Life time of a SE HPLC column

Discussions about gel permeation chromatography / gel filtration chromatography / size exclusion chromatography

9 posts Page 1 of 1
Well, it looks like I cannot get any more mileage out of another size exclusion column. I was wondering if anyone would share insights about which columns are more durable and how to make them more durable. I understand that it may be a sensitive topic for people who sell columns, so I would ask advanced users for their opinions. Which brands did you have better luck with? What shall you expect from a SE column? How does the flow rate (and back pressure) affect the durability of a column? I understand that the column shall be operated within the limits set by the manufacturer, but intuitively it seems that a fragile architecture of highly porous stationary phase may benefit from lowering the flow. In my case, the instrument time is not a problem, so I would lower the flow rate if I can get a significant increase in the # of samples before the column dies. Also, filtering of the mobile phase (we use 0.1 micron membranes) and spinning of samples goes without saying. I am separating antibodies and their conjugates and use silica-based stationary phases with pores around 300 A.
Use of guard columns is a separate issue. I've heard many arguments for and against using them, but, unfortunately, most of the "for" arguments came from sales people. What do users think about guard columns?
Yet another aspect is washing of SE columns. My experience is that if it seems like it is the time to wash the column, it is actually the time to order a new one. Does anyone had better experience?
Please at first follow the instruction manual of the manufacturer. Handle a SEC column with care like any other column.
It makes sence to use a guard column. Make your samples and mobile phase as pure as possible. Matrix on a column is always difficult to wash out. Avoid pressure peaks. If you get an average of 1000 injections on a column or only 500 depends on your samples and the matrix and the mobile phase you run over the column. If suddenly you get only 50% of the number of injections than something happened with the column. Good for sales people than now it is time to purchase a new column. Don't waste time and money with column washing and repair efforts. Same with increase of back pressure. Here maybe the change of the frit will help but there is no guarantie that this will help a long time. Good luck
Gerhard Kratz, Kratz_Gerhard@web.de
I am with Gerhard that buying a new column is the wise choice if your column has lost most of its performance. But if funds are slim I know that sometimes you have to try and keep the columns going for as long as possible.

I previously worked in a lab where we used GFC columns from GE which were see-through. After some usage we could see the packing material become compressed, giving quite a void in the inlet. We could improve the column life-span somewhat by gently back-flushing the column. However, it might not be recommended to do so so check the instructions for the column first before trying it.
Actually, the manual for the column provides instructions for back flushing it with aqueous-organic mixture (say, 20-30% acetonitrile in water), or high salt buffer with low pH, or 6 M guanidine to wash away proteins (did not try this one yet). It did not work for me. My guess is that by the time I am washing the column, it got voids and reverse flow only disturbs the stationary phase even further. A couple of time that I did it, the pressure went way down at the beginning of the back flow wash, but it quickly climbed up such that I had to reduce the flow. It kind of recovered overnight as the mobile phase used as a wash was substituted with water, but went through a similar cycle when the flow was switched back to normal and the mobile phase was returned to a buffer.
I've been battling nearly the same issue for years. I work with antibodies and antibody conjugates on silica based HPLC columns was well, and have always been disappointed with the number of injections I get before the peaks become too broad to work with. On average I may get 250-300 injections out of a column, and I think ONCE came close to 500. What's strange is, if I run the manufacturer's QC test (usually something like PABA, or some other small molecule to determine the plate count), on a brand new column and a column that is "dead", I get the same result. The peak broadening only seems to apply to the antibodies themselves. I have found a few things that seem to help a little:

1. ALWAYS filter the mobile phase. We use a totally aqueous mobile phase, with sodium azide added to retard microbial growth. One thing to be aware of with sodium azide however, is that it is not stable at acidic pH. Silica however, doesn't like high pH, so you're stuck between a rock and a hard place. Also, if you adjust the pH of an azide containing mobile phase using an acid, add the azide last, as when adding the acid you create local "pockets" of extremely low pH, and some of your azide decomposes.

