Mobile phase with acetonitrile, tetrahidrofurane and water

[kkovats]

I have developed an isocratic method for a steroid with the mobile phase H2O : CH3CN : THF = 66 : 21 : 13 (v/v/v). The column is Inertsil ODS3 3µm 150 x 4.6 mm. The temperature is 40ºC.

During the development one column failed, it lost completely the efficiency. It happened one or two days after experimenting the method at 60ºC. At this higher temperature only 2 injections were performed, but the supplier concluded that probably it was the high temperature the cause of the column failing and changed the column.

We went on with the new column and begin the validation. After one week work we noticed that, although the column did not fail completely, it "visually" lost the efficiency when compared to a new column.

At this point we stopped the validation, it is not worthly to validate a method and have a 1-2 weeks column life.

We tried several "equivalent" columns from different suppliers and choosed one of them that gave the more similar selectivity (Luna C18 3µm 150 x 4.6 mm). This column did not even resist one day of work, after overnight it failed completely.

I do not know what is happening. Theoretically the mobile phase is not so aggressive to distroy columns overnight. We have several methods with acetonitrile / water or tetrahidrofurane / water that do not give problems, but we do not have any method with the mixture of the three solvents.

Would the mixture of them be so aggressive?

Did anyone have the same experience?

The other columns that we tried and had acceptable selectivity were:
- Luna Aqua
- Supelcosil ABZ Plus
- Symmetry C18
- Symmetry Shield RP18
- Atlantis dC18
- Platinum EPS C18
- Nucleodur C18
(all type of columns that we are currently using...).

Does anyone have experience with these columns using mixtures of H2O : CH3CN : THF?

I would very much appreciate your opinion/suggestions.

Thanks in advance!

Kati

[DR]

The only thing I can offer is to note that THF is much more viscous than other organic solvents, so you should expect exceptionally high back pressures with the MP you're using. A lower flow rate may help.

[kkovats]

The pressure with a flow rate of 1.2ml/min was 2500-2700psi.

[Bryan Evans]

You may want to try running a blank injection, using a gradient,
just to make sure no analytes are sticking to the stationary phase.

If that does't work, we have had a similar method developed on our C18 columns. Our columns have polyfunctional ligand bonding and
polymeric endcapping, which does help to make them very durable:

http://www.imtakt.com/TecInfo/TI183E.pdf

[Uwe Neue]

You have used a broad range of columns under what I would consider to be rather mild conditions. I am wondering, if one should not look for the problem somewhere else, and not with the columns.

1. What is your sample (other than the fact that your analytes are steroids)?
2. Which of the columns in your list showed the failure that you described? Only Inertsil ODS 3 and Luna C18, or all of them?
3. What is the quality of your THF?

Let us start with that, and maybe this will lead to other questions.

[kkovats]

Thanks for the replays!

Bryan,

Usually overnight I washed the column with 100% CH3CN, so if there were sticked materials, they surely went away with the washing. In the morning then, after washing, the column just showed lost efficiency.

Uwe,

1. I suppose the problem is not the sample because we have analysed this product for years with two different methods. One CH3CN / H2O gradient method and another THF/H2O isocratic method. The columns were Nucleosil C18 and C8, respectively. The methods have been working well, the challenge now was to develop a unic method, selective for all the impurities that till now we only could "see" with two methods.

2. I only worked more time with the two mentioned columns (Inertsil ODS3 (one week) and Luna C18 (one day)). With the other columns I only performed stabilization and 2-3 injections, just to see the profile and selectivity, to be able to choose a new column.

3. Merck, HPLC grade THF.

Kati

[syx]

Does anyone have experience with these columns using mixtures of H2O : CH3CN : THF?

We have used ACN : THF : 0.02M ammonium acetate buffer pH 4.0 (27:20:53) with Luna C18(2) column for a species of statin determination. The column temperature is maintained at 25°. The flow rate is set on 1.5 mL per minute. We found nothing strange after several injection.

[Rafael Chust]

Being THF a stronger eluent than ACN, I would try 100% THF (assuming back pressure is not too high) or a mix between THF + ACN.

A column can lost efficiency basically due to pore clogging or C18 chains restraint.

I will send you a column maintenance procedure, to see if you can recover the columns.

[dr_Pyrex]

I really supprised about kkovats problems. I use succesfully CH3CN:THF:H2O (96:3:1) phase with Luna column (150 x 4.6 mm, 3 um) and 40C.

kkovats: what kind of tubing do you use: PEEK or steel?
THF can be dangerous with PEEK tubing...especially with high temperatures.

[syx]

THF can be dangerous with PEEK tubing...especially with high temperatures.

Dr Pyrex, how far is too high? And how much of TFA is needed to break the PEEK tubing on that temperature?

[dr_Pyrex]

I do not have any strict rules about high % of THF in HPLC. I found on the web some tips:
Phenomenex
"However, in line with the instructions from the manufacturers of PEEK plastic materials, we would recommend that you do NOT use 100% pure Tetrahydrofuran (THF), Dichloromethane (DCM) or Dimethylsulfoxide (DMSO)"

Agilent
"Expose tubing to specific solvents. DMSO, THF, and methylene chloride cause the PEEK tubing to swell. Concentrated nitric acid and sulfuric acid weakens the tubing."

Varian
"PEEK is chemically inert to virtually all solvents used in HPLC. Only methylene chloride, THF, and concentrated nitric acid are to be avoided"

In my opinion you should change (if you have an opportiunity) your tubing for steel and tried analysys again with new column.

[JM]

How old your HPLC grade THF is?? THF known to generate peroxides on long storage if no antioxident is present in it ( the grade you have may not have em) . This could creat problems.

JM

[kkovats]

We are using THF of 2.5l bottles, thay last 2-3 days, so every 2-3 days we open a new bottle.

The equipment we have been using does not have PEEK tubing.

Thanks!

Kati