I was told that dioxane should not be injected onto the MS. What would be the reason for this restriction?

I just got started with LC-MS. It seems that ammonium acetate and ammoniun formiate are the only buffers compatible with LC-MS. However, they result in decreased selectivity compared to non volatile buffers. Ammonium carbonate, the other volatile buffer and which gives better selectivity is dramatically unstable. Sometimes preparing a new mobile phase with this buffer result in a totally different separation. Only few hours after preparation also lead to tremendous changes in the separation. Are there any clues MS compatible buffers or mobile phases, which can also give good selectivity?

I'm afraid I can't add much to your first question other than to suggest that perhaps the reason stems from the low volatility of dioxane. But in the case of your second comment, while there are definitely a few other possible options such as the corresponding methylamine or dimethylamine salts, the choices are definitely pretty limited. In essence, you are limited to salts of weak acids and weak bases both of which must be rather volatile. This doesn't leave you with very many options. By the way, most likely your problems with ammonium carbonate are related to the tendency of this salt to be present as ammonium carbamate (if you look on your reagent container label will most likely mention this). Since the ratio of ammonium carbonate and ammonium carbamate is dependent upon solution conditions and generally present in the crystalline form supplied, one can expect changes in this ratio to occur over time until equilibrium is established.

On dioxane:
Here are several issues that might be directly related to the MS, not including any chromatographic considerations: (1) With some ion sources, there is PEEK tubing built into the ion source, between where you hook up to the LC and the actual spray needle. PEEK is incompatible with certain organic solvents. Check a PEEK compatibility chart. (2) Dioxane is flammable. If compressed air is being used as a drying gas, rather than nitrogen, there could be an explosion hazard. Check to see what is being used as a drying gas. (3) If your sample is being flow-injected or infused directly in dioxane, you might not get much ionization. I've never tried dioxane, but I've gotten little or no signal when attempting to infuse a sample in THF (for example).

On buffers:
This depends on what sort of compounds you're analyzing. I find that many compounds chromatograph well with 0.1% acetic or formic acid, on a modern endcapped C18 column. If these acids or their ammonium salts don't provide satisfactory performance, there's always trifluoroacetic acid (TFA). But beware that it will suppress your signal in positive mode, and completely kill it in negative mode. The amines that Chris mentions can also be used on a very limited basis, but be warned that they will kill your signal in positive mode.

Formate and ammonium formate, acidic acid and acetate work well in the acidic pH range. ammonia and ammonium bicarbonate are good for the alkaline pH range. There is no MS compatible buffer for pH 7, and there are difficulties at very low pH. You can use TFA, but ion suppression is common.

Now I got some questions too:
What about MFA or DFA or dicarbonic acids instead of TFA?
Has anybody ever tried some other organic acids with lc/ms?

Thank u for all ur comments. Personally I also suspected the low volatily of dioxane (also THF?). The fact is that these organic modifier usually result in a drastic improvement of the separation. Can any one suggest other solvent which have the same properties, as the commonly used solvents (Acetonitrile and methanol) often give less good separation. Any clues to improve the separation. I'm analyzing basic compounds.
On the other hand, I believe that TFA is often used if one needs a low acidic pH. Is it recommended to the adjust the pH at higher values, using ammonia (thus, resulting in a ammonium salt of TFA)?
Moreover, other clues to improve the selectivity, thus allowing the use of the common solvents and buffers?

Furthermore, I would like to know if it is the acid or the base, which should be used to calculate the molarity of the ammonium formiate or ammonium acetate solution. By the way, I was told that the overall concentration of the buffer in the mobile phase should never exceed 10 mM for LC-MS purpose.

A couple of comments to:

Andreas:

A bit of caution on your proposed alternative volatile acids: fluoroacetic acid is extremely toxic with a lethal dose for humans estimated to be in the range of 2-10 milligrams per kilogram! The most toxic route is inhalation with cause of death being cardiac arrest! Interestingly, it occurs naturally in Gifblaar which is reportedly the most toxic plant found in Africa. Difluoroacetic acid is significantly less toxic (a lethal dose being 100-180 milligrams per kilogram) but this might still be too hazardous to be considered usable for such an application. Furthermore difluoroacetic acid is fairly expensive (10 ml of 97% purity acid is $137 from Fluka).

Jean:

I think the general practice is to calculate molarity based on the minor component (assuming one of the components is present in excess). Regarding solvents, I would think that THF or acetone would be much better choices but I don't know if they are viable from a MS point of view.

Jean,

I've read some papers on lipid analysis by LC/MS, where chloroform, isopropanol, and acetone were used (not necessarily in the same method). Some of these used APCI, not ESI. See the above caution on flammability. It's mass spec lore that some people have launched their ion sources.

Even with volatile buffers, you may start to see some suppression above 10mM. But I wouldn't say to never exceed 10mM. If you've got plenty of sensitivity, you can go higher if needed.

