Please, could you explain me what is "hydrophobic collapse"in RP-HPLC? Does this effect exist is NP-HPLC (hydrophilic collapse?)?

It is the sudden loss of retention observed with some reversed-phase columns in mobile phases with a very high water content, mostly 95% or larger. The immediate origin is often a drop in the column pressure, either by accident or by a reduction in flow rate. The phenomenon is an exclusion of the very hydrophilic mobile phase from the pores of the hydrophobic packing. If the packing has a large amount of residual silanols or a polar group that is incorporated into the packing, this "hydrophobic collapse" does not occur.

I think that what you are refering to is flattening, or collapsing of the hydrophobic 'tail' when there is insufficient 'organic' phase present in the mobile phase. This 'falling over' is common with so called 'mono-functional' hydrophobic phases, that is the hydrocarbon 'tails' are attached to the solid silica support by a single bond (through a silicon-oxygen link). The effect is reduced by using 'poly-functional' linkages, actually only two of the possible three linkages can be made, so they should be called 'bi or di-functional'
The moment of collapse will occur at different water/organic concentrations for different column packings, depending on many things, including %c content, or surface coverage. If the resolution of the peaks is sufficient for your needs, then there is no problem. Check the efficiency of the column with say pure methanol, then with only 10% methanol/water (using different compounds to give you adequate retention), or at whatever combination you are using. There should be little difference, for the same k';if there is alot of difference then you have lost that efficiency.
This phenomenon happens with long hydrophobic moeities (C18), and less with shorter moeities, C8, C4 etc. and obviously not at all with pure silica, alumina et al. but also be aware that the linking groups may add to some problems with diols, CN, NH2.
Cheers
Bryan Wallwork

Bryan, I am sorry, but I disagree. The phenomenon of hydrophobic collapse has been associated with a loss of solvent in the pores. It is a simple wetting phenomenon, or rather dewetting. One can calculate what the forces are that are needed to keep the mobile phase in the pores. They depend strongly on the wetting angle of the mobile phase to the stationary phase. If it exceeds a certain angle or if the pressure gets too low, the mobile phase is forced out of the pores. It is the same phenomenon that is used otherwise to determine the pore size distribution of a packing by mercury porosimetry.
Of course, it depends on the character of the stationary phase. For phases with a low content of bonded phase and conversely a high content of surface silanols, the phenomenon does not occur. One can run ye olde Resolve C18 in 100% water without ever seeing a collapse. It appears to be also absent from phases with trifunctional silanes with a low coating level, but I am not aware of any studies with truly highly coated trifunctional phases and very good endcapping (or other indications of low silanol content). One of the complications with trifunctional phases is the fact that for every bonded phases one creates at least one surface silanol. Thsi complicates the picture.

Uwe, very interesting. I'm not sure I understand (agree with) everything you mention, do you have any suggestions as to where I can find out more about this phenomenon, as I think it maybe particularly relevant to a problem I am having with the separation of haemoglobin variants on a weak ion exchanger? It defies all the common sense approaches to chromatographic separation!
On short columns (5mm yes, 5mm!)if a slug of air is introduced, the back pressure increases 2 to 3 fold, and returns to original conditions only after 10 minutes or so of intermittant running, this may suggest the exclusion of mobile phase (totally water based) from the pores(a mix of 100 and 1000A°). The stationary phase is polyaspartic acid covalently bonded to an amino phase on silica.
Any ideas?
Cheers and bon Noel
Bryan Wallwork

Here is the stuff on the surface wetting in pores. The background is the equation of Laplace and Young which describes the pressure needed to penetrate a pore as a function of the wetting angle and the surface tension. I am not sure how to do this well on a web site, but let me try:
P = - 4*gamma/d * cos(theta)
P is the pressure needed to penetrate a pore, gamma is the surface tension of the liquid, d is the opening diameter of the pore and theta is the contact angle. For a pore dameter of 10 nm. water with a surface tension of 72.8 dynes/cm and a wetting angle of 111 degrees, one needs over 10 MPa to get the liquid into the pores. (This is based on a calculation done by my colleague Ed Bouvier).
What is the stuff that is in your tiny column? Especially the particle size is interesting. And what is the actual pressure that you are observing? The polyaspartic acid stuff should be fully water wettable. On first glance I would therefore say that the wetting issue does not apply.

