Category: Analytical

Calibrations: The More the Merrier?

Actually, more isn’t necessarily better but none is not an option either.

Calibration curves are used to actually determine concentrations of compounds of interest in your samples.This idea being that the lowest and highest unknown concentrations are bracketed by at least 2 standard concentrations from your standard curve.  If a unknown is outside you calibration range, then you need to redo, to bracket that concentration in your curve, as long as the curve is still linear. If the curve cannot be increased, then dilutions are needed to bring the unknown concentration in the calibration range.

One can talk about the time interval between calibrations  (weekly, monthly, etc.). This interval depends on how complex or dirty the samples are. If degradation of the separation is occurring, then a calibration is warranted. EPA test methods, in particular, say what the recommended calibration interval is. Also, if the column is removed for any reason, then a calibration is warranted.

Also, there is the use of duplicate samples, calibration check samples, and spiked samples that can be used during a run. These samples are to access the viability of the calibration curve.  All of these should be used to some extent.

There is quite a bit of information in this post. The point is, one should not ignore these points because they are hard or make analysis more complicated. These should be used to better the analysis.

David Slomczynski, Ph.D; GeoMetrick Enterprises



HPLC Method Development around the World: Why it is so hard

Taking a method that was developed and published from another country and trying to use it in another country – why should that be so hard? Well, once, I was trying to reproduce an HPLC method from a Japanese group published in an Analytical Chemistry Journal (Honda et al 1991 Anal Biochem 199: 256). It turns out the columns used, in the paper, were not available NOR for sale in the US. I tried many columns, with similar characteristics (packing material, coating, etc.) and none worked (even with help from companies selling columns based on their knowledge). I basically had to start from scratch and develop my own system, which I did and it was totally different and more complex.

Another time, I was using an older method, which used high pressure solvent mixing, versus what I had, which was low pressure mixing and could not reproduce the separation. Again with modifications, I was able to get the right results. As it turns out, high pressure mixing, involves solvent compressibility calculations in its mixing algorithm and if not calculated properly can actually throw off the solvent mix the column sees.

So the take home points:

  1. Make sure all the instruments are the same (avoids void volume issues, etc.)
  2. Make sure Solvent and water quality are the same (this is easier now).
  3. Make sure the columns are the same and available.
  4. Use set standards and if possible the same standards

David Slomczynski, Ph.D; Geometrick Enterprises