basic question

[mk12]

I am asked to make 60:40 Water:methanol mobile phase. I am not supposed to mix them on volume basis I am asked to make on weight basis.
how to do that?

[juddc]

Weigh 600 gm of water into one vessel then weigh 400 gm of methanol into another vessel, then combine the two aliquots. Actually, if you're careful there's no need to use two vessels - just weigh it all into one.

Alternately, combine 600 ml of water with 400gm/0.79gm/ml = ~506 ml of methanol and you'll have the same solution, but I never told you that.

[mk12]

thanks

[mk12]

thanks

[Consumer Products Guy]

Most likely someone became attached to the principle that (all else being equal) weight is considered to be a more accurate way of measuring than volumetric. However, considering that the composition of the mixed mobile will change slightly over time due to selective evaporation, that part is moot. As long as the directions are specific and not ambiguous(like they are in this case), and you follow them, no problem. I'm starting to wonder if adding buffer acids and salts on a weight basis isn't more repeatable than using pH electrodes to go to a pH, that always seemed a little strange to me, and oftentimes not specified whether the aqueous part was taken to that pH, or after mixed with organic, which brings along issues of its own....

[juddc]

FWIW, I never adjust the pH of a buffered MP (bulk). I figure out how to make it such that it ends up at the desired pH, then check a small aliquot of the aq phase (only) to be sure that I did what I thought I did. I'm rarely more than 0.1 pH units off, which won't impact anything in a properly designed separation anyway. As long as one is working within the proper buffering ranges of the stuff they're using, I find this technique to be quite rugged. I don't want a pH electrode near my (bulk) MP anyway...ick.

[Uwe Neue]

I do not agree with the preparation suggested above. Please verify with the person who gave you the instructions about what you need to do.

Most compositions specified in LC are based on volume. a 60/40 mixture means to me to mix 60 mL of A with 40 mL of B. If I am asked to make this mixture by weighing out the components, I will look for the densities of the components, and calculate the weights needed to make a mixture with a 60/40 v/v composition.

Water has a density of about 1 at room temperature, so you use 600 grams of water for 600 mL. For getting 400 mL of acetonitrile, you weigh out 400 [ml] x 0.79 [g/mL] = 316 mL of acetonitrile. When you mix both, you will end up with a volume that is slighly short of 1000 mL due to contraction. (I do not have the exact densities in my head, so take these values as approximate values, and find the real values in Google).

For high-precicion mobile phase compositions, we do exactly what I just described.

[DR]

I'm w/ Uwe on this. My first inclination is to measure out 600mL water, add to MP container, then measure out 400mL MeOH, add to MP container, then swirl & He sparge the mixture. The problem to be avoided is to assume that volumes are additive and add the MeOH to the water in the graduated cylinder. You would come out with too much MeOH that way (for proof, try measuring out 50mL of each into separate graduated cylinders, then pour both into a third graduated cylinder - you'll find yourself about 5% short of the anticipated volume.

[juddc]

Uwe -

Please check your units.

400 ml x 0.79 gm / ml = 316 gm.

While I disagree with w/w preparation of mobile phase as well, I think my calculations were correct and I figured that he was working off of some established but admittedly unusual procedure. Also, he did ask a straightforward question, which I felt deserved an answer.

I should have simply told him to make a 54:46 v/v water:meoh solution, I suppose, as that would be the same and more easily transferrable to a gradient mixer, should that ever be desired.

[Uwe Neue]

Right:

400 mL x .79 g/mL = 316 g.

Sorry for the typo. But the remainder of the statement stands.

[juddc]

I don't disagree with Uwe at all, far from it.

My concern revolves around whether this is a pre-existing method. If it is and we tell him to use the conventional (v/v) method and simply to combine 600 ml water with 400 ml MeOH, his chromatography almost certainly will be significantly altered.

I'd much rather tell him how to do it in the conventional manner using a v/v recipie that better approximates what was done in w/w; i.e 600 ml of water + 506 ml MeOH.

[mk12]

If I am using isocratic method than mixing with wt basis is ok but if writing a gradient method mixing on wt basis is not practical i think.

[tom jupille]

Correctness is nice, but consistency is more important.

A well documented method should either explicitly describe the mobile phase preparation procedure or reference an appropriate SOP ("Standard Operating Practice"). In ambiguous cases like this one I would try to find out what the original developer intended (descriptions like "60:40 Water:methanol" usually refer to volume ratios by addition), and then append the appropriate procedure to the method to make things clear for whomever follows you. If that's not possible, then do it by weight (being mindful of density issues). Then, when you're sure that it works, append the appropriate procedure . . . .

[wanda50]

Always this same discussion over MP prep. The convention where I work which is not in SOP form as suggested by Tom, is to use actual volumes and mix in a separate container. Every time a new person comes in, we have to tell them what is meant by the method, even though our methods specifically state 400 mL MeOH and 600 mL water.

I have seen people make aq. buffers in a graduated cylinder, and often wondered how they measured it all. Then I wonder if they actually got all that buffer out before I have to use the cylinder.

[Victor]

The point of all this is that it is much more accurate to weigh out mobile phase components than it is to measure them volumetrically. Many measuring cylinders claim an accuracy of +/- 1% which means that if you measure out a volume of 500mls of water with a 500ml cylinder you could be getting 505mls or 495 mls. A decent balnce can do much better than this. I guess an inexperienced operator could do worse than this with a cylinder.... It is actually interesting to measure out what you think is 500mls in a cylinder and in parallel weigh the cylinder before and after dispensing the 500 mls- you may be surprised at the error, or you may revise the way that you have been looking at the liquid meniscus.

On the other hand I would guess that most people would use a binary solvent delivery pump to make 60:40 v/v methanol water on line. Am I right on this? I would presume that most modern pumps could do this much more precisely than a manual cylinder method. However, I do not know what accuracy is achieved by such pumps-somewhere around 0.5% I seem to remember. But if in any doubt I would use weighing as a gold standard method, as documented by Mr Neue.