Uwe, thank you for that lucid explanation. How do you reconcile this surface tension effect with the recent NMR and spectroscopic studies that seem to indicate that binary mobile phases "de-mix" and leave the stationary phase coated with a layer of organic solvent.

Best Regards,

Tom

Since appearance of this chain I have been looking, unsuccessfully, for an article on NMR and phase collapse in my material. It was hoped to save us all time if more refs were presented here. However, I did find some secondary literature regarding collapse:
R. G. Wolcott, and J. W. Dolan, LC GC International, 12(5) 260, May 1999, and
JW Dolan, Readers´ Questions: Carryover, Mobile Phase Temperature and Column Care,
LC GC Europe, 12 (12), 754 - December 1999
755.
Therein, and in other places, it is mentioned that reversing a phase collapse can take considerable time (30 min . . .). I am very skeptical as it has never been observed here, and because diffusion of organics in a monolayer or pore (if the effect mentioned by Uwe Neue is pesent) should negate the effects in fractions of a second. Is there a third phenomenon active in some cases and confused with collapse or wetting?
Very common exposure to wetting effects are experienced with microfilters: water does not go through a teflon filter, air goes easily through a polar filter when the filter is dry, but not when wet.

to H W Mueller: water can pass through a ptfe (teflon) filter membrane (at least for a while) if it is wetted with methanol first. Solid phase extraction cartridges (C18 etc) are conditioned with methanol (and the like) to 'solvate' the hydrophobic tails, and to present a larger surface area to the analyte.
to Uwe Neue: The stuff in the column is polyaspartic acid phase, used to separate proteins.It has a nominal pore size of 300A (30nm) but is probably a mix of 100A and 1000A! The silica particle diameter I am working with is 5µm, with a back pressure of 250psi at 0.5 ml/min (not your typical conditions) on a 4.6mm x 50mm ss column. The microcolumn used is 5mm diameter by 5mm yes 5mm bed length, with a flow rate of 0.6 ml/min, and a back pressure of typically 40 (forty)psi.It is the latter column which has the problem with air bubbles causing high back pressures (>100psi) this is truely LOW PRESSURE liquid chromatography.
Cheers
Bryan

Tom,
I have not yet seen the study on the NMR work, and I would appreciate the reference.
However, I am not surprised. There have been a ton of investigations based on adsorption studies that have shown that the organic solvent is adsorbed on the surface of an RP packing (e.g. Karger a long, long time ago, and Jaroniec more recently).
H.W.,
The reversal of this effect takes no time whatsoever, if you rewet the column with 100% of an organic solvent, such as methanol or acetonitrile.
Bryan,
I suspect that the phenomenon that you encounter on the 5 mm column at 40 psi is nothing but a trapping of air in between the particles. Imagine this as being similar to the situation of two imiscible solvents in a packed bed. The migration of such a thing is really very strange and very unusual. No reasonable uniformity as we know it in normal chromatography. You could test if this is the issue by increasing the backpressure behind the column with a restrictor for a short period of time, maybe a minute. If the air is completely removed with this pressure pulse, this would demonstrate that the problem is along the lines described here.

Uwe,

If the organic is adsorbed then I am still unclear regarding the mechanism of hydrophobic collapse. The data that I recall seemed to indicate an organic layer of something like 7-14nm thus I assume that the pores would contain organic and bulk mobile phase surface tension would not be relavent. I will try to find some references when I have a chance. Any thoughts would be appreciated.

Tom,
The thickness of the adsorbed layer must depend on the solvent composition. A layer of the thickness that you describe is not realistic in a mobile phase with a high water content. I even have my doubts about such a layer in a low water content of the mobile phase. Don't forget that the average pore size of a "normal" C18 is only 10 nm.

Dear Uwe, Tom, Bryan and Mueller:
Thank you very much for your help!

Once collapse occures, does it mean the column is dead and can not be reversed by passing through the high percentage of organic solvent?

Thanks!

No, it's reversible, but it can take some time to reequilibrate. Probably the most insidious aspect of the problem occurs in gradients when you start the gradient at too low a % organic. The column then partially reequilibrates during the gradient itself, but small changes in conditions (temperature, for example) can change the equilibration rate so that you get irreproducible results from run to run.