Mobile phase and acetonitrile mixing

[dragosb]

Hi,

I have a basic qustion concerning mobile phase delivery.
Is it better in a gradient to use two mobile phases with different acetonitrile concentrations (considering that the ion strenght and the buffer used is the same in both mobile phases) or is it better to use the pump to mix the 100% acetonitrile with a mobile phase with no acetonitrile in it in the desired ratio.

Second question:
Is it a good idea to switch the flow abruptly after each sample (ex. from 1.5 ml/min to 0.4) without a progressive flow change.

Third question:
I have a change in the retention time from one sample to another eventhough pressure remains the same, temperature is controlled in an oven. The change is of more then 2 minutes and is random, I can not say it is getting longer or shorter in time. Please let me know what else I can check to eliminate this problem? Might a column conditioning change anything, or should i just let some mobile phase flow at a low flow rate for a day?

Thank you

[danko]

Ion strength and buffering is a separate matter than the dominating – in your case – issue. So maybe it belongs to another discussion. Generally speaking it might be an advantage to reduce the ion strength during the gradient and it might be a disadvantage.
Mixing 100% organic component/liquid (acetonitrile or whatever) in water (aqueous phase) may cause considerable air bubble release and worst of all exactly where it is extremely harmful namely in the GPV (in low pressure mixing systems). Consequently you lose control over the gradient composition – hence retention time, separation etc.
I would never do that once the method development is completed and the parameters determined.

Changing the flow rate from 1.5 to 0.4 mL/min is completely harmless – both to the column and the system. So don’t worry. If – on the other hand – it has something to do with the retention time variation, it might represent a potential factor but in the following ramp-up phase – i.e. when the/if the flow is to intended to go up to 1.5 mL/min reproducibly. Again – I personally – wouldn’t do that kind in a sequence.

The third question might become irrelevant once you’ve corrected the above issues. Otherwise,
Best Regards

[paulw]

When you are using a buffer in the mobile phase, I prefer to use two premixed mobile phases at the respective ends of the acetonitrile concentration. When you pump mix the acn and water you can sometimes get a salting out effect from a composition that if premixed is quite soluble.

[Consumer Products Guy]

Drag - I read Danko's response, and I'm answering below with my observations over 30 years of HPLC, with Agilent 1084, 1050, 1100, and 1200 units

1. We typically pump in acetontrile 100% from the "B" reservoir, either with isocratic or gradient elutions, and in our validated procedures (cGMP and GLP).I have a basic qustion concerning mobile phase delivery.

2. In the old days of columns this was much more important; today's modern columns can handle this better.

3. For varying retention times I would try a longer post-run equilibration time between injections. Typical re-equilibration is 5 to 10 column volumes of the inital concentration of the solvents. Note this is in column volumes, but you set the post-run time accordingly, in minutes. And make sure no one has bumped the on-off switch on the degasser. And that ALL FOUR channels have been burged that day before starting/equilibrating, even if only two are used to deliver solvent for your assay - this is important.

[dragosb]

Hello Consumer Products Guy,

What do you mean by burged for the 4 channels?

Thanks for the answer

[danko]

To Consumer Products Guy.
Ad 1. It is a well known fact that mixing organic and aqueous phases (f.x. ACN and water) results in air bubbles release. So if you say you never experienced a problem in that context it could either be because you didn’t attribute related problems to it or you’ve just been lucky.

Ad 2. Neither in old days or now it has been a problem to go from “flow” to “zero or very low flow” rapidly. It could be no more than the opposite – i.e. if you go from “zero flow” to relatively high flow in a rapid manner. And yes, this is more a theoretical than a real concern especially now-a-days.

Ad 3. If the retention times fluctuate randomly (i.e. get shorter end longer in an unpredictable fashion) I wouldn’t worry about the adequacy of equilibration time. It would be a concern if the retention times are predictably shifting to a shorter time - but stable after a couple of injections.

To Dragosb.
I’m sure it should have been purge instead of burge – typo I’m sure.

Best Regards

[AA]

I cant resist tossing in my comment on this:

Ad 1. It is a well known fact that mixing organic and aqueous phases (f.x. ACN and water) results in air bubbles release.

True enough, however this only happens at atmospheric pressure. When using binary, high pressure mixing pumps it is not an issue. There is a small effect on solvent compressability, but it is minor. Low pressure mixing systems, usually quartanary systems, can suffer from this issue. This is why most decent LP systems have in line degassing (we used manual vaccuum degassing followed helium spage in old days). High pressure system do also gain a small, longer term performance gain from degassing by eliminating that compressability issue.

