Best way to prime up a HPLC (Waters) before use

[EmpowersBane]

I run a gradient method using Waters 2695 HPLC with UV detector but I only use lines A and B for the gradient and line D for washdown post-run. Is it still better to prime up all the lines, A B C and D when preparing the system, or only the lines you use? I have heard a few different opinions on this, ranging from only prime the lines you use for 5 mins each, and also to prime out all the lines for 10 minutes, I'm looking for some advice from experienced users.

My own method when preparing a HPLC for a run is to first prime out seal wash for 20-30 minutes, prime needle wash 4 times, purge injector for 40-50 cycles, then wet prime the lines A B C and D for 12 minutes each, before slowly ramping up flow rate on initial conditions for gradient and leaving the flow rate equilibrate through the system for at least 60 minutes before running any injections. Is this too much/not enough?

[Hollow]

Hey

this sounds a bit too much for me. Of course, it depends on your applications.
And btw. a better indication than time would be the volume pumped for priming

For me, this is what I normally do for start-up:

- keep all lines primed (at least filled with solvent), because all lines share a single vacuum chamber inside the degasser. If one line would be empty, the efficiency of the degasser could suffer (degassing the empty line not your solvent...)
- so normally I do a wet prime with about 15-25 ml (2-3 min @7.5 ml/min) on lines with replaced solvent, and about 10 ml on lines without any changes. So about 10-12 min in total.
(watch the system pressure to see when the air bubbles/new solvent reaches the pump-head, then wait the pressure to stabilize with the new solvent)

- Priming seal wash depends if I have replaced the solvent or if the line still looks ok. Then maybe just a few cycles

- Injector: I normally do a purge with MeOH as part of the "goto-standby" sample set, so the liquid inside the injector is exchanged too. Then when preparing the system, I check the syringe for bubbles, and if necessary, do an injector-purge with methanol for 6-10 loops with activated compression check. Maybe additional measures need to be taken to remove any sticky bubble from the plunger. Afterwards I include an injector purge at the beginning of the samples set with the runtime parameter of about 7 (needs to be between 6.5-15).

- Needle wash depends on the application, maybe give it 3-4 cycles, an check if solvent sputters from the yellow line (no steady flow is normal)

Preparing for runs is also straight forward.
- as my columns are normally in acetonitrile, I start with 100% (or close) of organic (ACN or MeOH) for about 10 column volumes or longer (wetting all the pores of a C18 column).
-- normally I do this in "manual mode", either directly on the 2695 or from the control panel in run sample
-- ramping up the flow just in one or two steps (e.g. start with 0.3 or 0.5 ml/min until pressure stabilize and check for leak, then go to final flow)
--- during this step, I can write the sample-set table
- then either manually change the composition to initial composition, manually or with <equilibrate> function in the sample set
- next do an injector-purge (runtime 6.5 – 15)
- then have the gradient run once with the <condition column> function, so the first injection has the same column equilibration as any other injection
-- special caution has to be considered when dealing with salty buffers, which may precipitate. Then introduce an high aqueous intermediate step when going to high acetonitrile


- as part of the "goto-standby" sample set, I normally do:
-- flush my column with about 10-20 column volumes of acetonitrile or appropriate storage solvent
-- do a wet prime for about 2 min (with default parameter of 4) on 100% MeOH
-- do an injector-purge with 100% MeOH, runtime ca. 7
-- if equipped with a column selector and bypass line, let about 3-5 ml of MeOH flow through the detector as well. (Otherwise let it sit in ACN until the column is removed.)
-- stop flow
-- if not used for longer period, replace (wet prime) the aqueous solvents with MeOH, remove the sample vials and power-off.

Doing so, I rarely have any problems with sticky check valves or when restarting the system

[Multidimensional]

The Waters 2695 system has four liquid channels and FOUR SEPARATE vacuum degasser channels (*All the lines Do NOT share a single degasser chamber inside the degasser. They are separate, 0.5 mL volume chambers). The degasser channels must be kept full of liquid (if you do not use a channel, the liquid inside quickly evaporates, leaving only air stressing the degasser). Make sure the channels are chemically compatible with your mobile phase choice. Connect any unused degasser positions in-series to actively used channels (with a 'C' tube). This maintains flowing liquid inside them and also improves degassing efficiency.

