Chromatography Forum

First of all, "20% ACN in aqu. solution" is not a simple statement, as one doesn´t know what is meant by it. If one states 20%ACN + 80%H2O, v+v, one comes closer to a simple repeatable recipe. Usually what is meant by this is that one added 20 v units of one to 80 v units of the other (not putting 20 in a graduated cylinder, then 80 on top!), and one should probably state it as such. Before you do something with this it should stand for some time anyway (to equilibrate) and should be brought to the temperture at which you want/need it.
If one adds 200 mL of ACN into a 1 L volumetric flask, then adds water to the mark one will be, strictly speaking, doing this forever (it should be at the mark only at 20°, or whatever is stated on the flask, etc.), but I guess it´s ok if one states exactly what was done, nevertheless, I consider that ridiculous. Certainly, you will get something different when adding 800 mL H2O to such a flask, than adding ACN to the mark.
If you ignore the temperature changes then why use volumetric glassware? If one neglects temperature changes here than one probably soon does it for MeOH-H2O mixtures, that´s getting more than sloppy.
All in all, in a publication a "simple" statement like the one you made above would be an attempt to keep things equivocal.

I would weigh 200mL*0.782g/mL=156g of acetonitrile and 800g of water into a reservoir. I can do this far more reproducibly than using a graduated cylinder. This is standard practice at Dionex. Modern electronic balances are wonderful.

I agree with Mark

The weighing of the components is the most reproducible way to prepare a mobile phase. When the ratio of water/organic modifier is a critical parameter regarding resolution , SST... then weighing yields in the most robust methods (and the use of a balance results in very small variations when testing robustness regarding variations of the ratio water/org. mod.). If you have many samples and much mobile phase is needed, the error of a large volumetric cylinder is relatively high, whereas the error of a balance in this area is much smaller.

Florian

Expanding a bit on HW Mueller's last sentence, that the question comes up at all illustrates the point that it is important to be as specific and explicit as possible when describing mobile phase preparation. The (American) "Murphy's Law"* has one variant which states "Anything which can be misinterpreted, will be misinterpreted."

*For our non-US members, I believe this is also referred to as:
"Sod's Law" (UK)
"Le principe du maximum enm*rdement" (France - sorry about my French spelling!)

I'll start a thread in the "Around the Water Cooler" section in case anyone wants to add other local variants.

There has been a wonderful description of the effect of different ways to mix solvents in Veronika Meyers English version of her German book on "Fallstricke and Fehlerquellen der HPLC in Bildern". I believe the English version was called "Tips and Tricks". I have written an "HPLC Troubleshooting Column" in the American Laboratory using her chromatograms. I will send you a copy, if you contact me.

You will get different results, if you start with water and add organic until the gradient is full, or if you start with organic and add water until the graduate is full. The best (and only valid) approach is to measure the water and the organic separately and then combine the two. However, this will give you a volume that is a bit less than 1 liter, even if you wait until the thermal effects are gone.

As a small side effect: due to the solvent contraction upon mixing, two perfectly working HPLC systems, one a low-pressure mixing system and the other one a high-pressure mixing system, will give you a different flow rate, and thus different retention with the exact same mobile phase composition. With premixed solvents, they both will give the same results, of course.