Protocol and solvent help needed for gel-permeation chromatography

  1. So I have been asked to separate a polymer (polyurethane) into high molecular weight and low molecular weight components. I have access to Sephadex LH-20, a simple glass column, acetone (a List II chemical), ethyl acetate, and hexane. I may be able to get access to other solvents, as long as it is not anything exotic. I also have a couple TLC plates. I have experience with simple column chromatography (with silica gel) and GC-MS, although the labs I'm curerntly using are probably at high-school to simple undergraduate level with no fancy equipment or anything.

    In theory all I need to do is to pack the column with sephadex, run the appropriate solvent through it and add my polymer sample (dissolved in the same solvent) on top. Simple enough, except for two problems:

    One, I cannot find any appropriate protocols for what I am supposed to be doing. Most of the protocols I found are for separating proteins, and they all require a buffer of some sort. Do I need a buffer, even though I'm separating polyurethane? Or do I just soak the sephadex in my solvent until it swells, pack the slurry in the column to make a gel bed and just run my sample through the whole thing (while adding solvent continually until my arm cramps)?

    Two, as sephadex is expensive, I need help in determining an appropriate solvent to use. TLC doesn't seem to help much here as it uses a different stationary phase (silica). Anyone have any ideas on what solvent or mixture of solvents I should use as the mobile phase?

    Thanks in advance.
  2. jcsd
  3. chemisttree

    chemisttree 3,717
    Science Advisor
    Homework Helper
    Gold Member

    Try this for preparing your column (ignore the buffer references) and this chart indicates that the solvent you should be using is DMF or dimethylformamide. I hate working with dimethylformamide! How will you detect the passage of your polyurethane fractions?
  4. epenguin

    epenguin 2,496
    Homework Helper
    Gold Member

    You don't need to stand there adding it continually. You can have a standing column of solvent above the bed, maybe 30, 40 or more cm, and just top it up from time to time, it doesn't have to be absolutely constant. The solvent depends on the solubility of your polyurethane. Make it as aqueous or otherwise convenient as you can.

    Note as explained in chemsitree's instructions, the sample volume can be only a small fraction of total, by the volume exclusion principle. A delicate moment is when you have absorbed the sample and you put the solvent on top, you tend to get remixing you have to minimise by care and tricks.
  5. Thanks for the protocol. So it's basically just the usual steps for packing a column except that I'll need to presoak the stationary phase in solvent until they swell completely. Got it.

    Unfortunately I don't have access to DMF (or fortunately, as I don't want to work with that stuff either.) Neither do I have DMSO or acetonitrile, the usual go-to substitutes for DMF. Basically, if the solvent requires protective gear beyond a fume mask and gloves -- well, I don't have it, as the labs are meant for teaching low-level chemistry anyway (up to an introductory organic chemistry course). They are not research labs, unfortunately.

    (We do have a centrifuge though, but I'm not sure if it's worth the trouble to purify the sample beforehand to remove dust.)

    I suppose pure acetone is the closest substitute I can get --it being a polar aprotic solvent and all. Ethyl acetate may also work, but its dipole moment and dielectric constant differs a lot from DMF. I'm not sure if that matters as I am not familiar with the theory behind gel-permeation chromatography (other than the whole "pass a bunch of molecules through a network of porous beads, smaller molecules enter pores and elute slower" thing). Also, the acidic hydrogen on acetone won't matter right?

    As for detecting the polyurethane fractions, I'm supposed to just let the solvent evaporate from the eluents and check for the presence of crystals (yeah, very low tech). Then I'll redissolve them and send the samples to another person with access to a mass spectrometer.

    Well, the column I'm using is quite short (approx. 35 cm). Still, thanks for the tip, I'll just position a burette over the column and use it as a reservoir. Alternatively, I can plug a burette and use it as a makeshift column, but that does make packing harder.

    So, in theory any solvent should work as long as it dissolves the polyurethane well, right?

    Chemistree's instructions state that the sample volume should only be around 1~2% of the total bed volume. In theory I could just increase the concentration of the sample, but it tends to become viscous. Does viscosity matter when the sample is initially added?
    Last edited: Sep 13, 2013
  6. epenguin

    epenguin 2,496
    Homework Helper
    Gold Member

    That is most of it.
    Therefore in theory, unlike in adsorption chromatography, everything should elute within the column volume. In practice there will be some adsorption more often than not, therefore you will only know this by experiment. Use twice the column volume unless anyone knows better. And do not dismount the column till after you have done the analyses, because if anything is still left in the column you can then still elute it.

    Rather it is a combination of very lo-tech with relatively hi-tech. Useful would be some fairlyornery-tech. What polyurethane is it? - doesn't it have a UV absorbance?

    Or for elution you might set up a siphon arrangement from a large beaker.

    It could even be to your advantage helping you layer eluant on top of sample minimizing mixing.

