Chromatography Forum

hi everyone

im using an agilent 1260 series and im having issues with my quat valve/pump . i cannot get the same pressure profile.
i run 100 % (very low conc phosphate buffer) and then as soon as i introduce B (acteonitrile)its wavy. obviously the peaks are shifting as a result.

things ive tried:
-the quat valve has been removed and cleaned (multiple times)
-with a different quat valve i cleaned all the lines with 0.1 mol HNO3 and ran the gradient again with no columns and a restriction - ok
-with the different valve i ran the gradient with old columns - ok
-got back my quat valve, ran the gradient with brand new columns - bad
-the quat valve has been checked externally (with a gradient) and the pressure increases linearly across the entire gradient. (water/water with acetone)
-pump pressure fluctuations are <1%.
-completely purged lines at 5ml/min
-removed all lines and sonicated in 10% hno3. including frits
-replaced purge valve frit
-tried the same gradient with different channels (B/C and A/C) same wavy profile.

i have had issues with algae in the past. so i am going to clean my valve with HNO3 and repeat the same steps.

does anybody have any suggestions what to go to next because i am totally out of ideas.

below is pic of the pressure profiles of 3x injections one after each other. just MQ water and running gradient



if image doesnt work here is the link: https://imgur.com/a/fFSNOG0

Extra info: the valves are not leaking (switch off the quat pump module and all the valves automatically close - no leaking)

Hi Kate,

You've tried a lot of the things I would have suggested, but have you confirmed that your pumps are pumping at the correct flow rate? I checked your image, and it looks very interesting-- what does the black line represent? The method itself?
It seems like something is slipping and then returning to it's normal state in relatively regular intervals. If you are concerned about microbial growth though, just take some extra from your waste (probably not the line, it's probably filthy, but maybe a fraction collection?) and determine if microbes are present/growing in the solution. Whenever I have issue with potential contamination I just throw on the IPA and run it for a few hours making sure to keep the purge open first prior to flushing the entire system (just incase the growth is in my frits or somewhere in between which happened to me during my last PM.... thanks Ivan)

I retired before Agilent 1260 systems were out there. But on 1050, 1100, and 1200 systems we ALWAYS purged ALL 4 channels 5 minutes at 5ml/minute, even if we were only using two channels. That would be each day of use, at start up.

With the old 1050 degassing unit, the hold up in the degasser was 40 ml, so when changing solvents we had to also account for that.

If you run without a column, only a restrictor in place (< 40 bars backpressure at all times), what happens? Basic troubleshooting should guide you.

Also, please STOP with all the nitric acid rinses and cleaning. NONE of these are ever needed unless you are back in 1980. You are likely to do more damage. MeOH or IPA is fine. Make sure the system has been properly serviced AND TESTED afterward. Check and recheck. Poor quality degassing, sticking valve, obstructed solvent pickup may contribute to cavitation. If the pump does not run 100% perfect, then no analysis is worth running. Check components with a restrictor only, running one channel as at time, initially, to verify operation. If you have a spare MCGV, swap it and document any changes. Keep it simple and troubleshoot/change one thing at a time.

BTW: The profile you show (though I have no idea what the running condition are) appears to show that your pump has a leak. See how it drops off as the second piston refills? Please service the pump (new seals and maybe new pistons too), then run a pressure leak check on each channel.

Hi all

thanks for the replies!

so to rule out a few things:
-i cut off the time from my image. therefore its not super clear. the gradient is running over about 50 mins with a flow of 1.3 mL/min. no pump leak (as someone said as the second piston refills) the pressure change is over 10-15 mins not seconds.
-I've flushed with stronger solvents (IPA i.e) and still had growth. thats why i switched to something oxidising like HNO3. it is only dilute. < 1wt% (about 0.15 M) and its only flowing through the MCGV not the pump. Everything in there can handle easily x10 more concentrated HNO3. i am not worried about damaging anything.
-i never use channel D. its plugged. so i never flushed it. i am wondering if that is the cause of my problems, so ive also flushed this with acid and water. for the other channels i purge them all before i run for 10-15 mins each.

i will test the flow rate today. each channel at a time with a restriction. i checked with a restriction and the gradient looked fine (using a spare MCGV). i want to compare it today.

