Noisy baseline

[nan0guy]

Hello,

I work in the pharmaceutical industry, and have a few years experience running HPLC, but I still consider myself to be learning the field.

I am running an HPLC column, using ion-pair additives (1-octanesulfonic acid sodium salt) per a USP method to analyze an HCl salt of a drug compound on a C18 column. I am plagued by an irregular noisy baseline.

I have gone through the typical troubleshooting efforts re: a noisy baseline with which i am familiar - leaking (none inside or out); detector lamp (strong), bubbles(none) or electronics (isolated); mobile phase miscibility (water/ACN) and quality (clean glassware, HPLC grade solvents). The column is not losing silica, but it is older, and has been used for many injections.

Here's my actual dilemma - How do you decide to trash a column? Are there objective criteria. These columns are costly, but time and effort aren't cheap, and I hate to be wasting either.

Literature suggests that extended washing (liters) can remove ion-pairing reagents, but is the column recoverable, truly? Have too many of the silanes/retention phase been washed from the silica surface, exposing silanols? Could I expect TEA to make a difference in the noisy-ness?

Finally, can I explain my noisy baseline as material being released in a non-uniform way from the column, since silanes are lost and silanols are exposed, or ion-pairing reagents are clustered?

Thanks for the help in understanding this phenomenon!

Thanks,

Lloyd

[Blazer]

Here's my actual dilemma - How do you decide to trash a column? Are there objective criteria. These columns are costly, but time and effort aren't cheap, and I hate to be wasting either.

I'll tackle this part of your post. The columns aren't cheap for you or for me based on our salaries. However, for a company, columns are cheap and especially when they are viewed through the lens as one piece of the entire manufacturing/testing process. They are a consumable part and should be replaced as necessary. It's the cost of doing business.

John Dolan has a pretty solid rule of thumb for LC columns. Once you get to the point where you're at $1/injection, the column doesn't owe you anything. For the average column this means somewhere around 700 injections. As you pointed out above, troubleshooting the column can be a timely and in turn, expensive process. Your time, the cost of materials and solvents can all quickly surpass the cost of a new column.

[James_Ball]

Usually when I get rid of a column is when the back pressure becomes too much to deal with, which means it has become clogged or the silica has begun to break down into smaller particle sizes. Only a few times have I had to dispose of a column due to poor peak shapes or contamination. I had one column I used for over 10 years with no problems, but that was a rare one.

I find noise in the baseline is more often a flow problem than anything. Even the smallest leak in a check valve or piston seal can cause baseline problems. If you are running an Agilent 1100 or 1200 system you may want to replace the Active Inlet cartridge, when those get old you will get noise in the pressure line of 1-3 bar which will cause a noisy baseline, and that has been one part that fixed my problems in the past. Also you will get similar problems if you are using a vacuum degasser and it stops working correctly, causing tiny pressure fluctuations due to dissolved gasses in the mobile phase. There are many things that can cause noise at the detector, not just columns and lamps.

[Gerhard Kratz]

I would run the QC run stated on the COA of the column.
Ion pair always bonds to the surface, not possible to wash 100% from the packing material.
Check your degasser and pump heads. Good luck

[ATAlR]

I would run the QC run stated on the COA of the column.

Running CoA is always first think I do with new column, so after a problem arise I can always compare current performance with the first run.

[BostonFSE]

Noisy baseline is typically lamp or dirty flow cell.
Run constant flow of water for about 10-15 min and monitor baseline. Note the the max and min drift on the y axis (I like to see a maximum delta of only a few milli aus on the Y axis)
Now pull out the flow cell and do the same thing for 15 min

If the noise gets alot better when you pull the flow cell out, you have a dirty flow cell lens. if it stays the same you need a new lamp

this is assuming you have no column on (just a union) and you have good quality Mobile phases.

If column cost is an issue, I have reverse flushed columns with 100% organic (run column out to waste) for a few hours at low flow rate to get a longer life out of them.

[JI2002]

Before trashing the column, run your method without the column. If the noise level is the same with and without the column, the noise baseline has nothing to do the column. Also what detector are you using? The ion pairing reagent you use has no UV absorption if UV is the detector for the method.