Buffer preparation

[ananda]

Hello Everyone,

I have a question about a buffer preparation for a CEC HPLC method. if I want to prepare a buffer of 20mM K2HPO4 and 1M NaCl at pH 3.1 (we noticed K2HPO4 is better than KH2PO4 for this particular method and pH 3.1 gives us the optimum resolution), what I do is dissolve appropriate amounts of phosphate and NaCl in 1 L measuring sylinder and adjust the pH and fill the cylinder upto 1 L mark -everything in one solution.
Another person makes it by mixing 500 mL of 40mM phosphate solution and 500 mL of 2M NaCl solution and then adjust the pH.
It seems thearoatillcaly there shouldnt be a difference in resolution retenition etc. but is there any little things that can affect a resolution of closely eluting impurity peaks with the main peak if someone make the buffer in this second way??(eg. volume of mixing, etc?)
Appreciate your comments.

best Regards!

Ananda

[Mark]

Ananda,

There should be no difference which way you do it assuming the NaCl is pure. If it were I doing this, I dissolve and adjust the pH in a wider container to make sure there was adequete mixing during the pH adjustment, then adjust the volume in the graduated cylinder. I also assume there is no organic present or the pH measured will be afirly meaningless.

Regards,
Mark

[tom jupille]

Also, if you are doing trace analysis, do not put the pH electrode in the bulk buffer. Pull an aliquot into a test tube, measure the pH, and then discard the aliquot. This avoids the possibility of contaminating your buffer with leakage or carry-over from the electrode.

[Einar Ponten]

Considering the pKa's of phosphate the pH is not within its most optimal buffering capacity so the final pH may vary for that reason.

[tom jupille]

pH 3.1 is on the edge for phosphate, but what else is available that has a decent UV cutoff? Formate would seem to be the only other choice.

[Bill Tindall]

Ananda, If I understand the two procedures correctly- in one case the pH is adjusted and then the solution is diluted to 1L. While in the other case, the 1L of solution is adjusted to some pH, resluting in a volume greater than 1L.

First, for the best reproducibility neither method is best. To be most precise figure out what weight of each component is needed to achieve optimum results. Weigh these amounts into the container and dilute to volume. Weighing is greatly more precise than measuring pH, probably several orders of magnitude more precise.

But, to answer your question, there could indeed be a difference. In elementary chemistry we learn that diluting a buffer does not change it's pH. In fact dilution changes ionic strength, ionic strength affects activity coefficients, and the pH electrode measures activity, not concentration. So in theory there will be a difference in the actual pH of the final solutions prepared each way because a different dilution is made after the pH is adjusted. But in your case the ionic strength is very high so small differences in dilution will be insignificant, certainly less than the likely error in pH measurement.

[Bill Tindall]

Ananda,

There should be no difference which way you do it assuming the NaCl is pure. If it were I doing this, I dissolve and adjust the pH in a wider container to make sure there was adequete mixing during the pH adjustment, then adjust the volume in the graduated cylinder. I also assume there is no organic present or the pH measured will be afirly meaningless.

Regards,
Mark

Mark, it is perfectly reasonable to measure pH in mixed aqueous/orgainc solvents. The glass electrode responds just as in water to hydrogen ion activity in many solvents and water-solvent mixtures, especially if they have some dipole. In some cases it is the only measurement that yields any useful information. I will advertise my class here. I will be teaching a class on pH in mixed solvents at Pittcon (providing 10 people sign up).

Just as the pH scale for water was developed, pH scales for several other solvents are available. NIST has standards for some methanol water- mixtures.

[HW Mueller]

As a partial convert to "Bill´s method" (still check the pH with a pH meter after preparing with the correct weights ....), I would like to mention that these correct weights can be calculated for you by some internet sites, like
www.bi.umist.ac.uk/users/mjfrbn/buffers/makebuf.asp
Most likely this site will ask you to weigh out acidic and basic forms of the buffer, which means KH2PO4 and K2HPO4 here.
If I remember correctly, citrate does not absorb (UV) much either, and buffers well at 3.1.
Also, Beynon´s site (the above URL) works with activities and ionic strengths, so one can "see" these effects.

[ananda]

Hello Everyone,

Thank you very much for all your replies and comments and every comment is great but as you know there are certain things that I can and can not. This method I am using for a peptide therapeutic and method is even validated here but Qc is having a resolution problem with the second method of buufer preparation. However Bill's suggestion to check the final pH after adding my 35% ACN is new to me..and I am sure to some of you as well. It sounded like he is very confedent about it and can we all communicate here further to assure that his suggestion si a god idea or still it is something that we think not appropriate. Thanks Bill.

Again thanks everyone!

Have a great Weekend!

Ananda

[Uwe Neue]

I know that Bill has done a lot of homework on pH measurements in the presence of an organic solvent. I have done such things myself, if I needed to do so. However, I violently disagree with him that such a thing would be enlightening for Ananda's problem. To Ananda, I recommend to stick exactly with the procedure used until now, and to not consider to measure the pH in the presence of the organic solvent.

