"Invisible" samples

<p><span>Hi, </span></p><p><span>I have a question about some "invisible" samples. That's the problem: I am currently working with solution of NCE in DMSO/buffer solution to check real solubility prior to cell test. Commonly my solutions have a content of sample varying from 10-5M to 10-7M of compound in 1-2% DMSOin buffer.</span></p><p><span>When tested in UPLC (same method transferred from HPLC) this solution shows no compound, apart a little bit in the 10-5M solution, while when tested on Alliance every solution shows compound around nominal concentration (more or less than 50% of the nominal)</span></p><p><span>If same solution are made in 100% DMSO they display a peak both in HPLC and UPLC.</span></p><p><span>Does anyone has an idea of what happen?</span></p><p><span>Just to remeber:</span></p><p><span>-Same solutions</span></p><p><span>-same gradient and phases (scaled by software) UPLC/HPLC</span></p><p><span>-XBridge C18 100x2.1 in HPLC </span></p><p><span>-UPLC BEH C18 50x2.1mm 1.7µm</span></p><p><span>I am thinking about a sticking to the column, but this is not common for small molecules.</span></p><p></p><p><span>t</span><span>hank you for any ideas</span></p><p><span>Metropolis</span></p>


  • Hi,

    Try to drop the draw rate on syringe down to 30ul/min (Advanced buton on sample manager page from Empower or Autosampler tab from Inlet editor in MassLynx), you may be having a problem with the viscosity of the DMSO preventing sample being drawn into the needle.


  • Hi,

    thanks for the information, I will try, but my sample have low amount of DMSO (maybe I wasn't able to explain in the right way!) ,around 1-2%, so viscosity doesn't seem to be the reason. I will try this morning and let you know.



  • Hello

    I am thinking that this is a solubility issue as the samples like the DMSO at 100% best. Alliance is a direct style injector and it pushes all the sample and its diluent on to the column, even when the samples are not ideally soluble. In the case of the ACQUITY a loop based injector there is the extra transfer step to the loop. Great for speed and reducing dispersion. But if your samples drop out of solution in the loop they will not make it to the column. The dual wash system will then efficently wash them out of the loop too, hence their "disappearance". I would thus optimize the sample diluent (100% DMSO) and as Outback says there is also the viscocity effect possible with DMSO and the syringe rate should be slow.

    I have asked a few other chemists to weigh in on this too.


  • Hi

    I have seen a similar problem to this once before, where an HPLC system worked fine and the Acquity UPLC did not show any peaks. This was due to sticking of analytes to the PEEK injection needle. One of the main differences between the Alliance system and the Acquity is the material the injection needle is made of. In Alliance the standard type needle is made of stainless steel, whereas the Acquity standard needle is made of PEEK. We do have other types of needles available, so if you are using the PEEK needle try switching to the stainless steel one (part number 205000362). This solved the problem in the case I was refering to.

    Best regards


  • Hi Liz,

    thank you for your reply.

    The question is : if sample is "in parking" in the loop has the same solvent in wich is dissolved (vial) so why does come out only in the loop?

    This would be a common problem: a lot of people do solubility testing with UPLC (like Waters Application note of P.Alden; D. Shave, K. Yu et.al).

    I think that there are other problems.

    For 100% DMSO you are right about viscosity.

    Let see what other people suggest.



  • Hi Rune,

    thank you for your interesting answer/suggestion: that sounds a better explanation, considering the low solubility of my compounds.

    I will try buying a steel needle.



  • I agree this is definately what I was alluding too I stated "loop", but I should have been more general, I agree with Rune my thought is that the action of drawing the samples through the injector that allows the sample to decide if it prefers the sample diluent or certain of the surfaces. Definately try the SS needle and as mentioned there are other choices as stated too, and as Outback said sometimes faster or slower syringe speeds can affect the outcome.

    Let's see what else comes up from the others if you do not mind...


  • Here is additional information from out chemists

    You should go ahead and order the SS needle but in the mean time, given that you are able to recover the sample when injected in 100% DMSO, they recommended that you should investigate solubility issues a bit further. The lower system volume combined with the fact that the gradient does not pass through the needle can set issues in borderline solubility situations. Are you using PLNO or PA? Could you be using Partial Loop with Pressure Assist (ie not PLNO) if so it could be that the weak wash is incompatible with the sample? You might try to formulate your weak wash for improved solubility and use the Partial Loop with Pressure Assist injection mode to "combine" the sample with the weak wash to improve delivery of a "solubilized" analyate to the column.


  • Hi Liz,

    commonly I used PLNO but I have tried also a full loop (10µL) because I wanted to exclude a "strange" low sensitivity.

    As a weak wash I use a 50/50 solution of water/Acetonitrile (my gradient starts with 70/30 TFA0.1%/CH3CN): does this affect PLPA injection mode?

