Problem on HPLC determination

[Crystal]

column: C18
UV 254nm
Flow rate: 1.5mL/min
Mobile Phase:
A: 0.1M NaOAc,pH5.0
B: 50:50 ACN:water
Gradient: 5% to 60% B over 15mins

I used this system to determine the vit C(ascorbic acid),B6(pyridoxine),B1(Thiamine) and B2(riboflavin) in multivitamins tablet. Good separation is obtained. However, there is a broad peak and tailing of B1 instead of a sharp peak. I tried to increase the % of ACN and flow rate to 1.7, the peak width decreased but still broad. How can I solve this problem?

Moreover I want to know is it right to dissolve the standard and sample to the mobile phase A(NaOAc)? I found that it is difficult for B2 to dissolve in this mobile phase, how can i improve this problem?

[Mark Tracy]

Thiamine is very poorly retained, and does tend to tail. If possible, use an aqueous-compatible reversed-phase column (polar embedded or polar endcapped) and start with 100% buffer to separate thiamine and ascorbate. You may need to evaluate a couple of different columns before finding the one you like.

Riboflavin is poorly soluble in everything; aqueous buffer is better than any organic solvent. You just have to live with a stock solution of around 100 µg/mL.

[Uwe Neue]

In addition to Marks suggestion, I recommend to use the Atlantis dC18 column from Waters. It works in 100% water without difficulty and has more hydrophobic retention than polar embedded phases. With this column, you can start your gradient in 100% water.

[Einar Ponten]

You use these:
A: 0.1M NaOAc,pH5.0
B: 50:50 ACN:water

Thus, you run two gradients at the same time. You change both solvent and salt/buffer concentration when going from A to B. I don't think that is healthy considering that thiamine is a base.

Once we did a method validation of an ion-pairing method for thiamine and found that we had simliar problems as you when the initial ratio of "strong solvent" was too high (methanol >50%). Acetate is your counter ion and it would be a good idea to have that in eluent B as well.

Normally your method works fine. Maybe you now have an "old column" that have more free silanolic groups available. That would surely give you a broad and tailing peak, especially without a counter ion (in B).

[Mark Tracy]

One other comment. I have found that the analysis of vitamins is matrix dependent. The interferences found in Brand A vitamin pill are not the same as Brand B, and certainly not the same as juice beverages. You may need to fiddle around with column type, gradient and pH to optimize the separation of the vitamins from the matrix.

If you want to see how I did vitamins (more than one way), browse the Acclaim Library on the Dionex website: http://www1.dionex.com/en-us/acclaim_li ... 29396.html

[Bryan Evans]

We have a great C18 column for vitamin B assays.
http://www.imtakt.com/TecInfo/TI163E.pdf

You may also want to try a pure silica column for
vit c and vit b assay:
http://www.imtakt.com/TecInfo/TI159E.pdf
http://www.imtakt.com/TecInfo/TI154E.pdf

[Einar Ponten]

What Brian suggests as an alternative is HILIC and if you consider that we are also able to support you.

http://www.sequant.com/sn/ufiles/SeQuan ... 00-12A.pdf