GPC Baseline issues? (Binary Pump G1312A) - Faulty check valve? Malfunctioning proportioning valve? Priming issues?

I am new to GPC (and HPLC) and am assessing a GPC system that was sitting for some time.

I am having baseline issues and have tested various things to figure out how to resolve it:

I have a baseline issues that have a profile corresponding to deriva. It was more like ruido ciclico, but after hooking up the channel B which had not been hooked up before and replacing the following pump parts: Active inlet valve, Outlet valve, pump seals, Solvent filters on A and B I am now still seeing long-term noise that looks like deriva (Delta is approximately 500 nRIU). 

I can rule out detector issues (RI signal is stable if pump is switched off) - temperature variation within 4-5 hours i approximately 1.5-3.5 and within 1 h approximately 1.2Celsius.

I primed the in-line degasser and pump, but did not use IPA for it. Instead I used THF that I would like to use for the samples. I just ordered IPA to do it properly. Maybe there is still air in the system and I can remove it that way

I read that the check-valve could be faulty or a malfunctioning proportioning valve (I am pulling THF from the same bottle) could be causing issues. Where can I find these on the Binary pump G1312A)? Is the check valve one of the active inlet or outlet valves to the pump? Is the proportioning valve the solvent selection valve?

I installed new columns, so there should not be issues with column contamination. However, bypassing the column leads to a more stable and acceptable pressure (delta: 0.75 bar (constant) vs 2.5 bar (varying)) and a slightly more "stable" RI signal (deriva type variation, delta 500 nRIU vs. more ruido ciclico type variation. Amplitude approximately 250 nRIU but on top of that an increasing RI signat with a delta of 1250nRIU in h), but not anywhere near the +/- 2.5nRIU that it should be.

Signals for system with bypassed column and sampling port:

Pressure Scale:

RIU scale

  • Hello Bchristle, 

    The RID will be extremely sensitive to solvent composition and therefore when using a Quat pump the Gradient valve should be entirely bypassed. Route the solvent line from the degasser directly to the Inlet valve. If using a Binary pump, place both solvent inlet tubes into the same bottle. Program the pump to draw 50% from each Channel. These steps prevent 'crosstalk' where other solvents are inadvertently drawn into the stream. See if this resolves the instability. Also insure the RID and Column temperature is 5 degrees above ambient.  


  • Thanks. I I have a binary pump and you can see our set up in the pictures below (lines labeled solvent reservoir are coming from degasser). I already p put boyj solvent inlet tubes into the same bottle. Are you saying I should bypass the solvent selection valve in the middle? Would it not be confusing if I then program pump to draw 50% from each channel?

  • The baseline variation of the RI signal is +/-250-500 nRIU and not +/-2.5 nRIU as stated in the manual. The pressure variation is inconsistent +/- 0.75 to 1.5 psi. When I inject a standard, it looks like the chromatogram below. I was told that the chromatogram looks great and the baseline fine before and after the sample and that there is nothing to worry about using it the way it is. What is the typical pressure/RI signal variation on a HPLC/GPC?

    Pressure Scale

    RI Signal scale

  • Hello,   I agree with the statement the chromatogram looks great and the baseline is fine and there is nothing to worry about. The better question to ask is how much ripple is present and is the pump operating within its specifications? The answer here is Yes. 1.5 PSI is about 0.1 bar.  The specification is <3bar. The pump is passing with flying colors. If you want to calculate the %RSD for the RID, make repeat injections of a standard material and calculate the average areas and relative standard deviation. For GPC in general the tolerance is quite high, 10% typically. Hope this helps. 


Was this helpful?