    I don't know how this Sephadex is these days, but I remember that after swelling one used to stir and let settle a bit, throwing away the fine particles that don't settle so fast. Several times. They are a nuisance and enemy of resolution when you apply sample and eluant.
    Last edited: Sep 14, 2013
  7. chemisttree

    chemisttree 3,717
    Science Advisor
    Homework Helper
    Gold Member

    A small bed of washed sand at the top of the column will minimize mixing. Elute solvent through the bed until you see air just touching the top of the resin bed. Add the sample to the sand bed slowly while slowly draining solvent from the resin bed. Be careful not to allow air to contact the resin bed. When the sample has been added, remove mobile phase (through the bottom of the column) to draw the sample into the resin and then continue adding mobile phase in small portions, same volume as you were using for your sample, while slowly eluting solvent from the resin bed. This washes the interstitial volume of the sand plug free of analyte. Repeat three times and then fill the column and reservoir to the level you will be using during the chromatography. Be sure to maintain a constant head during elution so the pressure doesn't vary.

    As I see it, detection will be your most challenging hurdle and it's one that you should think carefully about. Typically the eluant is collected in fractions and they are each analyzed for the presence of analyte. The usual way to indicate the passage of analyte is with refractive index (during elution) but you will have a bunch of fractions so that isn't possible. It would be nice to have a chromophore that would show on a TLC plate with fluorescent binder but that will be very difficult with DMF (a spot won't dry!) but not so much with acetone. Just be certain an acetone blank doesn't contain impurities that will give you false positives (ie. use HPLC quality acetone!) If you don't have TLC plates with fluorescent binder or a sample that will produce a dark spot when illuminated with the UV lamp, you can visualize the plate in the usual ways.
    Last edited: Sep 14, 2013
  8. I don't know. It's in a bottle labelled "polyurethane resin" and is a thick, viscous liquid with roughly the same consistency as treacle. There's also some sort of dye in it so it appears red. In theory the dye should be of low molecular weight and therefore be the last to elute out of the column, but I'm not quite sure about that.

    As for UV absorbance, the sample does seem to quench light (very weakly) on my TLC plates, but I neglected to check with an acetone blank to see if my solvent has any impurities.

    I don't have a UV lamp though, so I can't shine one directly on my column. I'm currently using a transluminator I 'borrowed' from the bio labs.

    I'll double-check with an acetone blank to see if it has any impurities. I won't be able to use the labs until Tuesday, UTC+8. I'll keep you guys updated then.

    (Oh and as a side note, I don't have microcaps so I'm currently making substitutes with the ol' "heat a pasteur pipette with a bunsen burner flame" trick. And disposing of the used ones discreetly so the lab staff won't complain too much about me wasting equipment. The teaching staff are totally fine with me using the pasteur pipettes this way but the maintenance guys will raise a fit because they apparently have to fill out some paperwork. For a couple cheap glass tubes. Sigh.)
    Last edited: Sep 14, 2013
  9. Alright, I apologize beforehand if this counts as double-posting, but technically this is a new update.

    So I ran some tests with the TLC plates. The acetone has trace amounts of impurities in it -- no spots, bands, or splotches can be seen on the acetone blank (acetone is spotted onto the TLC plate with a microcap, then the plate is developed in more acetone), but it does look slightly different from an unused TLC plate when both are viewed under UV light (the luminance looks less... smooth?). I also exposed both the acetone blank and and unused TLC plate to vapors of 'a certain solid halogen that I cannot name due to forum rules'. The former is slightly darker than the latter, 'though again, no spots, bands, or splotches can be seen.

    My polyurethane samples (dissolved in acetone) consistently result in a spot on the origin and a broad splotch about halfway from the origin to the solvent front, even when I vary the concentration of my sample ('though I haven't tried a more concentrated sample yet, as those always clog my microcaps.)

    The spots and splotches can also be seen when the TLC plates are exposed to the previously mentioned halogen vapors. No additional spots were seen.

    I also accidentally left a TLC plate in the developing chamber (beaker filled with solvent with solvent-impregnated filter paper on top with a watchglass) for too long (approx. 1 hour). When viewed under UV light, all the analytes appeared to be pushed to the top edge of the TLC plate, as expected. Interestingly, after I exposed the TLC plate to the halogen vapors, a clear spot can be seen on the origin, in addition to the expected band at the top edge of the plate.

    Repetition of the above yielded a similar result. Repetition with an acetone blank loaded with the same microcap (cleaned with acetone multiple times before) yielded a negative result on both tests (UV and halogen vapor) Does that mean that my polyurethane has a fraction that doesn't absorb UV light?

    I also did some tests with ethyl acetate as the solvent instead of acetone, but they all yield a broad, balloon-shaped blotch starting from the origin, so I didn't pursue that further.


