Can my standards be in a different solvent than my samples?

My samples are typically in heptane and my standards are in either acetone/heptane, heptane, or methanol.  Would there be an effect on quantitation?  I am using a S/SL injector at 240C with a 10:1 split flow (hydrogen).  Thanks!

Parents
  • Hi Anatoly,

    Actually I'll have to agree and disagree with .

    I was developing a method to quantify amylene in dichloromethane, first on a GC-MS, then on a GC-FID.

    Amylene has a boiling point (Bp) lower than CH2Cl2 and lower than any of the other solvents I had available in the lab to prepare my calibration solutions. So I started to try several, that wouldn't have the same m/z peaks (I was going to run in SIM mode) and I had to tweak the injector temperature and column starting temperature. This was not easy when Bp's were quite different.

    Since the amylene peak would come out before the solvent, the nature of the solvent was influencing the shape of the peak. Because I want peak symmetry between 0.8 and 1.2, that was an issue. So I settled for a solvent very similar to CH2Cl2 but without amylene as stabilizer of course. The best would have been to use CH2Cl2 stabilized with something else but that did not happen.

    Another issue is when I inject pure compounds for traces quantification. There is no solvent for the sample but I have to have one for the calibration solutions. That's where I use and internal standard to compensate for any differences (like viscosity) between the sample and the solvent of the standards.

    Btw: you are aware that hydrogen can reduce unsaturated compounds on hot metallic parts?

Reply
  • Hi Anatoly,

    Actually I'll have to agree and disagree with .

    I was developing a method to quantify amylene in dichloromethane, first on a GC-MS, then on a GC-FID.

    Amylene has a boiling point (Bp) lower than CH2Cl2 and lower than any of the other solvents I had available in the lab to prepare my calibration solutions. So I started to try several, that wouldn't have the same m/z peaks (I was going to run in SIM mode) and I had to tweak the injector temperature and column starting temperature. This was not easy when Bp's were quite different.

    Since the amylene peak would come out before the solvent, the nature of the solvent was influencing the shape of the peak. Because I want peak symmetry between 0.8 and 1.2, that was an issue. So I settled for a solvent very similar to CH2Cl2 but without amylene as stabilizer of course. The best would have been to use CH2Cl2 stabilized with something else but that did not happen.

    Another issue is when I inject pure compounds for traces quantification. There is no solvent for the sample but I have to have one for the calibration solutions. That's where I use and internal standard to compensate for any differences (like viscosity) between the sample and the solvent of the standards.

    Btw: you are aware that hydrogen can reduce unsaturated compounds on hot metallic parts?

Children
  • Now it's me to agree/disagree. Low boiling hydrocarbons and GC/MS can be a different story than classical detectors, especially because a MS is not fully rate-sensitive AND not fully concentration-sensitive (it switches sensitivities depending on amount of sample) and if EI or CI is used.

    Also in your case the type of phase/column and its (over)loading behaviour plays a role. In those cases it would be best to use a higher boiling solvent instead one close to the bp of the analyte (yes, you tried it but there might be a better solvent). And this would be true for very low polarity phases (close to bp separation for hydrocarbons), not more polar ones (Waxes as phase, polar organics as analytes) or PLOT.

    Maybe the original submitter can shed some light on what analytes need to be analyzed and what phase is used. Same would be true for you otherwise we keep speculating.

  • Nice discussion between csiquet_flex2000 and Dr.No.. But what about my comments i mentioned in this thread...

  • Hi ISH,

    peak area calibration is used for rate-sensitive detectors, peak height calibration is for concentration-sensitive detectors. Peak area is the first case is (almost) independent of retention or asymmetry, at least within some limits. Peak height is the same, but the peak area of these peaks is directly dependent of retention. As software often is used in peak area mode, even for concentration-dependent detectors you will get errors when a peak shifts due to being on the tail of a solvent peak. This is because your analytes interaction is not only with the stationary phase itself, but a solvent-saturated stationary phase, this changes selectivity. And ... you need good separation, true.

  • The GC-MS was only used to know where my analyte (amylene) was during method development because, most of the time, it was hidden within the solvent peak. As my sample is dichloromethane (with some ppm of amylene), I tried to stay in its boiling and polarity range when choosing the solvent for the standards.

    The method was then 'translated' for a GC-FID. Columns phase is HP-5.

    So I was just trying to explain that solvent choice was not straightforward for this method. Maybe because the analyte is a small molecule with a low Bp (38ºC) within a solvent with similar Bp (39,6ºC). Furthermore, even with a lab conditioned at 23ºC, it's not easy to have a column temp below 28ºC when trying to condensate the sample in it.

    So, as I said in another thread, I am still on the learning curve of GC and GC-MS and I appreciate all the insights.

Was this helpful?