Pressure spikes upon return to original conditions

[Arne]

Hello,

I see a pressure spike of about 40 psi (from 240 to 280 or so psi) upon returning to initial conditions in an HPLC method. At the end of the gradient the mobile phase composition is 95% MeOH and 5% H2O and I return to 95% aqueous in 6 seconds. Over 5 min the pressure will normalize.

Why is there such a pressure spike? Can it be avoided by returning to initial conditions more slowly, e.g. over 30 s?

We are using a Surveyor LC pump.

Thank you so much!

Arne

[bisnettrj2]

I think you're seeing the pressure effects of going through the 40%:60% water:methanol to 60%:40% water:methanol portion of the gradient. The rate of change of your mobile phase composition at the pump head or mixing tee is not the same as the equilibration of that mobile phase through the column, so you're seeing the column transition through the range of compositions, albeit in a rapid manner. At some point, the mobile phase composition is between 40 and 60% water in the column, and this is the most difficult composition of a water:methanol mixture to pump due to its viscosity. Therefore, you see a brief pressure spike.

[James_Ball]

I think you're seeing the pressure effects of going through the 40%:60% water:methanol to 60%:40% water:methanol portion of the gradient. The rate of change of your mobile phase composition at the pump head or mixing tee is not the same as the equilibration of that mobile phase through the column, so you're seeing the column transition through the range of compositions, albeit in a rapid manner. At some point, the mobile phase composition is between 40 and 60% water in the column, and this is the most difficult composition of a water:methanol mixture to pump due to its viscosity. Therefore, you see a brief pressure spike.

I also agree with this. With water methanol you will see that spike because the highest viscosity is in the middle of your gradient range near 50:50. With other solvents like Acetonitrile you wont see this as the more organic you add the lower the viscosity will be.

[Arne]

Thanks for the reply, guys!

I knew the methanol/water viscosity behaviour but I didn't think about that. Good call.

What still confuses me is that the pressure takes many minutes (>5 column volumes after quick change in conditions) to go back down to the level expected at 95:5 water:MeOH.

[James_Ball]

Thanks for the reply, guys!

I knew the methanol/water viscosity behaviour but I didn't think about that. Good call.

What still confuses me is that the pressure takes many minutes (>5 column volumes after quick change in conditions) to go back down to the level expected at 95:5 water:MeOH.

Methanol/Water always seems to me to take longer to equilibrate. What is your flow rate and how much dead volume is there before the column?

[Arne]

James,

thanks for pointing out that MeOH/H2O takes longer to equilibrate. I originally chose MeOH over ACN because ammonium formate is more soluble in MeOH.

Column + guard column volume is:~245uL
Total system volume is: ~270 uL (Solvent peak at 1.35 min at 200 uL/min)
Total dead volume: ~25uL (including 20 uL inejction loop)
Before column: ~ 23-24 uL is my best guess. I did not measure what part of the dead volume is before and after the column.

Would changing the mobile phase back to initial conditions more slowly attenuate the pressure spike, e.g., taking 30s instead of 6s to go from 95% B to 95% A?

[lmh]

There's no benefit to changing slowly(*). The fundamental problem is that 50% MeOH is more viscous than water or methanol so no matter how slowly you make the transition, when it's at 50%, the back-pressure will be high.

In fact, I haven't tried this, but I'd guess a rapid change is better than a slow one, because if you change rapidly, I can imagine a situation where most of the column is full of water or methanol, and only a short zone somewhere is currently sitting in the danger-region at 50%, whereas if you go slowly, at worst the entire column will be full of the most viscous mix. Another thing it'd be great to try out!

(* on modern silica-based columns; some columns may not like rapid changes of organic content for other reasons)

[Arne]

lmh,

Interesting points. I agreee with you. Thanks for chiming in.

ok, now it's time to analyze 80 samples of biomass hydrolysate....