User experiences of fused-core particle columns?

[mardexis]

I don't know how these halo particles are made but I do like the idea of taking a mono-disperse collection of 1.5 uM Stober spheres and applying a superficially porous layer. You should have improved performance due to the extremely narrow PSD and the reduced stationary phase diffusional distance. I haven't yet seen Agilent take advantage of this in their poroshell line but maybe that is the basis of the suit.

That said, one would not expect the larger halo particles to be more efficient than well packed sub 2 micron.

[adam]

I have looked at a number of publications on this topic. One in particular did a nice comparison and concluded that the fused core particles gave 80% of the efficiency of sub-2-micron particles (i.e. UPLC). This work was done on an Acquity system so it should be a fair comparison. And it was done with small molecules. One would expect even better results with large molecules.

So the take away is that fused core particles are not quite as good as UPLC (for small molecules) but pretty close. It seems these particles do offer a way to get pretty high efficiency or speed without having to work at very high pressures.

- J. Sep. Sci. 2007, 30, 3104 - 3109

[mbicking]

Here are some more articles if you want more background. I use some of this material in my training seminars.

Kaczmarski, Guiochon, Anal. Chem., 79, 4648 (2007)
"Modeling of the Mass-Transfer Kinetics ..."
Heavy on theory but some useful plots. Makes that case that the big advantages are for big molecules. I know Georges Guiochon, and he is one of the most respected chromatographers around. I trust his judgement.

Kirkland, Langlois, DeStefano, Am. Lab., April 2007.
"Fused Core Particles for HPLC Columns"
Kirkland et al. work for the company that makes the particles, so they have an interest in the outcome. But the article is very readable with some nice plots. Like Guiochon, Kirkland is also one of the "gods of chromatography."

Hsieh, Duncan, Brisson, Anal. Chem., 79, 5668 (2007).
"Fused-core Silica Column HPLC/MS ..."
A comparison of fused core and 1.7 um particles. Makes the case that they produce comparable performance for a pharma compound. Introduction has a readable section on van Deemter curves.

[Uwe Neue]

I am on vacation for a few days, without access to journals, so let us take this with a grain of salt.

My recollection says that the publication cited by Adam required a pressure at or over 6000 psi for the comparison cited by him. Thes are UPLC pressures and the work was done on a UPLC instrument. I will verify next week.

Hsieh and colleagues are mass spectrometrists. They accomplished very fast (sub-2-minute) separations with a MS system attached. While this is good work, this does not mean that they were able to truly evaluate column performance. Typical MS chromatography is dominated by extra-column bandspreading, and is not suitable for measuring column performance. One needs to take this with a huge grain of salt.

More next week after I (re)read the publications.

[mbicking]

I think we should take all of this with a grain of salt, including all the marketing claims by Mac-Mod, Supelco, Agilent, and Waters (which covers both sides of the argument). All have a vested interest in the outcome, and all are going to present data that promotes their agenda (seems a lot like political ads, doesn't it? ).

Both particle types (sub-2 um and fused core) offer specific advantages, depending on what you are trying to accomplish, and each has issues that must be taken into account. You cannot say one particle is "better" than the other unless you can define "better." And in my opinion, there are several definitions of "better" that could apply to each type.

The original question asked whether anyone had any experience with fused core particles, since they have received a lot of marketing attention lately.

[Wayne Way]

To answer Uwe,

The pressures cited in the Cunliffe article mentioned by Adam are:

All conditions (instrument, etc. the same)

sub 2
Acquity 6531 PSI
Zorbax 5551 PSI
Thermo 5461 PSI

Fused-Core
Ascentis Express 3078
Halo 2806

One of the concluding remarks was:

The Halo C18 and Ascentis Express C18 columns offered a lower backpressure option to sub-2-micron columns with only a slight sacrifice in peak efficiency.

[Uwe Neue]

I promised to get back with some facts, after (re)reading some of the literature references.

As anticipated, the van Deemter curves by Hsieh et al. are off by a significant amount: the HETP values at the van-Deemter minimum are about 2-3 times worse than what I know about 1.7 micron particles. The authors appear to be aware of the problem, since they state: "The use of MS detection might seriously affect the true value of the plate height efficiency of a column due to the relatively large MS inlet volume." Consequence, the paper is not a proof that fused-core particles are as good as sub-2-micron particles.

