Buffer preparation

[tom jupille]

This one will certainly become a case study in our Troubleshooting course!

[ananda]

Yes, this would be an ideal case study and people should be very careful when we are dealing with pH senstivie methods such as IXC and if methods are to be transfered to QC. Like Bill suggested, I am calibrating our pH meter exactly the same method our QC did and will let you all know the results.
Still dont know whether it is the pH meter or the electrode or the buffers or the way 2 labs do the calibration. So far it looks like not the calibration buffers but others have to be checked one by one.

Regards!

Ananda

[Victor]

This is an old problem. When I worked in industry we used to circulate a buffer solution for all our laboratories (not just chromatography!) to measure the pH blind using the standard method that they used. The results were amazing with a general consensus as to the correct result but some wild outliers and some disturbing results where the pH was out by +/- 0.5 or so.

I think your lab should establish a protocol for standardising pH meters, specifying an exact protocol as to how the buffers are made., and how the meter is standardized. We found this approach improved things dramatically. The first thing I would do is for Anan to take all the buffer solutions etc to the other pH meter and standardize it Anan's way. Then see if the pH meter reading is different. This would eliminate some variables immediately.

By the way, there was a throw away remark that making phosphate buffers with KH2PO4 or K2HPO4 was basically the same thing. This is not true because the concentration of potassium ions in two buffers made with these different salts is not the same. This could be a problem if you have negatively charged sites in your column (ionized silanols?) and you are trying to analyse cationic species.

[HW Mueller]

A very common source for errors in calibration (and reading samples) is that one dips the electrode in, lets say a pH = 4 standard then maybe in a pH = 6.88 standard, etc. Even though the standards are buffers, the second standard will no longer be 6.88. A 0.3 change in standard pH is easily attainable that way, one has to wash the electrode carefully before dipping into calibration/unknown solution.
On Victor´s buffer preps with phosphates,
if you prepare the buffers wrongly (namely by adjusting the K2HPO4 with HCl, etc., etc.) than you will, of course, becorrect in that two different buffers will be obtained, depending on starting point (K2HPO4 or KH2PO4). As a matter of fact, you would have to be superhuman to get the same buffer twice. If you adjust K2HPO4 with an equimolar KH2PO4, as an example you will always come to the same buffer as if you start with KH2PO4.......

[Bill Tindall]

Well, I have been pretty lucky at guessing so I will go for the jackpot:

The target pH is 3. If the claibration buffers are 4 and 7, which is typical , then the target value is out of range. Small details in the calibration can become big problems when extrapolated beyond range.

Some modern meters curve fit the 2 or 3 calibration buffers and all that goes on is hidden from view. I never trusted this approach. Being basically conservative in these matters, I think the old style meters of knobs and dials give one a sense of how well calibration is proceeding and hence a feeling of what errors may be present.

If the calibration buffers used were 4 and 7 then I suspect the observed differences in measurement between labs are a result of small calibration differences magnified by extrapolation outside the calibration range.

Why so? Electrodes are arranged to read about 0 mV at pH 7. This convention aided the old fashioned slope and intercept way of calibrating with knobs and dials (temperature[slope] and what ever else the other knob was called[intercept]) There is little slope error for readings around pH 7 because there is little slope contribution to the result. But, by the time one gets to pH 3(about 240mV), the mV reading from the electrode is almost entirely determiend by slope and any slope error is magnified the farther one gets from pH 7.

If I was stuck with making a measurement at about pH 3 and needed to do it with great precision, then I would prepare a tartrate or phthalate pair of bracketing buffers around the target and calibrate off these. There is NIST standard buffer at about 3.56 that could also be used as a single point check.

[Einar Ponten]

Maybe this fits better at “The Water Coolerâ€

[Ron]

I was at a workshop several years ago where a presentation was given on errors in common analytical techniques, and the property found to be measured incorrectly most often was pH. I certainly agree with Bill that all buffers should be prepared by weight to avoid the problems inherent in adjusting the final pH. The fact that two pH meters differed in the pH reading by a significant amount should be a warning about the type of error commonly occurring in pH measurement.

[TOT]

Thanks all of you for this useful discussion, I learned quite a lot about pH of the buffer. However, I am still not very sure about how to prepare buffer by weight. For example, to prepare 25mM phosphate buffer pH 3, I go to http://www.bi.umist.ac.uk/users/mjfrbn/ ... akebuf.asp to get the recipe and the result is "Dissolve 0.0025 mol of acid component. Dissolve 0.0224 mol of basic component (No added neutral salts, I due to buffer alone). Make up to 1000 ml with pure water" and another way is measuring the pH. My quetion is: which is acid component and which is base component? Are they H3PO4 and KH2PO4, respectively?
Another website for preparing buffer http://www.zirchrom.com/Buffer.asp also ask us to choose acid and base? May I choose H3P04 as acid and K2HPO4 as base for the buffer? Are there any difference between the buffers with the same H3P04 as acid but different KOH, KH2P04, K2HPO4 as base? What does "% of Max Buffer Capacity" the Zirchrom website mean?
I read in http://www.alltechweb.com/productinfo/T ... u333-2.pdf , they told that potassium phosphate buffer at pH 3 works really well for amines and I am working with amines now so I want to find out which is the best way to preapre this buffer.
Thanks in advance for all of your answer.

