non-linearity of UPLC method

Options
BJ
BJ
in Instrumentation and Mass Spectrometry
<p>Hi,</p><p></p><p>I'm currently developing a new method for an API.</p><p>While testing linearity for related compounds, I found that my method isn't linear at the lower side of my calibration line (0.1-5% of my normal concentration).</p><p>Here are some data:</p><p>%      Response
0.1       2459.051963
0.25     6745.959772
0.5      15412.28079
1         35414.32085
3         124463.9612
5         216200.0982</p><p>I repeated this a few times, so I'm quite sure that the method isn't linear.</p><p>The API is a small molecule with a pKA of about 9.3, my method is a gradient of a buffer pH 7.0 and acetonitrile.</p><p>Dilutions are made up in buffer pH 3.75 (switching to pH 7.0 doens't solve my problem, already tried).</p><p>Injecting 10µl full loop on a 10cm x 2.1mm  BEH C18.</p><p></p><p>When I switch the entire method to HPLC (using a 10cm x 4.6mm XTerra MS C18), my linearity problems are solved.</p><p></p><p>So I think it must have something to do with the injection system of an UPLC, but I don't know what to do to get my method working.</p><p></p><p>Any ideas?</p><p></p><p></p><p>BJ</p>

Comments

  • lizh
    lizh
    Options

    Hello

    Can you send the exact details of the gradient table and the sample diluent. Can you confim that you are using Full Loop injections for each dilution.

    I am not at a computer but am tempted to ask,what does the curve look like without the two highest concentrations? However, I will aslo asked some applicain chemsits to review.

    BR

    Liz

  • BJ
    BJ
    Options

    Hello Liz,

    my gradient table:

    Time10mM ammonium acetate pH 7.0ACN
    09010
    3.44060
    3.69010
    59010

    Flow: 0.5ml/min on a Acquity BEH C18 10cm x 2.1mm


    My sample dilutent is 10mM ammonium acetate pH 3.75, but I also tried pH 7.0 and even plain water, all with the same (bad) result.

    I'm using full loop injections on a 10µl loop.

    Without the 2 highest concentrations, the curve is still non linear. To overcome questions about detector overload: the data I gave were only part of the whole data, I only gave 0.1 - 5%, but I also have data from 5% up to 130%, which is still linear, so no overload for even 130%.

    Thanks for your time

    BJ

  • lizh
    lizh
    Options

    Thank you for this

    I have spoken to one of our chemists, please try making up the dilutions with his buffer and ACN in approx. the same ratio as the related compound elutes in the gradient and see if that helps. She thinks the loss may have to do with solubility. Let us know. And I have sent her your method details.

    BR

    Liz

  • Newman768
    Newman768
    Options

    Hello,

    maybe you experience an absorption effect of your molecule within the injection needle (Question: do you use stainless steel, silica or PEEK needle?).

    You can check this by adding a similar "counterbase" with at least 100 times higher concentration and repeat linearity at conc 0,1% and 5%.

    Andreas

Categories