-i never use channel D. its plugged. so i never flushed it. i am wondering if that is the cause of my problems

Sure doesn't help !!!


Also, whenever I did gradients, I always experimented with long re-equilibration times (yes: re-equilibration volumes) and only after establishing enough re-equilibration time did I reduce that for more throughput.

It is most likely unintended channel D's "contribution" to the mix that is at the heart of your issues.
Air finds a way, especially in low pressure mixing systems, which is why purging all of the lines is so important. Just a little air bubble from line D getting into the wrong spot can cause unreliable flow in the proportioning valve or in a pump head.

hi everyone

so i swapped to a new valve and brand new columns and for a few injections it was fine. but now the problem has returned...

same thing as before. any ideas????

kate

hi everyone

so i swapped to a new valve and brand new columns and for a few injections it was fine. but now the problem has returned...

same thing as before. any ideas????

kate

Do you mind showing what the gradient looks like when it's "fine" vs when the problem returns?

Edit: is there a reason why you are hand drawing these pressure traces and not doing screenshots?

Please give us the information about the chromatographic parameters (gradient table, flowrate, column, temperature) as well as the full pressure traces incl. axis-scales.

At the top, you wrote something about 90%B. Is that meant to be 90%B at the top or going to 90% after the top?
if 90%B would be at the top, then something is fishy here, as the pressure trace would not correspond to aqueous/acn mixtures.
If at the top is only 10%B, which would be plausible from viscosity point, and you tell, that this is after about 50 min, then the change is only about 0.2%B/min, which may be to flat for the pump/proportioning valve to reproduce/mix well.
Then it may be better, to prepare eluent B as 10% acn in solvent A and run the gradient from 100%A to 100%B. if a column flush is needed, then use high acn as eluent C for that.

another thing to check is the compressibility settings for your eluents. If I remember correct, this should be done with a special calibration capillary from Agilent (but I'm not that familiar with Agilent systems).

hi all

here is an overlay of normal vs not normal pressure
light blue, green and brown are not normal. you can see the problem starts with the introduction of ACN.
dark blue and pink are normal.

i included solvent ratios. the absolute time doesn't matter but i can tell you that its over 2.5 hours injection time. flow rate is over 1 ml/min. pretty bog standard reversed phase column.

it starts at 100% A which is water with 1% ACN.
B is ACN

https://imgur.com/UeLF3uW

i realised i cut off the pressure but its around 200 bar. the range from the peak to the start is about 20 bar.

it starts at 100% A which is water with 1% ACN.
B is ACN

Just sayin': in my career, I never started a gradient with 100% of a channel. Let an expert comment...

Just sayin': in my career, I never started a gradient with 100% of a channel. Let an expert comment...

we have ran this method for many many years this way. it has never been an issue until now.... so i dont think that this is the route of my problems

i ran my gradient with a restriction after the pump (nothing else in the flow path) and here is the result:

https://imgur.com/I6XIPPw

2 injections (black one i stopped manually)

i ran my gradient with a restriction after the pump (nothing else in the flow path) and here is the result:
2 injections (black one i stopped manually)

uugh...

can you run that (with or without column) again but with some chromophor in the eluent B and record that trace? e.g. aceton or thiourea (if with column)
Because pressure may not be telling us enough details.
Or maybe run a gradient step test on this system, even if the GPV was tested externally.

It seems that the GPV indeed has some issue.
Even the good traces earlier (red and blue) seem to have some "corners" in there and are not smooth enough, (in my opinion).
But in that picture, the bad traces already start at higher pressure levels. If that is not due to some differences in the plumbing (guess they are with the replacement GPV?), then either eluent A is not 1% ACN, or channel B has some issue. Or
Maybe the GPV is at a point, where it doesnt "like" the ACN/buffer anymore, just like issues with checkvalves, which can be ok until a certain ACN concentration is pumped.
This would not be have been discovered with the water/water:aceton check that you mentioned.