2. If you're working with antibodies/conjugates from multiple hosts, watch out for situations where, for example you inject an antibody from a mouse host, then later inject an antibody directed against mouse immunoglobulins. I don't care how inert the manufacturer claims the stationary phase is. Some of your sample is always going to remain on the column, and your next sample may then interact with the previous one, and so on.

3. Some manufacturers recommend that before using a column, you inject a few large slugs of some "inert" protein such as BSA to "block" any potential sites on the column that might interact through some mechanism other than size exclusion. I don't know how much good it does, but if I get a new column and do several injections of BSA, I do see the peak area increase slightly with each injection until it finally plateaus. There are some mobile phase additives which have been shown to reduce secondary interactions.

As far as cleaning goes, I too am guilty of waiting until I notice degradation in performance to clean (low pH/high salt, organics, guanidine, urea, etc). It hasn't worked once. Perhaps if I cleaned more often it might show some benefit...

And finally, I do use guard columns (cartridges) in addition to a pre-column filter, even though I filter all samples. I can see on the cartridge when I change it that it usually catches something. Either particles shed from the rotor seal, or in the case of conjugates, the color of whatever conjugate was injected most often.

I agree that it's usually best to just bite the bullet and buy a new column, but I wonder why SEC HPLC columns seem to be more expensive than, say, C-18. If someone could shed some light on that, I'd appreciate it.
I'm not going to weigh in regarding the stability of some other company's SEC columns, but I can address the price issue. SEC columns tend to cost more than RPC columns because they're larger; 300x7.5 mm or 200x9.4 mm vs. 250x4.6-mm for a typical SEC column. That means that their volume is 4x greater. There's a good reason for that. In a gradient mode like RPC, analytes typically diffuse about 5-10% of the way into the pores before they diffuse out again (assuming the material is totally porous). In order to get maximum resolution in SEC, you need to give time for small solutes (those that elute in the Vt peak) to diffuse into the full pore volume in order to accentuate the difference in column volume available to them vs. the volume available to larger solutes. That requires a flow rate about 8x slower than with a gradient mode. With an SEC column of the dimensions listed above, that would be at a flow rate around 0.5-0.6 ml/min, a rate well within the range of most HPLC systems. With an SEC column with an i.d. of 4.6 mm, the optimal flow rate would be about 140 µl/min, a rate that some HPLC systems would have trouble delivering with precision.
PolyLC Inc.
(410) 992-5400
aalpert@polylc.com
I agree with Andy.
It is not only the larger amount of packing material in the SEC column that justify the higher price. It is the application work done in the application lab. And therefore I would say the price also for a standard C18 column is too low. When ever it is asked how to separate A from B column manufacturers need to present applications and need to have enough experiance to answer such questions. Every column should be handeled with care.
Gerhard Kratz, Kratz_Gerhard@web.de
Thanks everybody for feedback. I'll answer Scott's post first and then share what I learned from talking to a sales rep and to members of my group.

1) Filtering mobile phase is a VERY useful thing. We do it twice: filtering it into a bottle and with an on-line filter between the pump and the autosampler. Both filters are 0.1 micron, We prepare 4-5 liters of mobile phase at a time, so the cost of bottle-top filters is not a big problem. I was originally using a 0.5 micron on-line filter immediately in front of the column, but decided to switch to spinning of all samples before injection to avoid peak broadening and sticking of hydrophobic material to the filter.
I decided to avoid sodium azide. This compound is only 4 time less toxic than cyanide which is still extremely toxic. I used to work at a plant which uses azides on a multi-kilo scale, and I am running away from it ever since whenever possible. My line of thinking is that if something starts growing in the mobile phase, the on-line filter will catch it. To reduce the chance of doing "inadvertent" microbiology, I use sterile bottles to filter mobile phase in. I never combine leftovers of mobile phase with a new bottle, and discard mobile phase after ~two weeks after I start using it. So far, I did not notice any problems.

