Separating vitamin D2 and D3 in RP column

[vahur]

Hello,

Im trying to separate vitamine D2 and D3 and am using isocratic method with 93% water and 7% MeOH. The retention times are about 13 minutes and the peaks aren´t fully separated. Can anyone suggest ways to separate them more efficiently? Or is that mixture near to ideal for separation and my column isn´t just good enough?

[tom jupille]

Impossible to tell without a *lot* more information:
Column type (the exact brand and name)?
Column dimensions and particle size?
Flow rate (or what was the dead time)?
The exact retention times of your peaks?
The widths of your peaks (or what was the plate count)?
Did you see significant tailing (what was the asymmetry factor)?

[unmgvar]

in most cases of RP the 2 compounds will not fully seperate.
we use to have a method using merck lichroshpere that we used and we switched to a cosmosil cholester column.
it performs a lot better for these 2 compounds.
the difference between the 2 is the presence of a double bound in one, and a single bound in the other on one of the positions.
this gives them a difference in molecule "volume", and cholester rigid structure gives a better selectivity for them.
nacalai have an application on line for it on their web-site

[zheyin]

Hello,

Im trying to separate vitamine D2 and D3 and am using isocratic method with 93% water and 7% MeOH. The retention times are about 13 minutes and the peaks aren´t fully separated. Can anyone suggest ways to separate them more efficiently? Or is that mixture near to ideal for separation and my column isn´t just good enough?

Have you tried Acetonitrile?

[vahur]

Impossible to tell without a *lot* more information:
Column type (the exact brand and name)?
Column dimensions and particle size?
Flow rate (or what was the dead time)?
The exact retention times of your peaks?
The widths of your peaks (or what was the plate count)?
Did you see significant tailing (what was the asymmetry factor)?

Column: Agilent Hypersil C18 4,0x125mm, 5um
Flow rate: 0,8ml/min
asymmetry: (D2, D3) 1,221; 0,866
width: 0,416; 0,4365
dead time: 1,35
tr: 11,43; 12,12

in most cases of RP the 2 compounds will not fully seperate.
we use to have a method using merck lichroshpere that we used and we switched to a cosmosil cholester column.
it performs a lot better for these 2 compounds.
the difference between the 2 is the presence of a double bound in one, and a single bound in the other on one of the positions.
this gives them a difference in molecule "volume", and cholester rigid structure gives a better selectivity for them.
nacalai have an application on line for it on their web-site

D2 has 1 methyle group also more if wikipedia has it right.

Have you tried Acetonitrile?

I haven´t tried acetonitrile. What does it change if I switch between MeCN and MeOH?

[vahur]

One thing also fascinated me, the original eluetion program was following
Time,MeOH%,water%,flow rate
0;93;7;0,6
4,5;93;7;0,6
5;93;7;0,8
16,4;93;7;0,8
16,8;93;7;0,6
What is the reasoning behind changing the flow rate?
Also as you can see, i mixed up the water and MeOH % in first post.

[tom jupille]

I have no idea why someone would want to change the flow rate (might have seemed like a good idea at the time). It does complicate things a bit.

As far as efficiency is concerned, if the widths that you gave are baseline widths, then you're seeing about 10,000 plates (the change in flow rates means that I can't calculate it exactly), which is a respectable value for that size column.

As far as retention goes, your k' values are up near 10 (again, the change in flow rate complicates things), so that's about as good as you're going to do.

That leave selectivity; trying to move the peaks relative to one another. Basically, the parameters you have available are:
- temperature
- solvent chemistry
- column chemistry

Soooo, try again, changing the temperature by 15 degrees (either way) and see what happens. If resolution gets worse, then try again in the other direction. If it gets better, then optimize. If there is little or no effect on resolution, go on to the next parameter.

Switch from methanol to acetonitrile (you'll probably have to cut the concentration a bit to get equivalent retention) and see what happens.

Finally, go searching for a better column.

Good luck (your're going to need it!).

[zheyin]

I have no idea why someone would want to change the flow rate (might have seemed like a good idea at the time). It does complicate things a bit.

As far as efficiency is concerned, if the widths that you gave are baseline widths, then you're seeing about 10,000 plates (the change in flow rates means that I can't calculate it exactly), which is a respectable value for that size column.

As far as retention goes, your k' values are up near 10 (again, the change in flow rate complicates things), so that's about as good as you're going to do.

That leave selectivity; trying to move the peaks relative to one another. Basically, the parameters you have available are:
- temperature
- solvent chemistry
- column chemistry

Soooo, try again, changing the temperature by 15 degrees (either way) and see what happens. If resolution gets worse, then try again in the other direction. If it gets better, then optimize. If there is little or no effect on resolution, go on to the next parameter.

Switch from methanol to acetonitrile (you'll probably have to cut the concentration a bit to get equivalent retention) and see what happens.

Finally, go searching for a better column.

Good luck (your're going to need it!).

Temp for these 2 cmpds does not help.
C18 columns are all the same to such two hydrophobilc nonpolar cmpds.
acetonitrile...

[unmgvar]

http://www.nacalai.co.jp/cosmosil/data/ ... .png&lc=JE

[tom jupille]

C18 columns are all the same to such two hydrophobilc nonpolar cmpds.

I was going to reply that C18 is not the only type of column out there, but unmgvar beat me to it with a very nice link!

[vahur]

http://www.nacalai.co.jp/cosmosil/data/ ... .png&lc=JE

Unfortunatelly changing the column is not an option where I work. Thank you all for your help. I will try to use acetonitrile.

[unmgvar]

for stereo isomers, going to very low temperature most of the time improves the separation
we are talking going to 5-10 degrees set in the cooler
i wonder, if not switching the column is because of budget reason, then remember that ACN costs about 3 times more than MeOH.
so actually buying another column instead of switching between the 2 solvents can be cheaper over time, and the more the application is used, the faster the column repays itself.

[Scott F]

Dear Vahur,

When you state that changing the column is not an option does that relate to;

1) That specific brand,
2) That type of stationary phase,
3) Specific column dimensions,
4) Is it a pharmacopoeial method or fully validated method?

Thanks,
Scott

[vahur]

for stereo isomers, going to very low temperature most of the time improves the separation
we are talking going to 5-10 degrees set in the cooler
i wonder, if not switching the column is because of budget reason, then remember that ACN costs about 3 times more than MeOH.
so actually buying another column instead of switching between the 2 solvents can be cheaper over time, and the more the application is used, the faster the column repays itself.

I thought ACN could be more expensive but didn´t know by how much. So probably switching to ACN is not acceptable also. I´ll try to change the temperature tomorrow. Thanks alot for the info. I need a little reality check because just came from university

[Scott F]

I would be very interested to see whether reducing the temperature provided the resolution you require - you may find what you gain in selectivity you lose in efficiency and as such reduce resolution.

I have seen two applications where Vit D2 and D3 has been separated on C18 columns, I am sure there are many more, the first was on old Type A silica support, not too dissimilar for your C18 column you are currently using, the lack of endcapping and higher energy silanols helped the separation. The latter was using a polymeric C18 which aids in shape selectivity, hydrodynamic volume, such as that used in the application below. A lab I am working with has just employed the below application successfully in order to achieve baseline resolution.

http://www.chem.agilent.com/Library/app ... 5342EN.pdf

Like I said - please do post back if reducing the temperature works as I would be interested to hear.

Scott