Chromatography Forum

I'm currently developing, and almost at the stage to optimizing method using UPLC. I'm running a gradient programme (ACN:water) with post-time.

Problem faced :
1. Initial pressure at the start of the day is not always the same. (ie last week can be 175Bar, today can be 130Bar and next can be 200Bar). All mobile phase are being filter through glass filter. All lines are purged and equilibrium done before the analysis. But still I face this. How to go about this? This making the retention time shift. Or is this not a big issue as long as I inject my Standard solution every time with sample?

2. Few problems at once : Peak not symmetry, broadening of peak.
But this occurs only after months I run with the same method.
Observed non-symmetry peak (like tailing), tried to prepare sample solution in ACN:water instead of purely ACN. After many times of injections, the non-symmetry seems ok. But still see broadening of peak, double the size before the issue occurred. My point of view, the broadening of peak will influence the resolution of peaks, which I have 2 peaks elute at very close Retention Time. Please advice.
I would like to know also, why I observed sample in ACN gives better signal compared to sample in ACN:water?

3. Volume of injection.
I am expecting double signal height if double volume of injection is being made. This happens to sample in ACN but not to sample in ACN:water. Why is this so?

Owning only 1 column, thus cannot do anything with this parameter. But am getting a new column though.

Please help. Thank you.

Hello tzewoei,

Is the pressure stable or do you see it fluctuating?

Your first problem sounds to me like a failing inlet valve or a failing outlet valve. If possible take a look at your pressure curve over time when measuring.

At an Agilent HPLC each Active Inlet Valve has a replaceable cartridge in it. When this fails you're pumping less solvent in relation to the other channel so you're facing an instable pressure and shifts in retention time.

Same problem with the outlet valve.

Hello

Just run leak test and you will be able to check if your problem is AIV or OBV. If you have binary pump you have 2x AIV/OBV so leak test is very handy to diagnose that problem.

Regards

Tomasz Kubowicz

Hello asm,

Appreciate your reply. Still a newbies am I, please excuse me for not knowing how to describe some times.

Well, running in gradient, we are to expect a gradient in pressure curve, right? So, ya...a curve with gradient observed but I would say is not fluctuating?
If to compare the pressure curves of injections within the day, may get
I) as perfect as you can a RSD of Rt of 5 injections = 0.02%. Perfectly overlapped Pressure curve.
II) But another day, getting a difference in initial pressure (ie noticed an step-ly increase in initial pressure for 5 consecutive injections. From ~175Bar for first injection to ~190Bar for the fifth injection.) RSD of Rt of 5 injections = 1.0%

Tried cleaning the inlet valve and outlet valve...sonicate the whole valve in IPA. Sometimes, this helps but problem may reoccur. But never change a cartridge before. This is supposed to happen to a new machine? Just wondering..

How about the broad peak issue? What could be the cause?


Hi Tomasz

Thanks for the advise. That never come to my mind. May I know, is running a leak test the first and most important to do? Hopefully I can remember how to run a leak test...thanks again.

Hello

If you have binary pump you have 2x AIV and 2x OBV. So how can you know which one is defective? Leak test can help you diagnose problem. What LC/software have you got?
You can try to sonicate valves and cartridge but sometimes it is not worth to do it (try with hot water and methanol).

Regards

Tomasz Kubowicz

Before worrying about chromatogram problems such as peak broadening, I'd only worry about getting the HPLC into shape, i.e. getting rid of the pressure fluctuations and retention time changes.

Do you equillibrate the HPLC well before starting runs? Purge all the solvent lines well (ALL lines, even if you use only 2 of 4), run at 100%B until pressure is steady, then run at 100%A until pressure is steady, THEN start the gradient. If this doesn't give you the same pressure profile every time, you'll have to look into the hardware such as the check valves, as already suggested...

Hello

If you have binary pump you have 2x AIV and 2x OBV. So how can you know which one is defective? Leak test can help you diagnose problem. What LC/software have you got?
You can try to sonicate valves and cartridge but sometimes it is not worth to do it (try with hot water and methanol).

Regards

Tomasz Kubowicz

The unit I have is a quaternary pump, running with OpenLAB. So, to run a leak test, I will have to run with Isopropanol?
May I know why did you say so that not worth to sonicate the valves and cartridge?
I read something about dwell volume and dead volume. Are both the same? What really are they?

Before worrying about chromatogram problems such as peak broadening, I'd only worry about getting the HPLC into shape, i.e. getting rid of the pressure fluctuations and retention time changes.

Do you equillibrate the HPLC well before starting runs? Purge all the solvent lines well (ALL lines, even if you use only 2 of 4), run at 100%B until pressure is steady, then run at 100%A until pressure is steady, THEN start the gradient. If this doesn't give you the same pressure profile every time, you'll have to look into the hardware such as the check valves, as already suggested...

Hello,

I think I equilibrate the system well enough? Normal practise : Filter ACN, water differently and purge solvent line with ACN and water respectively at 5ml/min for 5minutes. Then run the initial ratio of the ACN:Water in my preset gradient programme for at least 15mins, or till no fluctuation of pressure observed. Then start the injection.
If something looking abnormal, i.e. pressure increasing and really a time shift or doubting any clog of the filter, extra step added before the normal practise - Purge all lines A,B,C and D with filtered IPA (5mins each line), then perform the step as per normal practice.

HPLCaddict, may I know why are you suggesting to run 100% B and 100% A at two different time? Also another question, is this issue something might happen to a new machine? Thank you.


Regards,
TzeWoei

Hello

Generally ball valves have seat and ruby ball. If you have any particles on ball surface you will have micro-leak inside valve (unstable pressure). Sometimes you can "clean" valve using sonicator but it doesn't work in all cases.

The best solvent to run leak test is 2-propanol. If you're using OpenLab Chemstation channel D is used for leak test (Qpump). You need to plugged purge valve outlet with blank nut.

Regards

Tomasz Kubowicz