2 and 3) Ab-Ab sticking is always a possibility and, actually, even BSA can stick to or react with an activated antibody (see #6 below). Pre-coating of the stationary with BSA is a scary option because I have a MALS detector down the line and Mw readings will be all screwed up if there is any BSA leaking into other peaks. I usually run BSA as a validation standard after a blank in most of my sequences and keep fingers crossed that it does not stick to the stationary phase. Please tell me if it is not the case and that BSA does indeed stick to silica-based stationary phases.

4) News from people who sell columns. I got a piece of advise to include "prophylactic" shots of 100 ul of 6M guanidine HCl every 5 - 10 samples. It is kind of scary because 6M aqueous guanidine salt is >50% w/w solution, but I hope it will help. Also, I will probably set a rule to do a high salt pH 3 and/or 30% acetonitrile/water washes every 100 samples.

5) I am still hesitant to use guard columns. I still think of them as extremely expensive on-line filters. They can probably catch something outrageously reactive if somebody decides to inject it for whatever reason, but since SE phases are intentionally made rather inert, they will probably release most of the stuff anyway to the column behind.

6) Speaking of "outrageously reactive" stuff, I found out that some people in my group were submitting chemically activated antibodies for analysis with live maleimide, pyridine disulfide, or thiol groups. I think that this is a very effective way to destroy a column. If a protein sticks non-covalently to the stationary phase and is treated with one of these activated proteins, the blob of protein will be fast growing. One way to address this issue is to ban all "activated" samples and inject samples only after the final conjugation step. However, the conjugated samples may contain excess of activated reagents, and we are back to square 1. I am curious if anybody has a method for deactivation of activated groups on antibodies before analysis by SE HPLC. Any ideas are welcomed.
Andy Alpert wrote:
I'm not going to weigh in regarding the stability of some other company's SEC columns, but I can address the price issue. SEC columns tend to cost more than RPC columns because they're larger; 300x7.5 mm or 200x9.4 mm vs. 250x4.6-mm for a typical SEC column. That means that their volume is 4x greater. There's a good reason for that. In a gradient mode like RPC, analytes typically diffuse about 5-10% of the way into the pores before they diffuse out again (assuming the material is totally porous). In order to get maximum resolution in SEC, you need to give time for small solutes (those that elute in the Vt peak) to diffuse into the full pore volume in order to accentuate the difference in column volume available to them vs. the volume available to larger solutes. That requires a flow rate about 8x slower than with a gradient mode. With an SEC column of the dimensions listed above, that would be at a flow rate around 0.5-0.6 ml/min, a rate well within the range of most HPLC systems. With an SEC column with an i.d. of 4.6 mm, the optimal flow rate would be about 140 µl/min, a rate that some HPLC systems would have trouble delivering with precision.

I learn something new with every Andy's post. My 300x7.8 column came with a test chromatogram obtained at 1 ml/min, so I assumed that this was the optimal flow rate. The back pressure stays far away from the limit and the column seems to separate monomers from dimers of BSA or IgGs, respectively, almost to the baseline. Am I cheating myself out of something by not going down to 0.5 ml/min? I thought that slower flow rates will lead to broader peaks and larger errors in the concentration readings (UV or RI) important for Mw determination with the MALS detector.
9 posts Page 1 of 1

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 599 on Tue Sep 18, 2018 9:27 am

Users browsing this forum: No registered users and 1 guest

Latest Blog Posts from Separation Science

Separation Science offers free learning from the experts covering methods, applications, webinars, eSeminars, videos, tutorials for users of liquid chromatography, gas chromatography, mass spectrometry, sample preparation and related analytical techniques.

Subscribe to our eNewsletter with daily, weekly or monthly updates: Food, Environmental, (Bio)Pharmaceutical, Bioclinical, Liquid Chromatography, Gas Chromatography and Mass Spectrometry.

Liquid Chromatography

Gas Chromatography

Mass Spectrometry