[HW Mueller]

For pH ~1.1~3.1 (K1): acid:H3PO4, base: H2PO4-
For pH ~6.2~8.2 (K2): acid: H2PO4-, base: HPO4--
For pH ~11.3~13.3 (k3): acid: HPO4--, base: PO4---

I havn´t seen this % max buffer capacity but this can only be the ratio of your buffer capacity to that of the best pH (at max. buff...) for the concentration used. See the discussions on buffer capacity in this forum.

You should NOT use KOH for any of those buffers (discussed recently).

[TOT]

Dear HW Mueller,
I am sorry but I am really confused now, why I can't use KOH for phosphate buffer pH 3? Because in my lab, we always prepare the mobile phase by weigh H3P04 85% and after that adjust the pH with NaOH 5M(use pH meter). After reading the above discussion, I think the way to weigh both H3P04 and NaOH will be better, not depend on the accuracy of pH meter. And as I said before, someone reported potassium phosphate buffer worked better than sodium phosphate buffer so I try to prepare the buffer from KOH and H3P04 by weighing.
I also tried the buffer pH 3 from H3P04 and KH2PO4, as the recipe I got from Zirchrom website, I have to weigh 2.882g H3PO4 85% and 23.09g KH2PO4, make to 1000ml with H20, I measured the pH of this solution, it is 3; however, the amount of KH2PO4 is too large so when I mixed with 80% of MeOH for my mobile phase, it precipitated out. The buffer pH 3 from H3PO4 and KOH is OK when mixing with 80% MeOH.
Sorry for long description but I really need your help to clarify which is the best way for me to prepare the buffer. Thank you very much!

[HW Mueller]

You made the point already, the two buffers are different. Maybe you mentioned the concentrations, but not wanting to read the whole thing again I would guess that your buffer which was prepared with KOH + H3PO4 is much more dilute than if prepared with KH2PO4 + H3PO4.

You should want to make a buffer which is reproducible and defined in regard to pH and ionic strength. This is the case when you mix the acid and base versions of phosphate (in this case). Theoretically you could also do this by adding the desired amount of H3PO4 in a volumetric flask (also H2O) and then adding the right amount of solid KOH. Practically this is not possible (KOH is not as well defined as KH2PO4 and can not be weighed accuratly as it is hygroscopic, picks up CO2).
Now you probably prepared the desired conc. of H3PO4, then added (probably undefined) aqu. KOH and ended up with who knows what concentration of phophate buffer.

[TOT]

Dear HW Mueller,
Could you please confirm for me one more time that the way using H3PO4 and KOH to prepare the phosphate buffer pH3 is not reproducible, even with pH meter, because in my lab we have used this way for years ? And what do I have to do if I need the buffer pH 3 to mix with 80% MeOH, the buffer from KH2PO4 and H3PO4 will be precipitated out. Thanks!

[HW Mueller]

You are not alone in this, usually chromatography is quite forgiving. If you titrate to the desired pH (without ever overshooting) and your final volume is just under that of the volume for which you calculated the desired concentration you will be reasonably ok. Forget it!
If you have to have 80% MeOH use a lower concentration of phosphate.

[Chris Pohl]

TOT,

Since you didn't get an explicit answer to your last question let me try to answer this question. What Mueller is saying is that the buffer you prepare based on phosphoric acid and monopotassium phosphate will be a different composition (based on the phosphate concentration) than the buffer made from phosphoric acid and potassium hydroxide. But this has nothing to do with reproducibility per se unless you are sometimes making the buffer using phosphoric acid and monopotassium phosphate and sometimes making the buffer using phosphoric acid and potassium hydroxide. Either method should be reproducible provided it is done in the same manner each time and as previously mentioned the best reproducibility is obtained when buffers are made by weight. One additional detail which leads me to prefer your method is the specifics of reagents in terms of measured pH value. Acids and bases are of defined purity based on specific assays, making them preferred reagents for weight based preparation of buffers. Reagents based on salts are less precisely defined in terms of reagent pH. For example, if you check Fluka BioChemika Ultra anhydrous monopotassium phosphate, the product pH specifications list an allowable pH range of 4-4.5. This pH range represents a significant variability in the absolute amount the monopotassium phosphate necessary to make a given buffer. Thus, one would expect that buffers made from acid and base which you describe as common practice in your lab should fundamentally produce better reproducibility in terms of pH if you're using weight based preparation and better ionic strength reproducibility if you're using pH measurement as a basis for preparation than using salts to prepare buffers.

[Bill Tindall]

My goodness, will this thread never end.

Chris, your argument for not using salts is not supported by history, in particular the example you chose to illustrate. The reason the monobasic salt has a pH range for its solution is that at this pH this solution is not buffered. Miniscule variance in composition or water source will chage the pH of this unbuffered solution. This salt will never be used to buffer in this range. pH wandering around in this range will have no effect on a buffer prepared from this salt in a pH range where there is any meaningful buffer capacity. In fact one of the British primary standard buffers is prepared from monobasic potassium phosphate salt, by weight.