    So basically, what I should do is --

    1) Soak sephadex in acetone for three hours.
    2) Stir sephadex (with a glass rod, not with a magnetic stirrer to avoid breaking the beads) and decant fine particles. Repeat five times.
    3) Pack column with sephadex
    4) Flush column once with acetone
    5) Add more acetone, then elute until air barely touches the top of the gel bed
    6) Add a small amount of undiluted polyurethane sample on top
    7) Do the rest of the steps described by chemisttree.
    8) Collect eluent in a different container every few minutes (five minutes?).
    9) Spot TLC plates with eluent
    10) View TLC plates under UV light.
    11) Visualize TLC plates further with halogen vapors.
    12) Stop eluting when TLC plates no longer give a positive result for a sizable length of time. (1 hour?)
    13) Collect analyte fractions in sample vials and cap them.
    14) Flush column with three columns worth of acetone.
    15) Store column in a gas jar filled with more acetone.
    16) Place gas jar in fridge set to 4 degrees Celsius to store sephadex for reuse.
    17) Send sample vials to MS guy, plus a sample vial containing an acetone blank for further analysis.

    That's about it, right?

    Oh and there's seem to be another problem -- my columns don't have stopcocks. How do I regulate the flow of solvent? Should I use the burette as a makeshift column instead?
    Last edited: Sep 17, 2013
  10. chemisttree

    chemisttree 3,717
    Science Advisor
    Homework Helper
    Gold Member

    When you write "develop" the TLC, what do you mean? You don't need to place the TLC plate partly-immersed in solvent and wait for the solvent to move up the plate to near the top. You only need to visualize the spot itself. One TLC plate can visualize the presence or absence of material for the entire experiment.

    At the end of your experiment, you will have a test tube rack containing perhaps 30 or more samples collected from the column. Take a TLC plate and spot each sample in an array similar to the "array" of samples you have in your test tube rack. Each plate will have as many as 30 or so spots. These will never be treated with solvent but will instead go straight to the iodine tank. The samples that contain your separated polymer will be dark and the samples that have a weak (acetone blank) spot will contain nothing. Hopefully your results will show nothing at first, then one or more samples which show a dark spot which gradually becomes less dark, then nothing again and finally a dark spot followed by nothing again. Combine groups of the samples that contain separated polymer, ignoring the blank samples. A binary mixture should produce two groups of samples that you combine to produce two new samples... one which elutes early and one that elutes later. Of course that's the ideal case!

    You should dilute your sample with some solvent before applying it to the column. 10‰ is as strong as I would make it.
    Last edited: Sep 17, 2013
    1 person likes this.
  11. Results:

    I got three fractions. The fraction first to elute binds to iodine but does not absorb UV light -- this is probably the high molecular weight fraction that is also chemically inert (no functional groups). The second fraction elutes straight after, binds to iodine and absorbs UV light -- this is probably the high molecular weight fraction with functional groups. The third fraction eluted significantly later. This one absorbs UV light but doesn't bind to iodine to any appreciable extent. This is probably the low molecular weight fraction that I'm looking for.

    The first and second fractions are not separated properly though -- the second elutes right after the first, with no gap (blank samples) in between. Some components in the first fraction may also be in the second fraction. I will try again with a slower flow rate (by plugging to exit hole with a tiny amount of cotton wool).

    (I don't have any solvent-resistant tubing, so I can't use the "attach a tube to the column and tighten it with clip" method to control flow rate.)

    As a side note, what's the best way to aggressively clean suspect glassware with minimal equipment? I only have acid gloves and goggles (and a fume cupboard), with no face shield or acid apron, so chromic acid and piranha acid are out of the question. Currently I'm soaking suspect glassware in an acid bath, then a base bath (after rinsing out the acid), before rinsing and scrubbing with distilled water and rinsing with acetone multiple times.
    Last edited: Sep 21, 2013
  12. chemisttree

    chemisttree 3,717
    Science Advisor
    Homework Helper
    Gold Member

    Thats enough.
  13. chemisttree

    chemisttree 3,717
    Science Advisor
    Homework Helper
    Gold Member

    If you intend to collect the first and second eluting fractions for further separation, it might help to change the column conditions to achieve better separation the second time around. The first-eluting samples that may be co-mixed somewhat under the separation conditions you employed might resolve better if the stationary phase were further swelled with a little bit of "stronger" solvent listed in the table in the Sephadex LH-20 protocol I linked to in my first post. In this case a stronger solvent would be one that swells the stationary phase more. You have a lot to choose from that may be available to you. Methanol, Ethanol, Chloroform, Propanol, Isopropanol, Methylene chloride, added at 10% to 20% will swell the particles more than just plain acetone and should result in a better separation. You'll need to pack the column using more solvent and use the same concentration of the stronger solvent to elute. Methylene chloride would be my first choice if available. Make sure of purity with the TLC, of course.
Know someone interested in this topic? Share this thead via email, Google+, Twitter, or Facebook

Have something to add?

Draft saved Draft deleted
Similar discussions for: Protocol and solvent help needed for gel-permeation chromatography