I am not surprised that CPG has never had an issue with this, neither have I, but probably like him, I have always used some kind of degassing (manual, sparge, in-line) and taken care of my solvents.

Going back to the original question: I always try and use pure solvent in my bottles, 100% water or buffer in A and 100% ACN in B for either gradient or isocratic.

AA

[DR]

::tosses in $0.02::

Baselines tend to be quieter if a (for example) 20-80% A:B ramp is used than if values that are close to the limits (2-46% or 25-97% A:B, for example) get used to avoid premixing mobile phase(s). Things get relatively noisy when either pump has to run at a small percentage of total flow. Also, even if outgassing is not a problem, you will have temperature changes in the mobile phase and this can be a problem, especially if using a given method on different LC types. Low pressure mixing systems are more prone to outgassing, but a bit less noisy due to a little higher dwell volume (sllows for a little more mixing) while the opposite can be true for high pressure mixing systems.

Then there's the issue common to sapphire/ruby check valves and 100% ACN - not a good combination over time, so if you choose to disregard the above and drop your "B" line into a 4L ACN jug, at least use ceramic check valve cartridges.

As to pumping rate variations - with modern pumps, they tend not to just whack a column with an extra mL/min the instant you command them to. They suually ramp things up at pretty reasonable rates to prevent column damage (not that modern columns are as susceptible to this as older ones were).

RE: wandering retention times - provide more information about your mobile phase, please. Generally, if there is some column conditioning going on (common if you add -for example- triethylamine to mobile phase), peak retention times should generally wander in just one direction until things are "broken in". If the same peak can't decide when to elute, you probably have a check valve, mixing or outgassing problem.

[dragosb]

Thanks DR,

I use a Na2HPO4/NAH2PO4 buffer for my mobile phase with 10 mM buffer concentration, pH=6.8.

So this is the mobile phase that I mix with ACN.

Thank you and hope you can further help me with my issue

[lmh]

Perhaps some explanation useful for the less-experienced? (sorry if this is patronising!):

(1) Outgassing. Gas can only come out, if it's already in. If you've used a good online degasser, then there shouldn't be much dissolved gas in the aqueous and acetonitrile phases, and when they're mixed, although the solubility limit may be less than it was, the amount of gas that's present is still soluble. This is particularly true at the very high pressures the pump is creating.

(1b) But yes, purge your lines. Gas can redissolve back into the solvent lines, through the walls of the tubing, between the degasser and the pump. Purging also gets rid of any bubbles that have managed to appear in the last week that the instrument wasn't used...

(2a) Sudden pressure changes: a sudden pressure change is bad mechanically for a column, if the particles can move in response to it, creating voids. Modern well-packed columns with spherical particles are pretty much optimally in place (a bit like marbles arrange themselves on a tray), so a sudden pressure change won't mess them up. Older columns with irregular silica could never pack that way, so they were more vulnerable. Of course some people are still using irregular silica today, and these people, together with those using other less usual packing materials, should be more careful than the rest of us.

(2b) Any sudden pressure change is bad, but sudden decreases don't really happen so easily, because the pump doesn't encourage back-flow. If the pump stops, the pressure in the system has to be relieved by liquid flowing away through the column, and generally the system will have a pulse-dampener and some elasticity to be relieved. On suddenly starting the pump, it would be possible to raise the pressure more quickly if the pump is quick enough, but again it depends on the capability of the pulse-dampener.

(3) Pumping with pre-mixed solvents for the start and end of the gradient. A gradient pump has a certain accuracy and precision. Maybe when you specify 25% it could be anywhere from 24-26. If you want to run a (rather silly) gradient from 24% to 26% and you do this by putting 0% on channel A and 100% on channel B, and specifying 24-26 in your gradient table, you could actually be running a gradient anywhere between 23 and 27, and you don't really know whether it spans 2%. If you put premixed 24% on A and 26% on B, and specify in the gradient table 0-100%, the pump's +/- 1% precision is now suddenly improved 50-fold, because a 2% error as it passes through x% is now 2% of the difference between 24% and 26%...

(3b) But this depends on how well you mix your samples, and relies on solvent bottles not standing around long enough to evaporate significantly (which changes the mix). Labs who are good at mixing, and who habitually run the same gradients, will probably pre-mix. Labs who run a huge variety of wildly different methods are likely to use 100% ACN and rely on their pump in all but the most critical situations. No one wants to be throwing away half-used batches of expensive ACN-based mixes, or using buffers made up weeks ago.