Is it necessary to use a salt such as ammonium acetate to produce an ammonia/ammonium buffer (i.e., why not just use ammonium hydroxide)? I am using gradient elution and want a clean baseline.

Counter ion is not usually that important unless there is something specific you are trying to achieve or avoid. For example, acetate and formate would both have high absorbance at low UV wavelengths and thus would not be suitable for some application. However, acetate would be volatile and might be desirable for LC-MS. At 280nm for straight LC counterion probably wouldn't matter much.

It depends on what you are trying to do. If you only want to suppress the ionization of something, ammonia can be used in the same way as formic acid and acetic acid are used in the acidic pH range. If you want or need a true buffer, you need a counterion. Since you want a good UV spectrum, I am assuming that you are not using MS detection. Then you can use a range of counterions to the ammonium ions. Phosphate, perchlorate and sulfate are all very clean in the UV.

Ammonia (NH3) and Ammonium (NH4+) are the weak base and conjugate acid pair that are in equilibrium creating a buffering effect. A counterion such as acetate is a spectator ion at high pH, it plays no role in the buffering process. In fact, acetate only serves to complicate the UV absorption properties of the mobile phase since it absorbs more strongly that ammonia. In my situation, using a gradient, I prefer to maintain as flat a baseline as possible. So I chose to omit the acetate from ammonium acetate. My point is: if ammonium salts are being used at high pH, try ommiting the counterion by just using ammonium hydroxide.

I have used this as my aqueous portion of mobile phase: 0.3% NH3 in H2O. Pipet 10 ml ammonium hydroxide into a 1000 ml volumetric flask; mix and dilute to volume with high purity H2O.

Hmm... Let's recapitulate: In order to create a buffer, you need what you called the weak base and conjugate acid pair. For an acetate buffer, this is acetic acid, and an acetate salt such as sodium acetate. For an ammonium buffer, you need ammonia (NH3) and the ammonium ion (NH4+). From where do you get a controlled concentration of the NH4+ ion, if you just use a solution of NH3? Or, for that matter, from where do you get a controlled concentration of the acetate ion, if you just use acetic acid.
If you want to use a buffer, and the control that goes along with it, there are many options of "counterions" that can be used without disturbances in the base line. See my suggestions above!

At high and low pH, one does not need both an acid and its salt, or a base and its salt, to make an effective buffer. These are but limited definitions of a buffer. A buffer is something that changes only a little when a small increment of acid or base is added.
Most rigorously, the "goodness" or effectiveness of a buffer is related to the slope of the titration curve of the respective acid or base, dpH/d(increment of acid or base added). So, what does the titration curve of a strong acid or base look like at the begining of the titration (small increment of acid or base added)? Flat, little change in pH. Or, do the calculation. If you start with 0.01 M HCl and add enough base to neutralize 10% of the acid, the pH stays about the same, ie it is buffered. In fact, at the extremes of pH, strong acids and bases are several times better buffers than a 0.1M mix of a weak acid and its salt at a pH = pKa.

Certainly any solution with a pH of 12 or greater or 2 or lower is well buffered, even if one got to these pH's with strong acids or bases. Now in chromatography it seems to me that there isn't much present but a tiny amount of sample and some silanols to alter the pH of a solution, so the buffering capacity needed in chromatography is small. Hence, in chromatography one could make the case that 11 and greater and 3 and lower is adequately buffered, no matter how one achieved these pH's. One can achieve these pH's with ammonia and phosphoric acid. Therefore, I would argue that ammonia and phosphoric acid are respectable buffers for chromatography in the range of 11-12 and 2-3 respectively.

So, does anonymous have a buffer when they add 10 mL of ammonium hydoroxide to a L of water? Most certainly, yes. Does it have much buffering capacity? No. Will it work? Probably, because one doesn't need much capacity in chromatography. Will it be reproducible? Most likely, and that is the main issue, assuming the pH is high enough to do what ever it is one needs to do with the base.


To quell the howls of protest in this misunderstood facet of buffering, I will provide a reference, Treatise on Analytical Chemistry, Kolthoff and Elving editors, Part 1 Section B page 458 Acid-Base Strength and Protolysis Curves in Water by S. Bruckenstein and I. M. Kolthoff.

Because of the problem with guessing the buffering proficiency I have entered some formulars of the BUFFER INDEX into Excel and calculate this index for planned buffers. Sorry, donīt have time to put the present examples through. The calcs are extremely simple with Excell, but one needs time to sort out which value goes where...
So: the buffer index is dCb/dpH or -dCa/dpH for bases or acids, respectively, according to JN Butler, Ionic Equilibrium, Addison-Weseley, Reading, 1964, p. 149. (Cb,a is the conc. of strong base, acid added to the buffer)
Now I was very much surprised to see, in some example curves, what Bill mentioned: the super buff. capacity of strong acids at low pH, etc....
But, that only holds for adding more acid to a strongly acidic solution. If you add base you drop to a low capacity range very quickly, especially at pH = 2. A robust method calls for buffering in both directions. But, luckily, chromatography is often quite forgiving.

