Trouble separating Lipids

[kanryu]

Hey there,

I am really stuck with a problem here that drives me crazy!

So: I got a SOP from another group where they separate DPPC, DSPE-PEG and Cholesterol using a C18 (Acquity, 1,7µm, 2.1x50mm) column coupled to a CAD-detector.
They use 80% and 100% MeOH with 0,1% Et3N and the program is a follows: 80% MeOH for 1 minute, 100% MeOH for 8 minutes, another 5 minutes of 80%MeOH @ 0,2mL/min @ 50°C.

Thing is: The substances are only soluble in 100% MeOH and I need to inject them in 100% methanol.

I am using a 1260 from agilent and even bought a new column which I hoped would fit my needs: MN Nucelodur Isis, 3µm, ID 3mm, length 250mm.
I cannot use the program from the SOP because I don't have a CAD, we're only using a DAD (which is useless in this case) and an RID which means I can only run in isocratic mode (80% MeOH).
I tried injecting the 4µL they propose in their SOP (using a manual injector and build a loop from a peek capillary), tried 20µL, tried even 200µL but it seems that I only get one big peak everytime (probably the 100% MeOH?) no matter what substance I inject.

I tried to lower the temperature to 10°C and raising it to 50°C. I am running @ 0,2mL/min @ ~200 bar.

Should I try even more columns? The C8, C18 and MN Sphinx also gets me the same results. Could HILIC be an option? Is an RID the wrong detector? Besides: I can't go less then 80% MeOH, the stuff will just precipitate (and I can't heat up the injector). Injecting it in 80% MeOH also doesn't do the trick.
I am really desperate right now as we want to measure the lipid content of liposomes.

I hope somebody has some ideas I could try!!

Best regards
Marek

I am now trying for

[tom jupille]

What about going the other way and trying normal-phase (e.g. something like a CN column with dichloromethane / hexane) ?

[kanryu]

Our system is not that old and we don't have that much columns to try. We're running RP all time and never did NP, thus don't have columns for that. You'll probably even need other seals for the pump head and such?

[HPLCaddict]

Did you use run times that are long enough? With a 250x3mm column @ 0.2mL/min, the dead time is in the range of 5-6 minutes. And if your stuff precipitates with less than 80% MeOH, it seems to be veeeery hydrophobic, so even with a mobile phase of 80% MeOH retention might be quite high. Did you try a run time of, let's say one hour or so?

[kanryu]

My runtimes were about 15 minutes. I also tried with 0,5 and 1mL/min. I get a peak at about 6 minutes (the only one I get). I didn't see any ghost peaks or something in the later chromatograms. Do you suggest, I should try even more MeOH? i.e. 90%?
It really bugs me that I am limited to isocratic mode because of this shitty RID.

[Blazer]

What is the reason for you to try the method this group gave you? Meaning, is this a method transfer or are you just trying to apply their method to your work because you happen to work on a similar product? Regulated environment?

Essentially what I'm trying to get at is if you need to use the method as it is (or even just want to use it since the data shows it's pretty good), you'll probably want to get approval for a CAD. You could spend dozens-to-hundreds of hours trying to get a CAD method to work using an RID, changing the column and other parameters - essentially having redeveloped the method. I can almost guarantee it would be cheaper to just buy a CAD.

[kanryu]

The reason is: I have a starting point
They're also separating these three lipids with their method, unfortunately on UHPLC but still this should somehow work. No regulated environment, I can try everything I want on the machine .
I read some literature and at least got Cholesterol and DPPC separated from each other with a mixture (v/v) of 100:10:10 MeOH:Water:CHCl3. But still the PEG-DSPE doesn't show up OR comes right after the injection peak, meaning: it's hard to find a method that both retentates the peg more and still has the strength to separate the other two. I'll now try to shift the mixture a bit, I also tried with 90:10:10 but this also didn't really change too much in all the retention times.
I'd be happy to have a detector which is capable of gradients, that would make it easier in some cases, or at least: easier to optimize a method, when the isocratic one runs. ELSD should also be fine I guess. Maybe I'll ask if we could get another one. How much is it for a ELSD (I saw that Agilent doesn't have CADs in their portfolio)

[SciMania]

Hi,
Few suggestion to your problem(s) are;
1) Most lipids are not suitable for absorption detectors due to lack of chromophores.
2) RID is not a good alternate (limitation of isocratic mode which is not suitable for separating lipids).
3) ELSD (suitable for most lipids) with gradient mode has been reported by some groups to work well for separation (AAPS PharmSciTech, Vol. 14, No. 2, June 2013;J Pharm Biomed Analysis 51 (2010) 947–951). I recommend to try ELSD.
4) The column you mentioned should have void volume (time) around 6 mins (0.2 ml/min) so consider using void volume marker before analysis by any of the mode.
5) pH of mobile is an important parameter to consider while separation. Find the ionization behavior of your lipids at different pH and then try them.

Good luck