this correctly: You are worried about different phosphoric acid compositions? If the statements on the bottle are reasonably accurate one can recalc the amount needed for each bottle. Even better: One can get HPLC grade phosphoric acid (I am not in the lab right now, but I think ours is from Fisher?).

[JA]

The USP by default allows pH to be adjusted by +/- 0.2 units if necessary to meet system suitability, which implies that you can measure more accurately than that.

Should I interpret this as in their view one can repeatedly prepare a mobile phase which is titrated to within +/- 0.2 units of a prescribed pH via use of a calibrated meter? Do you subscribe to this opinion, or is my question irrelevant as regulators (including those outside USP jurisdiction or perhaps consultants you are trying to satisfy) would or should not have any concern since it is allowed by USP...

In the context of my question, how does one evaluate a method's robustness with respect to pH in the cases where a meter may not be accurate despite having successfully passed its calibration. Searching some old threads we can pick up on the issue of extrapolated pH measurements, i.e. those outside the calibrated range, which admittedly does happen here on occasion. Another case where a user found 0.5 pH unit variation on mobile phase prepared in two laboratories. I wonder if this just comes down to a simple case of whether the all procedures involved were exactly the same...


HW:
For a long time, we have used Fluka's HPLC grade phosphoric acid (Sigma-Aldrich 79606). I think I was trying to make a connection between the label assay (which may be often ignored, I wonder), the resultant variability, despite using a weight or volume approach, in absolute acid concentration in solution and the knock-on effect to the pH of your mobile phase.

I look at the label and think "wow, you allow up to a five percent variation in assay for this stuff?" Maybe the changes from one batch to another are rendered largely irrelevant due to the log- scale for pH?

[grzesiek]

"Is it possible to reproducibly prepare, by weight or volume, a buffer that is resistant to changes in pH without using a salt of the acid's conjugate base? I didn't think it was so." - reproducibly is a big word that means nothing without specifying the variation that is allowed

[HW Mueller]

Ja, I don´t remember that the specification uncertainty on the phospohoric which I use is as much as 5% (are you sure that this was not a batch to batch variation?), on the other hand we are talking about the mobile phase, not accuracy of the analysis. If one wants to copy a method which runs at a pH that doesn´t give robust results, one has to play with the pH anyway. I can´t recall ever copying a published method without having to change things around (outside of pH also). If one is worried about repeating ones own results after preparing a fresh buffer there are plenty of discussions here on how to do this.