HPLC Retention Interpretation

We already had mentioned nonporous packing material several times. What is it and how good is it compare to the traditional porous silica?

This development has some history. More than 15 years ago, Unger study the behavior of nonporous packings in HPLC. At that time his work did not attract much of the attention.

The progress of the silica chemistry recently allowed the synthesis of nonporous uniform spherical silica particles. These particles as it was shown by Kovats, has almost no particle size distribution. This distribution is less than 1%. This means that for the spheres with 1.5 m in diameter we do not have any particles with either 1.48 m, or 1.52 m.

The absence of the particle size distribution allows very uniform packing of these particles into the chromatographic column, which give very high efficiency. Since particles are nonporous we also do not have dual flow profile (usually flow around the particles in the column faster than the flow inside the pores), which also adds to the efficiency.

Columns packed with nonporous particles have some drawbacks. These columns has high backpressure, since particles are 1.5 m the interparticle space is small which gives high resistance to the flow, so only the short columns could be used.

Adsorbent surface area is small, roughly about 4 - 5 m2/column. HPLC retention is somewhat proportional to the adsorbent surface area. According to the adsorption approach to the description of HPLC retention, the retention volume, VR, of the analyte could be expressed in the form


where V0 is the column dead volume, S is the adsorbent surface area, and KH is the Henry constant, which represents the analyte interactions with the adsorbent surface.

To get analytes retained on that type of the adsorbent we have to increase the KH value significantly, which could be done primarily by decrease the content of the organic modifier in the eluent. On the picture below you can see the separation of parabens on regular column packed with the porous material (A) and the same mixture separated on the nonporous packing. Note the difference in the eluent composition, 70/30 MeCN/Water for porous, and 20/80 for nonporous.

This type of column could have the advantage if someone needs to run several hundreds of relatively simple and clean samples in the reasonable period of time.

Nonporous materials are very interesting from the point of view of understanding the HPLC retention process.

It is still a controversy between two different theories. One is based on the idea that the analyte could partition between two phases: mobile and stationary, mobile phase is considered to be the eluent and the stationary phase is that bonded layer of the C18 chains on the silica surface. But alkyl chains on the surface are anchored to the silica matrix and in addition to that they have limited mobility which depends on their bonding density. The higher the bonding density the lower their mobility. That is why this phase could not be considered as a liquid like phase where actual partitioning could happen. Also, monomolecular layer could not be considered as a phase from the thermodynamic point of view.

On the other hand in some cases this approach gives positive practical results. The behavior of certain analytes in the HPLC column fits to the partitioning description. This suggests that some type of the partitioning process is actually happen inside the column.

Another theoretical approach to the description of the HPLC process is adsorption. Adsorption assumes the interaction of the analyte molecules with the surface of the modified adsorbent, and its retention on that surface. The description of that approach is given in the other chapter.

Adsorption approach could explain some phenomena, like elution of some analytes before the dead volume, or tailing of only one analyte in the mixture. But it has also some drawbacks.

Retention of the analytes eluted with mixed eluent could not be easily described, since in this case we deal with multi-component adsorption. The description of the dynamic multi-component adsorption is very complex.

In the process of size exclusion chromatography we avoid any types of the interaction of the analyte with the adsorbent surface. Small molecules are able to penetrate in all pores and they are eluted with the dead volume, on the other hand, big molecules, which could not penetrate in any pores are coming out much faster with exclusion volume. This shows that there is a significant difference in the flows around the adsorbent particles and inside the pore space.

We mentioned dead volume several times already, but we did not discuss what is it. The definition is simple; this is the volume of the liquid phase inside the column. The question is how to measure it? Somebody will say, this is also simple, just inject "unretained" component and measure its retention volume. Here is another question, how to find this "unretained" component? Many suggestions has been made, uracil, KNO3, other salts, eluent components, etc. Unfortunately, if you try them all at different eluent compositions, they all will give you different values, so this is not acceptable. You never know how particular molecule will behave at the certain conditions on the certain type of the adsorbent. It could be slightly retained, or it could be excluded from the pore space. There are several ways to measure the true dead volume value, but all those ways are very complex and suitable only for serious thermodynamic study. For any analytical applications the rough approximation, but consistent one will be OK.

The simplest way to estimate V0 is to calculate the volume of the empty column and multiply it on 0.7 coefficient (you actually can use 0.65, or even 0.6, just be consistent). This method give you rough estimation with approximately 20% error. There is another way to do it with better precision.

We mentioned above that you never know how particular molecule will behave. Let now try to imagine how the molecules of the eluent will behave. If we have one component eluent it is simple. Actually, if we find out the way how to paint and detect some of those this will be the method to measure the dead volume. Unfortunately even the deuturated components show different retention.

If we have two-component eluent, the behavior of the molecules of different components will be different. Let us consider acetonitrile-water mixture on the reversed-phase column. We have relatively big interparticle space and tiny pores, which is about 500 times less than the particle diameter and they are comparable with the molecular size. Do you think the water molecules would like to go to these tiny hydrophobic holes? An acetonitrile molecule does, but water hardly so.

If we are pumping 50/50 acetonitrile/water composition, after a short equilibration we will have 50/50 in the interparticle space and somehow different, but definitely acetonitrile enriched composition inside the particles.

There is an indirect confirmation to this effect. The excess adsorption of the acetonitrile from water on the reversed-phase adsorbent has been measured many times. [] In those publications it has been attributed to the surface accumulation, and this accumulation may be pretty significant, up to 20 mMoles/m2. Where is this happens - inside the pore space.

We can easily estimate how much we need to fill the total pore space of one gram of adsorbent. Average parameters are S = 200 m2/g, Vpore = 0.7 ml/g for C18 modified adsorbent with original pore diameter of bare silica 120 .

Acetonitrile has density (dmole) approximately 45 ml/mole.

The amount needed to completely fill all pores could be related to 1 m2 as


This is just four times higher than the maximum of the excess adsorption isotherm of acetonitrile from water. I especially emphasized the word excess, this is the excess over the equilibrium amount, which is about 45 -50%. If we take 50/50 composition, this is already 40 mmole/m2 of acetonitrile plus the excess of 20 mmole/m2 which makes it over 75% the concentration of the acetonitrile inside the pore space.

As a result if we are pumping the 50/50 eluent composition we have that composition between the adsorbent particles, but inside those particles our composition is about 70/25, this is the big difference. There is certainly some concentrational distribution, since we have pore size distribution and active flow through the particles, but there is a difference.

What is this difference will give us? The whole new understanding of the HPLC retention process.

We can say now that HPLC retention process has dual nature. First the analyte has to partition into the porous space, which has different eluent composition than the space around the particles. After the analyte had partitioned inside the pores it could get adsorbed on the surface of the stationary phase.

This approach could explain some discrepancies in both existing theories of HPLC retention: adsorption and partitioning.

If we consider the situation when two completely different analytes are co-elute. This may be the consequence of two opposite effects. One analyte may effectively partition into the acetonitrile enriched space, but it may do not like to get adsorbed on the alkylated surface, and the other analyte may be able to get adsorbed on the surface but do not like to partition inside the pore space. This opposite behavior will result in co-eluting of those analytes.

This interpretation of the HPLC retention makes the porosity of the adsorbent a very important factor. The smaller the pore size the higher the difference in the eluent composition inside and outside the pores, and the more valid the partitioning effect.

Slide 8
Slide 9