Here is the reasoning: we all know that the pH changes, when I add the organic solvent. Also, the pK of the analytes change with the organic solvent. However, for chromatographic reproducibility, you want to have a consistent pH, and furthermore you want to know, if you are far away from the pK of your buffer. Since I do not know the pK of my buffer in the presence of the organic solvent, I am not able to know, how good my buffer is, if I measure its pH in the presence of the organic solvent. If I do the measurement in water, I know exactly what I am dealing with.

I only recommend to measure the pH in the presence of the organic solvent to people who understand these issues completely (like Bill), or who need to do this because they do some theoretical stuff that depends on the real pH (like me). For everybody else: measure it in water and don't worry about it.

[Bill Tindall]

I didn't (and would not in this case) suggest that Ananda prepare, or check, buffers after adding organic solvent. I was answering Mark'statement when I talked about that issue.

It appears to me from the response from Anada that the actual pH of this buffer is critical, such that subtle differences in the preparation of the buffer are important. I will bet a nickel that the origin of this problem is a difference in how the two labs calibrate and then measure pH. I have seen this problem before which is why I advocate preparation of critical buffers by weighing components and diluting to volume. It is very easy for two labs to be off by 0.1 to 0.2 pH units when buffers are prepared by adjusting pH to a target.

Another place to look is the concentration of sodium chloride. It must be added for some reason, but I am not familiar with chromatography that requires it. The concentration of sodium chloride will be different for the two ways to prepare the buffer. In the second case the buffer, as well as sodium chloride is diluted by adjusting the pH.

To further diagnose this problem each lab should prepare a batch of buffer and then each lab do the chromatography and check the pH of these two buffers. Also, calculate how different the sodium choride could be between the two methods and then test to see if this difference makes a difference in chromatography.

Please let us know the results. My nickel depends on it.

[Bill Tindall]

There was a time, and it was not centuries ago (well nearly half a century), when one did not, and could not, buy liter bottles of color coded buffer preparations for pH calibration. NIST, then NBS, provided the directions for calibration buffer preparation. If one needed a pH 7 buffer one WEIGHED the components and diluted to volume, just like making any other standard. All the other buffers used in that day, for example Clark and Lubs buffers, were similarly prepared by weights and volumes.

Given that historically buffers were prepared by weights and volumes, Vendors prepare the calibration buffers we now use by weigths and volumes, and most other standards have always been prepared by weights and volumes , why does the liquid chromatography community commonly prepare buffers by the less precise and more time consuming procedure of adjusting to a target pH?

If I had Ananda's problem I would determine what amounts of buffer components provided the desired chromatography. This can be done empirically using the chromatography as a response. And then for ever more I would prepare subsequent batches of buffer by weighing these amounts and diluting to volume. These operations can be done to 1 part in 1000 precision and accuracy. Given the log relationshop between measured pH and hydrogen ion activity, the adjusting to a target approach is greatly less precise and more prone to operation errors.

[ananda]

Hi Everyone,

Myself and our QC worked hard during the last few days swapping buffers, columns and samples to find the cause for our problem. If you remember the problem was lack of resolution of closely eluting impurity peaks from the main peak. First I found our QC prepare buffer component B (it is a gradiaten method) differently than me.

I thank everyone who sent their comments and suggestions. Finally I found where is the problem. BILL!!!!!!!!!!!!!!!!YOU DESERVE A BIG NICKEL!!!!!!!!!!!

YES..IT WAS THE pH of the 2 buffers. When I developed the method ideal pH for the resolution was 3.1 using phophate buffer. We found the 2 pH meters I use and our QC uses gives different pH for the aqueous phase of the buffer. My buffer that gave pH 3.10 showed pH 3.43 using our QC's pH meter.
pH difference of 0.3 is a big change and our pH meter and the electrode are from Beckman and the other one is from MettlerToledo. I knew it was the buffer and I thought it can be volume, ionic strength or pH. After Bill's reply earlier I wanted to check the pH and yes I think it is the cause. Now our big dilemma is why 2 pH meters show 0.3 difference in pH and which pH meter to beleive. I found the Mettler Toledo is a very fancy and latest model-pH meter called 'Seven Multi'. Now I am going to call both manaufacturers to find out which one to sue. (just kidding!). No it is not the calibration buffers, I checked.

Anyway, I want to send this special thank (of course along with well polished brand new nickel:)) to Bill and all others who send their comments.
You guys are great !

Cheers!

Ananda

[Bill Tindall]

thankyou for the kind reply.

I doubt the problem is with the meters, electrodes or buffers. The problem probably lies in the details of how the two labs calibrate the meter. These sorts of differences were common in my work environment and THAT IS WHY I DON'T ADVOCATE PREPARING BUFFERS THIS WAY (sorry for shouting, bye)

[Consumer Products Guy]

I fully agree with Bill's way of preparing buffers. I have seen numerous publications for which the buffer preparation steps are ambiguous.