HPLC Solubility vs polarity

Question

According to this data collected from PubChem, polarity of nitrobenzene and benzonitrile should be similar since solubility is dependent on the polarity:

Nitrobenzene: Solubility in water, 0.2 g/100ml

Benzonitrile: Solubility in water, 0.1-0.5 g/100ml at 22 °C

p-Chlorophenol: Solubility in water, 2.7 g/100ml at 20 °C

But according to the simulator (https://www.multidlc.org/hplcsim/4_2_0/) p-chlorophenol and nitrobenzene are more similar in polarity aspect, they both give peak at 2,2 min.

What is the reason of this difference? Is it H-bonding and being a charged molecule cause more affinity to the polar mobile phase, resulting a difference between solubility and polarity?

Answer 1

There are a couple of common misconceptions here. First a common mistake is to call a chromatogram a spectrum. A spectrum has nothing to do with a chromatogram generated by a chromatograph. Second common misconception is that aqueous solubility has very little to do with the order of elution in a typical C18 HPLC separation. What is perhaps somewhat relevant is the octanol water partition coefficient. You are completely ignoring the stationary chemistry in your reasoning. Third issue is that polarity is an ill-defined term. The fourth reality is that nobody can predict the elution time for a given black-box column chemistry from scratch for a given set of compounds. The simulator is built by a nice team. However, the retention time is based on experimentally observed data. There is nothing fundamental going on. Most of the retention models are based on interpolation of two or three experimental sets of data.

So look up the log octanol partition coefficients for three compounds and see if you can see a trend.

If you are deeply interested, then there is something called as a hydrophobic subtraction retention model. Forget about polarity, it is an umbrella term with very little utility/.