    (Before I used a 90/10 TFA0.1%/acetonitrile as weak wash but nothing changes)

    Are you suggesting something else about this type of injection? I am new in the UPLC world, so please suggest everything could improve his assay.



  • Fiat Lux!

    Using Partial loop with 1 minute of loop offline my samples magically appear!


    1) does it mean that sample now reach column and the dissolve into mobile phase?

    2) I cannot understand the real "best" value for Loop offline: could anyone explain it to me? Could it determine some differences between sample set with different values?

    3) What about linearity/accuracy and so on in this mode?

    Thank you in advance


  • Paulo

    Can you send us your gradient table with the time for the loop off line included as this will help illustrate whta may be happening.

    Depending when you take off line it will have a composition reflecting what the system gradient was at that time which will be the contents the loop will show to the sample and diluent when you load. That composition could be less or more compatible with your sample and diluent.

    Many TX,


  • Hi Liz,

    Attached you can find the table of the gradient.

    I tried 0.5-1 and 2 minutes of offline, with no significative differences.

    Cuold you explain?

    Thanks for your time


  • When your sample has limited solubility in either the diluent or the initial mobile phase there is a significant risk that it may deposit itself on the surface of the sample loop during the sample loading process. With a gradient, the mobile phase will sweep the contents of the sample loop to the head of the column, but the solutes which have dropped out of solution will remain in the sample loop until the mobile phase is sufficiently strong to ensure that it dissolves. Since the volume between the outlet of the inject valve and head of the column is very small, the difference between a solute focusing at the head of the column or "depositing" in the sample loop is impossible to detect in the resulting gradient chromatogram. You may recognize this as "at column dilution."

    When loop offline is employed, there is a risk of failing to "elute" or dissolve the solute from the sample loop. The shortest loop off line time should correspond to the retention volume of the last peak of interest plus one column volume. if the gradient has an isocratic hold at the final composition, it should be long enough to ensure that the column "flushing" occurs (one column volume at the final backpressure is the minimum). If the loop is switched offline before that time, you cannot ensure that all peaks have left the column. the risk of peaks that are "missing in action" becomes real. Generally, you can acheive the same time savings that are anticipated from switching the loop offline by using the load ahead funtion in PLUNO mode.

  • Paulo

    Bear with me, we will get you a full explanation shortly. I am on old windows so I cannot see you gradient, but I will get printed shortly.

    Best regards,


  • Thank you.

    I believe that I have understand.

    Now I will make some experiments to be sure that all compund is flushed off.

    Thank you for all informations.


  • Hi Liz,

    thank you in advance for your incoming explanation.

    I'm sorry for word, but it was the only one in the UPLC PC.


  • Hello

    Apologies that this took so long...and I keep procrastinating about updating Microssft!!

    The principal reasons why "Loop Off-line" might be used are:

    (1) Reducing dispersion from the injector

    Reducing dispersion is better done by using full loop injections with a small loop (1 or 2 µL).

    (2) Reducing the contribution of the sample loop to the dwell volume.

    For ACQUITY UPLC systems, dwell volume is only significant if you have a large sample loop installed, (> 20 µL).

    Loop Off-line adds a layer of complexity that can be very difficult to diagnose, especially when first developing a method as it can cause missing and wandering peaks. This happens when the contents of the sample loop are not consistently flushed onto the column and out of the column. We usually recommend a 2 µL full loop injections when first developing a method. It ensures that only sample and diluent are introduced and gives the best reproducibility performance. Once the conditions are established try PLNO then PLPA, usually, in that order. I would recommend PLPA when you are sample limited.

    Loop Off-line causes complexity, in that the contents of the loop will be the mobile phase composition at the point the loop is taken out of the flow path. If some of the solutes "deposit" in the sample loop because of poor solubility, they will NOT be transferred to the column until the gradient composition dissolves the sample and flushes it onto the column. At that point the solute has a very low retention factor (k') and will elute from the column in one column volume. Hence the recommendation that you ensure that all of the peaks are flushed out of the sample loop by pumping the gradient volume + one column volume before taking the loop off-line.

    So finally, to answer your question...there is one other detail for you to consider. When Loop Off-line is used the injection sequence for PLNO changes, (although the needle washes are the same). For PLNO with Loop Off-line enabled, there is no a pre-wash (before aspirating the sample since it would require two additional injection valve moves). We postulate the samples are therefore not washed out and so I available/elute when the loop comes back on-line, whereas before they are eliminated with the pre-wash sequence.

    As you can see the loop based injector is very flexible, but this can also add complexity. Therefore, please do try full loop mode first and then see where we are.

    Best regards,


  • Hello,

    no problem for timing: I can read/answer you only today because of the earthquake that had make "some trouble" in our city (L'Aquila), but we are restarting our activities.

    Thank you for all answers, as soon as possible I will use them!

    Best Regards


  • So glad that you are safe and back on line!


Sign In or Register to comment.