Column pressure rising

[Albany-12303]

I recently developed a method for a product (solution) that contains hyroxypropyl methyl cellulose (HPMC). This method is for the active and degs. I dilute 5ml of solution to 500mL with my pH 7 buffer (containing 2% ACN). The method uses a gradient going from 2 to 20% ACN in the buffer.

My problem was that I was getting increased pressure and a quick deterioration of peak shape after about 10 sample injections. I assumed this was due to the HPMC

I changed the method to precipitate-out the HPMC (in Acetonitrile) prior to analysis - stock (and taking an aliquot of the clear upper soln and diluting it with Buffer). My final organic conc in the MP ended up as 5% (to get it lower, ie 2% I would have to add more sample to make the stock at which point the active starts to precipitate (not freely sol in ACN). With the higher org content, I get a dip that interferes with a peak of interest.

So it looks like I have to inject the sample diluted in buffer/2% ACN with HPMC present.

I added a guard column, Gemini C18: 4.0mm (L) x 3.0mm (ID), and this seems to work for the most part. The peak-shapes are good afer 20 or 30 samples, but the pressure begins to build (in the column, not the guard - I checked for that).

I have been covertly (Scientists in our company generally does not do this) taking the columns apart, replacing the top mm of packing with fresh packing, and sonicating the frits in nitric acid. This works fine.

OK, that was a bit long winded, but my question is how can I prevent this rise in pressure? I do not want to change the sample prep - Is there a better kind of guard column that I can use?

Thanks

[Mark Tracy]

Try filtering your sample. I have tested Whatman Anotop+ filters that go from 0.20 to 0.02µm; they work and can protect even these new sub-2µ columns.

[Rafael Chust]

Have you tried to sonicate the sample as well? Besides the impurities, you might have not too solved sample...

[Albany-12303]

Filtering does not help - I believe that the HPMC passes through the filter (even 0.2 micron!)

The components are quite solublized. I am taking 5mL of a clear solution and diluting it to 500mL - the resulting solution is as clear as water (although the HPMC is dispersed as a colloid)

If I use a silica guard column instead of a C18, then maybe the silicia will trap the HPMC better (lots of hydroxy groups) and stop it from getting into the column.

[Mark Tracy]

That your column treatment works is prima facie evidence of fouling by insoluble materials. Prevention is the cure. Does the HPMC precipitate when mixed with mobile phase?

[Rafael Chust]

Why do I still think sonication could be a good tool? [/img]

[Albany-12303]

There is no precipitation in the mobile phase. HPMC is completely soluble - but not in the normal sense. Rather it is essentially a well dispersed polymer (geling or thickening agent). When you look at a bowl of 'Jello' it looks like the gelatin is dissolved, but it really is not (rather it has swelled throughout the entire volume)

I don't think that sonication would provide enough energy to break-apart the polymer.

[Mark Tracy]

I don't like to recommend ultrafiltration because it is expensive and slow, but that will remove the HPMC. Use a centrifugal 10000 MW cutoff filter.

[JZeszotarski]

Just a wacky out of left-field kind of idea, and I don't have any idea if it would really work, but could you crash out the HPMC by crosslinking it with borate? It would be an aqueous treatment, so hopefully chromatographic issues would be minimized, though I don't know where borate would elute on your method.

Just an idea... it works for making Gak with polyvinylalcohol, but not sure if it would crosslink HPMC.

Good Luck,
Jon

[Albany-12303]

A few year ago, I was working with a product Carbopol (a carboxylic acid polymer) that caused a similar problem. I was able to crash it out by adding a divalant cation (CuSO4). Since HPMC is non-ionic, I didnt bother to try this.

But using Borate? Sounds like a novel suggestion. What is the mechanism by which borate can cause cross-linking?

Thanks

[JZeszotarski]

I'm no organic chemist, but my understanding is that borate reacts with hydroxyl functionality. It can react with 4 hydroxyl groups and crosslinks PVOH by reacting with two hydroxyls that are in a 1,2 arrangement (it apparently doesn't react as well with 1,3 arranged OH groups) on each of 2 separate polymer chains.

As long as the HPMC has hydroxyls that are properly spatialy arranged I'm guessing it would work.

[TosohTechman]

Here's a question for you: how frequently is the column being cleaned? Over time, you'll start to get residual build up, even if you recovery all your A280. Make sure you follow the recommended cleaning procedure in reverse flow to eliminate that as a variable to why your column is experiencing back pressure.

[Albany-12303]

Washing does not help (but I wash only with water then ACN -I have not tried more exotic solvents such as THF)

The solution I inject is crystal-clear and even if I filter it the column does not last more than 10 injections (without the guard column).

[Alfred88]

Dear all:
I don’t want to address the “rising pressureâ€

[TosohTechman]

What is the matrix of the resin (silica, polymeric) and nature of the ligand(what conditions can it withstand)? What concentration of ACN are you using with the column. Depending on the matrix of the column (if this is RPC or analytical) you could hit it with a stronger conc. of ACN. Hope this helps. I only other thing I can think of if you're getting material crashing out on your column. You've completely ruled out any system pressure contributing to the problem?