why there is no kovats retention index in LC

[jiangds06]

it will be also as useful in lc as in gc?

[tom jupille]

it will be also as useful in lc as in gc?

No. If it were, people would be doing it . The fundamental reason that it's *not* done is that LC is inherently more complex than GC.

In GC the mobile phase is essentially a passive participant in the process; a gas is simply too dilute to have any chemical interaction with the sample (just as an example, the selectivity doesn't change if you switch the carrier gas from nitrogen to helium). That means that the only things that matter are chemical interactions with the stationary phase and volatility (boiling point). The Kovats index is based on a "ruler" of non-polar compounds (n-alkanes) which vary systematically in boiling point but have no other chemical interaction with the stationary phase.

In LC, the mobile phase plays a very important role in the process (switching from water to methanol makes a *huge* change in retention and selectivity, for example). As if that weren't enough of a problem, there is no good way to come up with "non-interacting" compounds for the ruler (remember, there are *two* phases with which they can interact!).

[sepscientologist]

You can do something sort of like kovats. I will look up the log Pow and can usually at least get an idea
of where my target analyte might come out in a reversed phase method. For those believers in the hydrophobic
effect being pretty dominant it's not a bad place to start. The text "Exploring QSAR" contains about 17,000
octanol water partition coefficients.

[jiangds06]

Thanks. Your statement makes great sense!
But non-polar compounds do have dispersive interaction with stationary phase, right? I guess they called out each other when calculating kovat index.

it will be also as useful in lc as in gc?

No. If it were, people would be doing it . The fundamental reason that it's *not* done is that LC is inherently more complex than GC.

In GC the mobile phase is essentially a passive participant in the process; a gas is simply too dilute to have any chemical interaction with the sample (just as an example, the selectivity doesn't change if you switch the carrier gas from nitrogen to helium). That means that the only things that matter are chemical interactions with the stationary phase and volatility (boiling point). The Kovats index is based on a "ruler" of non-polar compounds (n-alkanes) which vary systematically in boiling point but have no other chemical interaction with the stationary phase.

In LC, the mobile phase plays a very important role in the process (switching from water to methanol makes a *huge* change in retention and selectivity, for example). As if that weren't enough of a problem, there is no good way to come up with "non-interacting" compounds for the ruler (remember, there are *two* phases with which they can interact!).

[tom jupille]

But non-polar compounds do have dispersive interaction with stationary phase, right? I guess they called out each other when calculating kovat index.

Right. They are ubiquitous, consistent, and weak.

The "enabler" for any retention index is that there exist a set of compounds that meet the following requirements:
- the *selectivity* (the relative spacing of the peaks) does not change from one stationary phase to another;
- enough of them are available to cover a wide range of retention
- they are well-characterized and available in good purity
The alkanes meet those requirements for GC, which is why Kovats used them. There really is no general equivalent in LC; as soon as you add enough complexity to the molecules to avoid volatility issues, you add secondary interactions.

As sepscientologist pointed out, there are other metrics such as the octanol/water partition coefficient that give useful information.

As an aside on the whole subject, my personal opinion is that the major value of the Kovats index concept was that it formed the basis for the McReynolds constants characterizing the selectivity of GC stationary phases. McReynolds work demonstrated that the several hundred liquid phases then in use could be replaced by a dozen or so well-chosen materials. But that's a discussion for the GC section!

[jiangds06]

hydrophobic interaction in lc is equivalent to boiling point in gc.
heavy alkanes (liquid at RT) might also fit your requirement for lc (Use ELSD as detector)
If there are only 6 major types interaction of solutes and stationary phase in the reverse phase separation, only steric hindrance interaction will give different selectivity for these alkanes on different stationary phases. But I guess this interaction is small compare to hydrophobic interaction. As long as interaction between mobile phase and solutes, could we measure LC-kovat index under one gradient method.

But non-polar compounds do have dispersive interaction with stationary phase, right? I guess they called out each other when calculating kovat index.

Right. They are ubiquitous, consistent, and weak.

The "enabler" for any retention index is that there exist a set of compounds that meet the following requirements:
- the *selectivity* (the relative spacing of the peaks) does not change from one stationary phase to another;
- enough of them are available to cover a wide range of retention
- they are well-characterized and available in good purity
The alkanes meet those requirements for GC, which is why Kovats used them. There really is no general equivalent in LC; as soon as you add enough complexity to the molecules to avoid volatility issues, you add secondary interactions.

As sepscientologist pointed out, there are other metrics such as the octanol/water partition coefficient that give useful information.

As an aside on the whole subject, my personal opinion is that the major value of the Kovats index concept was that it formed the basis for the McReynolds constants characterizing the selectivity of GC stationary phases. McReynolds work demonstrated that the several hundred liquid phases then in use could be replaced by a dozen or so well-chosen materials. But that's a discussion for the GC section!

[Victor]

There was an attempt at doing this by Roger Smith many years ago. I don't know if the paper below was the last on this subject, or whether there are more recent ones. However, I don't think this method was widely adopted at the time, probably for the reasons that have already been given,

Application of retention indices based on the alkylarylketone scale to the separation of the local anaesthetic drugs by high-performance liquid chromatography Original Research Article
Journal of Chromatography A, Volume 355, 1986, Pages 75-85
Roger M. Smith, Tony G. Hurdley, Richard Gill, Anthony C. Moffat

[jiangds06]

thank a lot. will come back after I finished reading this paper.

There was an attempt at doing this by Roger Smith many years ago. I don't know if the paper below was the last on this subject, or whether there are more recent ones. However, I don't think this method was widely adopted at the time, probably for the reasons that have already been given,

Application of retention indices based on the alkylarylketone scale to the separation of the local anaesthetic drugs by high-performance liquid chromatography Original Research Article
Journal of Chromatography A, Volume 355, 1986, Pages 75-85
Roger M. Smith, Tony G. Hurdley, Richard Gill, Anthony C. Moffat