We have been experiencing intermittent stalling with our robo-columns, sometimes unrecoverable.
I was wondering if we are using to slow/fast flow rates? With very minimal information provided by Tecan we are unsure of what would be considered reasonable values. I am not sure whether it’s coincidence or not, it seems to always happen when we switch to a faster flowrate after using a slow flow rate. Note we usually only see a stall in one plunger at a time. Sometimes hitting retry the Tecan succeeds, hopefully this is not damaging the instrument?
For example we just had a plunger stall on dilutor #5 from this scenario:
your situation is not foreign - kinda the fall out of using an automated liquid handler to perform column based chromatography; which is part of the reason we developed inline pressure and flow sensors to better monitor flow characteristics through the system & robocolumns
assuming you are using 600 uL robocolumn, 30 uL/s equates to a residence time of <30 seconds - which is way too fast,
typically, for buffer & load phases, i would be running something at 4 minutes or more - which equates to 2.5 uL/s
Agreed that this is not uncommon. Something that Repligen and Tecan have both communicated in the last few years is that you may have to reduce your cycles/re-uses of the columns. There can be decomposition of the resin that causes pressure spikes. Even if your product quality data looks like you can achieve X reuses of the column, you may need to reduce that further to prevent these frustrating failures.
I’ve also found this to be very molecule/project dependent. Some combos of molecule and resin never have a pressure failure, and others are very sensitive.
I’ve switched jobs but my previous company used Robocolumns quite a lot. 30uL/s is way too fast and is definitely the source of your plunger stalling. IIRC, the highest we ever went was 5uL/s on 600uL columns. Typically that would be for CIP steps where we just wanted to fill the column and then let it sit. For sending actual material through the columns or eluting we were in the 1uL/s to 2uL/s range.
The Fluents had much better control for those slow speeds than the Evos did. We used Robocolumns on both platforms, but the Fluents were definitely better.
Thank you so much for your knowledge, this is the first time I have implemented chromatography so lots to learn.
To answer your question:
Resin: Sephadex G-25 Superfine
Column Size: 600 uL, 30 mm bed height
Concentration: N/A, error occurs between dispensing different buffers (Water + PBS)
Pre-filtering: N/A
Desired flow rate/residence time: Residence time 1 min
Syringe Size: 5 mL
We are right now only using the columns a handful of times less than 20, we thought they could handle a lot more. This occurs at the fast flow rates during our cleaning steps.
It is interesting I think we based our flow rates on the Akta (will double check with our scientific team). Tecan team programmed the method for us with these flow rates and did not raise a flag that they were too high! We have been operating under the assumption that they would have been ok since they were signed off by them. Thank you all for the info!
the typical usage is based on insertion/exit events damaging the O-rings & causing leakages (opposite of what you’ve been seeing),
using the standard Tecan model of 1 x insert/exit per 4-5 mL of liquid addition, and a typical experiment of 10 x different liquids across the column - you would only be able to use the columns about 5-10 times before the performance deteriorates,
the robocolumns are designed to be consumables and have been targetted for single use - but obvously, to be cost considerate, everyone wants to get more runs per column; as the o ring deterioration is the biggest problem, reducing the number of insert/exit events has been shown to improve the column lifetime to 20, 30 runs+ with minimal impact on performance
sadly, there is no pressure feedback through std tecan hardware - except the dreaded plunger stall, or a puddle of liquid under the columns after tip’s exit
the improvement to reduce tip insertion goes against the Tecan standard modus operandi for scripting - here we integrated larger capacity tubing & significany method development/flow control adjustments for better flow control & reduction in negative flow as the tip needed to exit the column for a re-aspiration of the same buffer (published work info available)
Our typical usage was setting a max of 10 and doing reuse studies with product quality.
Once we started seeing the pressure failures, we would try to bring it down to 5 if possible to derisk. I don’t know a magic number, that was Repligen’s advice to just assume worse degradation than we were seeing with PQ results.