Shodex KW column

Record: 527

Category: Specifications
Product: KW
Sample:

Question
I’ll go over what we are doing and observing to help answer your questions. First I’ll talk about the conditions followed by the observation.

Conditions: Waters Aquity H-Class UPLC in HPLC mode with e-Lambda PDA detector. The column is conditioned for 60 minutes using 1x PBS buffer pH 7.4 at 0.35mL/min. Sample run is isocratic 1x PBS buffer, pH 7.4 for 30 minutes at 0.35mL/min with subsequent injections (no next inject delay or equilibration between samples). Sample injection is 20uL of 0.2uM sample monitored between 210-800nm. Cleaning/flushing method is performed at the end of each run consisting of 100% water for 20 minutes, then 90/10 MeOH/Water for 2 mins, then 100% MeOH 2 mins followed by 100% water for 20 mins at 0.35 mL/min and stored over night. All changes are abrupt with no gradient between solvent conditions.

Observations: Each injection results in a sample with shorter and shorter retention times for the primary peak (graph below, red trace was first injection, green second and blue third). This observation is repeatable from day to day runs regardless of the initial equilibration time (tested at 10, 20 and 60 minutes) before injecting samples. The drift in retention times appears to be somewhat reproducible with each injection (first sample has the largest shift followed by less extreme shifts with consecutive injections). Peak shape and quality appears to be maintained despite the drift and this is only observed with this particular column which had approximately 30 injections before this problem was observed. This is observed with or without the guard column in place.

Please let me know if you need more information. The trace below was something I had available and is not the optimized PDA channel but the peak shape and drift is consistent. I picked this channel because it also shows baseline drift.

Answer
As I mentioned in my first email, it is an often reported phenomena that the KW columns needs some time for equilibration for a certain type of samples.
Since you are cleaning the column at the end of analysis, the adsorbed sample is washed off and at the next analysis, the column again needs to be equilibrated until it reaches the point it will not adsorb the sample anymore. It should not influence the following analysis. The reason you are seeing shorter retention time is probably because the sample does not “stick” to the column at later injections.
However, my concern is that I believe you have been using the same method with many of KW403-4F columns in the past, but not observed this before. Is this correct? Are you analyzing different sample? Was there any other change in method with this column?
If so, would you please tell me the serial numbers of your current working column and one of the old one? I will check if there was any difference in them at shipping time.

It is a good practice, but if you are using the column everyday basis, the cleaning step may not be necessary. You may simply flush the column with 100% water to wash off the salt when you shut down the HPLC system. Without MeOH cleaning, I think you should see stable retention time at start of the next analysis.
In addition, below is our technical notebook for KW400 series columns. In the booklet, it explains about the flow rate and injection volume suggestions.
I believe you will benefit the resolution advantages of KW400 series column if you lower the flow rate and injection volume…
http://shodex.net/data/downloads/Shodex_Technical_Notebook_5_KW400.pdf

 
 
Additional Comments
Since the negative peak (probably the solvent peak) observed is also shifting (23->21->19.5 min), shorter elution time may not be only due to the column.
What we suggest you to try are:
1. Check the column-back pressure at the beginning of the analysis (when the elution time is longer) and then check the back pressure again at some time after.
2. Also collect the effluent for a fixed time in the beginning and after several analyses to check if the pump is working properly.

Download or Print PDF