Wonderful! I am glad that we are having this discussion. Usually, this board is rather quiet.
Anyway, Bill, I think a little bit of math homework will help. I got a spreadsheet on buffer capacity. If I plug in 10 mM of a strong acid with a pK of 0 (which gives me a pH of 2), I get a buffer capacity of 0.0002. If I do the same thing for a 10 mM buffer at pH 2 (like a phosphate buffer), I am getting a buffer capacity of 0.005. This is about a 25 times difference. When I get back either tomorrow or after tomorrow, I'll give you the formula, the references and an explanation. Maybe HW Mueller will do so. Of course, you also have another problem. I can make my phosphate buffer at pH 2 at any concentration that is reasonable, but you can't do that with the strong acid.
Anyway: a buffer is a buffer if it buffers the pH. If it doesn't, it isn't.

Ewe, no need to send formula, references and explanation. I am well familiar with them. The first reference on the subject that I am aware of is D. D. Van Slyke, J. Bio. Chem. 52(1922)525.

To summarize my points:
1. Strong acids and bases, by themselves are good buffers at high and low pH, 2 and lower, 12 and higher.

2. The LC separation does not need much buffer capacity.

3. Therefore, strong acids and bases certainly can be used to buffer the eluent 12 and higher and 2 and lower, and probably 3 and lower and 11 and higher in many cases. A bit of phosphoric acid buffers an eluent just dandy for aliphatic acids. Preparing such an eluant from phosphoric acid and dihydrogen phosphate is unnecessary and more inconvenient. I have no experience at the high end, but the principal is the same, to answer the question asked at the begining.


I thought you might be disagreeing with these points till I got to the last sentance.

Certainly if I needed a high buffer capacity at pH 3 I would use an acid and a salt. At pH 1 and less and 13 and greater, nothing beats the buffer capacity of a strong acid and base respectively.

Actually, my statement was not quite correct. I don't have the time to fix it now, but it makes no difference if I use 10 mM HCl or 10 mM H3PO4. My program is missing the formulae for the pH extremes.
I still disagree with the statements that strong acids can be used to pH 3 or bases to pH 11. You don't consider a 1 mM concentration a strong means to control the pH, do you?

Gave the whole thing some more thought and time.
Uwe, that explains some, I just wanted to ask you about your formulars, as those lifted from Butler (ref. above) gave the following buffer indexes, all at 0.01 M:
Dissociated acid (Ka = 0): at pH2 ~ 0.023
Acetic acid: at pH2 ~ 0.023
at pH4.6 ~ 0.0056 (~at its "best")
Ammonia (a 10mM aqu. sol. has a pH ~ 10.6):
at pH 10.5 ~ 0.002 (not bad!)
at pH 9 ~ 0.005
at pH 8 ~ 0.001
at pH 11 ~ 0.0027
Ammonium Acetate: at pH 4.6 ~ 0.0056
at pH 5 ~ 0.0053
at pH 6 ~ 0.0012


The formulars are solutions of the differential given above, they are actually simple arithmatic but much of it, so if someone would like to get them, I can try sending them via e-mail.
The thought on this:
According to Butler, the buffer index was introduced by van Slyke and is sometimes called the inverse slope (compare Billīs differential). That brings me back to the titration curve, which has its strange form due to the water equilibrium.
Thus: isnīt the buffering of strong acids at low pH etc. etc. due to the water (also due to non-ideality, especially at higher conc.)?
The pH of a 10 M dissociated acid is (neglecting activities) -1, what is that?
Anyway, how much chrom. is done at pH below 2 and above 12?
So. I would tentatively modify Billīs points:
1. Strong acids and bases are buffered by water if you stay at low or high pH as mentioned
2. Many chroms, even of ionizable analytes do not need buffering, many only little, some as much as you can get.
3. Skip.

Hans

PS: I am having trouble with the following: Why is everybody mentioning the non-ideality, namely the deviations that require activities, but then decides to ignore this when it comes to calcs like those of the buffer index. Certainly with modern PCīs it should not be a problem to use activities. Anybody have equations? Refs?

First, to fix my previous statements: at pH 2 the buffering capacity of a 10 mM acid solution is 0.023. This is roughly 4 times stronger than the buffering capacity of a 10 mM phosphate buffer at pH 7, which is 0.0058. The buffering capacity of the 10 mM phosphate buffer at pH 7 is about the same as the buffering capacity of a strong acid at pH 2.6. The buffering capacity of 10 mM acetic acid at pH 3.5 is at 0.0019, about the same as 10 mM ammonia at pH 10.5.
To Bill’s point: how much buffering an LC separation needs, depends on what and how much you are injecting. This may not be only a question of the analytes, but also the sample medium. If your sample stems from a dissolution test, it will be dissolved in strong acid. If the retention of your analyte depends on the pH, you either need to adjust the pH of the sample to the pH of your mobile phase (assuming that it is not very acidic), or you need a strong buffer. In other cases, you may need a good control over the pH of your mobile phase, and this may be best accomplished using a buffer. An example is the retention of a compound, whose pK is around the pH of your mobile phase. Under these circumstances, you get a lot of retention differences with small variations in pH.
To Hans: it appears that the buffer index is the same value as the buffer capacity. My reference is a German book by Fritz Seel, “Grundlagen der Analytischen Chemie”. A lot of HPLC is done at pH 2-3, because silanols are largely eliminate at this pH. On the other hand, I am a strong advocate of using pH to change the selectivity of a separation. Of course, this works only if your analytes are ionic, and you work in the range of the pK of the analytes.