Separation of Ethylene Glycol, Formamide, and DMSO

[ebasarah]

Hi All,

I am trying to separate and quantify Ethylene Glycol, Formamide, and DMSO present in organ tissue.

I successfully separated DMSO and Formamide but E.G came at the same RT as Formamide.

Any Idea?

Also why at some certain PH and buffer the E.G peak ( U.V absorption) could not be observed? When I did not use buffer and normal solvent E.G has some U.V absorption near 190nm (not strong but detectable)

Thanks for the help!

[Uwe Neue]

I would use a silica HILIC column, 80% acetonitrile 20% water, UV detection at low UV 210 nm.

[ebasarah]

Done that!

Hilic column could not separate them all. infact, everthing came together...

Thanks!

[Uwe Neue]

What were the solvent composition, the column dimension, the flow rate, and the retention time?

[Einar Ponten]

What kind of "HILIC column" did you use. I am quite sure that you should be able to separate these compounds by a HILIC mode separation.

Please let us know more about your previous experience in this aspect.

[Chris Pohl]

ebasarah,

One option would be to use an ion exclusion column. Even though none of these analytes are ionic, such a column can readily separate a variety of polar solvents by a partition mechanism.

Chris

[ebasarah]

I used Hilic 250x4.6mm, flowrate 1ml/min UV absorbance 190-250nm.

following condition had been tried:
Gradient Elution
A: Water
B: Acetonitrile

Gradient:
Time(min) A(%) B(%)
0-2.5 10 90
10-12.5 50 50

2. 12.5 minutes Isocratic Elution: 95% Acetonitrile/Water

3. Gradient Elution
A: 50:50 Acetonitrile:10mM ammonium acetate pH=3.8
B: 90:10 Acetonitrile:50mM ammonium acetate pH=3.8
PH was adjusted by Acetic Acid

Gradient:
Time(min) A(%) B(%)
0-2.5 0 100
10-12.5 100 0

4. 12.5 minutes Isocratic Elution: 90% 1-Propanol/Water
5. 12.5 minutes Isocratic Elution: 40% 1Propanol/Water
6. 12.5 minutes Isocratic Elution: 10% 1-Propanol/Water


Spherisorb ODS (2) 10% methanol/water gave nice separation between DMSO and Formamide but E.G came at the same retention time with Formamide.

I am planning to try REZEX column and currently I am playing around with LUNA NH2 column. Hopefuly there is strong adsorption of Formamide to the amino group bonded phase.


Note: When I used ammonium acetate buffer above, all of the peaks intensity were greatly reduced from ~1000mau to about 50-100 mau
Why it happened like this?

[Einar Ponten]

You have really done a serious work on this.

However, I am quite surprised that Test 3 did not work for you. What kind of column did you use for this test. Was it Luna NH2? You actually had a very high content of ACN so most of the analytes should have been retained. What kind of solution is the sample, more than 50% water?

A "trick" to deal with water samples is to reduce injection volumes to an absolute minimum (<1% of column volume). Otherwise the strong solvent you already have in the sample will elute all analytes with the void (just as you will find in a RP system when the sample is in ACN).

Using a UV detector at low wavelength means that acetate from NH4Ac will contribute significantly to the background. Try to reduce the buffer concentration. Probably you need an another detector for a proper S/N.

[ebasarah]

thanks Einar Pontén,

all the test I described earlier were done on Hilic column.

I am still wondering why Ethlyne glycol peak didn't show when I used KH2PO4 Buffer at Ph=3 but It did show at ph=4.52??? weird huh?

If I am dealing with Amine not amide, the problem could be easier since I could protonated amine under low pH. But NH2 on amide can't be protonated due ketone group therefore the electron is more delocalized

[Einar Ponten]

I understand that you wish to use phosphate buffer due to its suitability for UV detection. However, the solubility is very low in 95% ACN. That will surely cause problems.

These compounds does not require much pH considerations with respect to the mobile phase, while I really think that you need to find an other column for HILIC mode separations.

[Uwe Neue]

Thanks for giving more detailed info, and - as you can read from the responses - a lot of very knowledgeable people are getting interested in your problem. Your condition 2 should have given you something useful on any HILIC column. What were the retention times of your three analytes under condition 2, and what column did you use for this study?

[ebasarah]

Good morning all,

When I used buffer the concentration of organic solvent did not exceed more than 50%.

retention times of three analytes on hilic column were 3.3 min
Gradient Elution
A: Water
B: Acetonitrile
Gradient:
Time(min) A(%) B(%)
0-2.5 10 90
10-12.5 50 50
I used Luna Hilic Column ( diol bonded phase) from phenomenex

I will try soon using ammonium formate with the same condition above.
I have some ideas to try this week...I will keep you updated!

Thanks again for your response!

[Uwe Neue]

OK, I understand now: here is the problem: HILIC works with mobile phases of the opposite composition than reversed-phase. You want to work with a LARGE fraction of acetonitrile, and a small fraction of water. You do not need ANY salt for your analytes. I am also not convinced that the HILIC column that you have is the best choice for the task, but we will see after you use the correct mobile phases.

As I said above, use 80% acetonitrile 20% water. Nothing else in it. Use a low wavelength, 210 nm. equilibrate you diol column with this mobile phase until you do not see any drift any more in the baseline. Then inject the three analytes. If they elute too early, go to 90% acetonitrile and 10% water. No salt needed.

If you do not get retention under these circumstances, you need to get a better HILIC column. My recommendation is the Atlantis HILIC silica column from Waters. Start with the same mobile phase as mentioned above (80% acetonitrile).

[ebasarah]

Hi Uwe,

Thanks for the response...
Yes, I had done 90% ACN or even 95% ACN in water as mobile phase on my luna Hilic column. Even I tried 90% Propanol in water on the same column. But They were all eluted together no separation at all.

Buffer in my case is not that significant since my analytes are not readily ionizable.

Currently I just need to separate Formamide and Ethylene Glycol...It seemed like they have more interaction with the solvent than the adsorption to bonded phase surface.

What is the difference between Luna Hilic and Waters Hilic?

[mbicking]

Here's one more option:
I have been able to see some retention of alcohols and glycols using a polar embedded phase and 100% water as the mobile phase. In my particular examples, I lowered the column temperature to 20 C, because that option was available, but room temperature should be similar.

I used RI detection, but if you think you can see it with absorbance, try it. I would expect some retention for DMF, but DMSO might be a problem.

One other question: are you sure you are seeing your peaks? At short wavelengths, many things absorb and sometimes vacancy peaks can be mistaken for real peaks.