Initial Waters Alliance specific NOTEs: ALL of the CHANNELS on the Waters Alliance 2690/95 (and 2670/75) must be primed and kept "wet" at all times. A syringe is very useful to get things moving from the bottles down to the degasser inlets. Wet Priming is best with the Waters brand system. They cavitate easily and lose prime when things are not running perfectly. A good initial prime is needed to keep the gradient proportioning valve (GPV) in liquid (for the seals) and to maintain fluid in ALL four vacuum channels (5 channels for the 2670/75). Do not leave any pure water or buffers sitting unused in the system. Flush them out after use, use water with some organic solvent (i.e. MeOH) to arrest any bacterial, mold or algal growth). If you do not need to use a solvent line, then be sure to flush it periodically to maintain the flow path with liquid (again, no standing buffers inside unless you want to contaminate the flow path or cause corrosion). No need to go nuts with long flushes and washes to get the system up and running. You are just wasting mobile phase. Use a logical procedure (know the volumes of the tubing and flow path, then wash several volumes of liquid through each one, ONE-AT-A-TIME, never 25/25/25/25 (a newby mistake). Set one channel to 100%, draw liquid through the system, pump it to waste and check for air bubbles or issues. If all is well, move on to the next channel. Start to finish, a 4-channel 2695 takes about 30 minutes to prime/flush and then start to equil a column (which will take as 'long as it takes').

Here is a link to a free article on general HPLC PUMP Priming:
"Tips and Advice for Priming your HPLC PUMP (or similar pumps, FPLC or UHPLC Pump) "; https://hplctips.blogspot.com/2020/09/tips-and-advice-for-priming-your-hplc.html

[Consumer Products Guy]

We used Agilent HPLCs (quaternary) with Agilent degassers.

I made a Method to pump each channel at 5 ml/min for 5 minutes, then do same with next channel, all with pure line open to waste. So this would take 20 minutes each morning of use.

While pumping an aqueous phase, I would check on the pressure to confirm that the pure frit was OK.

[Hollow]

The Waters 2695 system has four liquid channels and FOUR SEPARATE vacuum degasser channels (*All the lines Do NOT share a single degasser chamber inside the degasser. They are separate, 0.5 mL volume chambers). The degasser channels must be kept full of liquid (if you do not use a channel, the liquid inside quickly evaporates, leaving only air stressing the degasser).

yes and no.
You're right, each channel has its own chamber but as I remember, all four chambers are directly connected to one vacuum line/pump, without any valves. So they are not independent from each other.
Either way, keep all lines primed.

[Consumer Products Guy]

We used Agilent HPLCs (quaternary) with Agilent degassers.

I made a Method to pump each channel at 5 ml/min for 5 minutes, then do same with next channel, all with pure line open to waste. So this would take 20 minutes each morning of use.

I should add that when changing a mobile phase channel for use, we would purge that channel additionally, as the 1050 series degassers had a hold-up volume of 40 ml/channel. Later degassers used different tubing and their hold-up volumes were considerably less.

[Multidimensional]

Hollow, as stated earlier: [quote="Multidimensional"]The Waters 2695 system has four liquid channels and FOUR SEPARATE vacuum degasser channels (*All the lines Do NOT share a single degasser chamber inside the degasser. They are separate, 0.5 mL volume chambers).

The degasser's vacuum gas system is of course shared through one vacuum pump (common to most vacuum degasser systems. They use one pump). However, you stated that;"all lines share a single vacuum chamber inside the degasser", which I just wanted to be clear is false. For purging/flushing solvent through the HPLC system, it is important to know that INDIVIDUAL vacuum channels/chambers are used (not one shared chamber) as each contains it own mobile phase. That was the critical message to share, and that each line must be flushed independently on the four channels of the Waters Alliance 2695.

[lmh]

This, I think, is one of the biggest contributions of modern HPLC design. In the old days, I too primed each line for 10min or so, and that was after spending ages carefully filtering all my solvents; I would then monitor my pump pressure and re-purge at the faintest sign of ripple; I worried loads about my needle washes and flushing things, and it took an hour (at least) to get an LC system ready to use.
Now I struggle with users who intend to run 20 samples for 5 minutes each, and book 2 hours of instrument time. They forget that the autosampler may take the nearly a minute to do the injection, which means they've only got about 5 minutes to set up the system before their booking starts to clash with the next user. But the beauty of modern UHPLC's is that they've been built by engineers who have such an ingrained habit of minimising dead volumes and maximising speed of operation that it's actually possible to set them up in 5 minutes. Put the bottle on, press purge, and 2 minutes later it's purged, and really that's all it needs. I am so grateful for this; I'd hate to go back to the old days. It just seems so counter-productive to employ a skilled analyst and then have them spend all morning watching expensive solvent purging into a waste pot. On autosampler needle-wash purging, for example, our oldest Shimadzu Nexera autosampler is now about eight years old I think, and has been purged once in its life (when the engineer installed it, the engineer's instructions being that it would never need purging again, provided we kept it in a solvent in which things won't grow, and didn't let it run dry). Meanwhile, our Acquity sample manager (Waters) does benefit from being properly purged, but it takes about 5 minutes, maybe just a little more, for the necessary 2-3 cycles, which can be done simultaneously with purging the pump, so the whole thing can be set up in ten minutes if you get a move on and you're worried about an impending tea-break! We owe HPLC designers a lot.