Thank u Chris for ur comment. More clearly, the question is as follows: The pH of 0.2M ammonium acetate solution is about 7. If I need a higher pH, I should adjust with ammonia and consequently, the concentration of ammonia will increase. And if I need a lower pH, I will add acetic acid and the concentration of acetate will increase. It seems that if I'm investigating the effect of pH (pH 6 to pH 8 for example, there will be a biais due to subsequent change in the molarities of both 2 components. Isn't it? Instead of using the ammonia actate salt directecly, shouldn't I use separately ammonia and the acetic acid. For example, I think about measuring the necessary amount of concentrated ammonia to make 0.2M (after bringing to volume in a volumetric flask) and adjust the pH using formic acid before bringing to volume. In this case my solutions will really be 0.2M ammonia with different amounts of formic acid for different pH. I can also do it by fixing the formic acid concentration at 0.2 M (measuring the necessay amount to reach this final concentration in a volumetric flask) and adjust the pH by adding different amounts of ammonia. Although this procedure seems to be heavy, I would like to know about the right component that I should fix at 0.2M.

The molarity (or normality) of a buffer should be based on the itme that buffers. The molarity of ammonium formate at pH 3.7 should be calculated based on formate. The molarity of the same buffer at pH 9ish should be based on the ammonia concentration.

Jean,

I think this is more of a tactical question than a technical question, that is it all depends on what it most convenient for you. It's just that when reporting the concentration used one must calculate this based on the concentration of the least concentrated ion. My suggestion, to save time, is to make stocks of ammonia and formic acid, prepare known concentration mixtures, check the pH and following that just reproduce the concentrations without rechecking pH (since that shouldn't be necessary with high purity reagents).

Chris, this is not a tactical question, this is a question of what buffers. Therefore I completely disagree with you. If one makes an ammonium acetate buffer at pH 4.75, the acetic acid/acetate concentration is what makes the buffer, and it is there at double the concentration of the ammonium ion. Therefore it is an acetate buffer at the combined concentration of acetic acid and the acetate ion. If I am making an ammonium acetate buffer at pH 9.25, the ammonia/ammonium concentration is responsible for buffering and is there in excess, and I need to define the buffer based on the concentration of ammonia and the ammonium ion. Otherwise things don't make sense.

There are very rare exceptions to this rule. A while ago, I needed to study a phenomenon, where it was best and most convenient to keep the cation concentration of anionic buffers constant. In this case, the procedure of the buffer preparation needs to be very clearly specified.

Uwe,
are you talking about the same cation concentration at diff. pH of your buffer? That may take some neat calculation unless you add a large excess of cation as a neutral salt?

Uwe,

The common practice when describing buffered systems is to use terminology such as the following: prepare a 20 mM ammonium acetate buffer adjusted to pH 4.75 with acetic acid. In this case, the amount of ammonium ion is known but the exact amount of acetic acid is generally unknown so it's not possible to express the buffer in terms of the concentration of the ion in excess. My point was that when making up buffer is from the individual acids and bases one should replicate the process commonly employed when preparing buffer solutions. Admittedly, when I prepare buffers I don't use this practice as I prefer to prepare buffers by weight. The buffers I use in my work are specified in terms of grams of both the acid and the base diluted to 1000 g in water. This allows one to specify the concentration of both the acid and the base if such information is deemed relevant.

Chris,

I guess that these things really depend on the practice, and I do generally not agree with the procedure that you have outlined. In my world, a 30 mM potassium phosphate buffer at pH 7.0 is prepared by mixing a solution of 30 mM potassium dihydrogen phosphate with a 30 mM solution of dipotassium hydrogen phosphate. Or for a 20 mM acetate buffer at pH 5.0 you mix a 20 mM solution of sodium acetate with a 20 mM solution of acetic acid. The only issue with this approach is that you do not always get exactly a liter if you start off with 500 mL of one solution, but this is for the most part not relevant in LC.
Oh, and the individual 20 mM solutions are prepared by pipetting from a more concentrated solution that is stored for multiple use. This keeps it simple as well.

Best regards
Uwe

I have to prepare 10mM of acetate buffer pH 4.0. Should I prepare 10mM of acetic acid and 10mM of sodium acetate (i.e. is the molarity of the buffer an additive property of the individual components)or should I prepare 10mM of each component(i.e. is the molarity a function of the molarity of the salt)

Usually the buffer concentration refers to the salt. So you'd make up 10mM acetate and adjust the pH to 4.0 using acetic acid.
john

But is it not true that in the acetate buffer some of the acetate ion will be contributed by the acid and some by the salt.
Therefore, the sum of all the acetate is used to describe the molarity of the buffer.

there was a mistake I made in my question posted on 25 March. When the property is additive the molarity of the individual components should have been 5mM each.

All true but when a method describes 10mM buffer it always means 10mM with relation to the salt as far as I know.