The paper in J. Sep. Sci. 30 that was cited by Adam gives a value of 20500 plates at a backpressure of 6531 for the Acquity column, and values of 16600 to 17900 plates at backpressures between 2800-3100 psis for the Halo columns, all measured at 0.4 mL/min. I doubt though that these measurement were done at the optimum of the van-Deemter curve for the 1.7 micron ACQUITY column, since in figure 3B, the best performance for the Acquity column was obtained at 3.5 mm/sec, which translates roughly to 0.6 mL/min. Also, the plate count for the ACQUITY column at this spot translates to about 25000 plates, which is where the plate count for the 1.7 micron column is supposed to be.

So, nothing new. As stated above, the Halo particles are pretty good, but they are not a performance match for UPLC and UPLC particles.

[unmgvar]

i 'd like to take this to another angle of aproach, hopefully not firing the matter too much by doing so
i will divide the matter into 2 very broad categories for the sake of discussion.
the theoretical approach is that the smaller the particle size the more plates we have so the better the chromatography is undeniably correct and a physical fact.
on the more practicle side we need to remember:
1. the gain is only the square of the plates-sometimes the added hardware or application demands for that are not worth while.
2. the smaller the particle size the more stringent are the requirements for small extra column volume effects, such that they can ruin any added chromatographic benifits
3. the pressure toll on the instruments is such that we need to take it into consideration.

as shown in the given examples:

The authors appear to be aware of the problem, since they state: "The use of MS detection might seriously affect the true value of the plate height efficiency of a column due to the relatively large MS inlet volume."

this can also mean: it is not worth while for me to invest in a UHPLC system and columns since it will mean that i need to buy a new MS (anywhere from 250-600 K$) in order to really see the benefits.

The paper in J. Sep. Sci. 30 that was cited by Adam gives a value of 20500 plates ... for the Acquity column, and values of 16600 to 17900 plates ... for the Halo columns, all measured at 0.4 mL/min

square of 20500 is about 143.178
square of 17900 is about 133.791
that's 143.178/133.791= 1.07
7% increase in resolution for about 2 times more pressure. sometimes it is worth it sometimes it is not. having the alternative is always good i think.like for regular HPLC. you can work at around 1200 psi or 2500 psi. the instrument will take,you just know that at 1200 psi uptime is greater then at 2500 psi, it's a sad fact of life.

[JA]

This comes from the AMT pittcon poster.



I guess this illustrates the first sentence of the message from Tom Waeghe. The C-term coefficient appears to be similar to that of a fully porous 3 micron Ascentis packing at just above ambient temperature.

Maybe it's wrong to pick too much into one picture, but it also appears that the larger, fully porous particle is competitive with respect to the B-term for longitudinal diffusion. Proportional to the variance contribution from longitudinal diffusion is a parameter, y, which is "a constant dependent on the quality of the packing." Can someone expand upon what that quality is and why it doesn't differ between the illustrated columns?

So it leaves us simply with a comparison in the A-term. I wonder how a 2.5 or 2.7 micron porous particle (if one exists) stacks up against the fused-core material. Would the latter still trump with the size distribution card?

[mbicking]

I will make a few comments while the better theoreticians prepare their responses:

I don't think the B-term should be different, since it is based primarily on the diffusion coefficient of the analyte in the mobile phase. Particle geometry does not enter into this effect.

The A-term does have a particle diameter component, which may explain some of the shift.

The high flow shapes (C-term dominates) do look similar. But remember there are two factors in the C-term: stationary phase mass transfer and stagnant mobile phase effects. The former should be similar for similar bonded phases, but I assume it is the latter term that is smaller for the fused core materials. If so, that would explain why the shape stays the same, but is lower for the fused cores.

Maybe the big guns can shed a little more light on it, but that's the way it looks to me. But remember, the really bid advantages seem to be for the large (really large) molecules, where the C-term becomes really important.

[JA]

But remember, the really bid advantages seem to be for the large (really large) molecules, where the C-term becomes really important.

Strange that there are no comparisons for [really] large molecules in either their Pittcon poster or the brochure offered via the AMT website. Also, this packing material is "only" 90 angstrom.

Another question, is the C-term dependent on the size of the pores? If "no", can I get the same small-molecule performance out of a 300 and 100 angstrom packing?

[mbicking]

I am going to yield to the theory guys on this. But I do know that there are no wide pore Halo materials available yet, so that's why there is no comparison.

[JA]

My statement was somewhat rhetorical

On one hand, 'people' state that the mass-transfer advantage of the fused-core materials is realised for large molecules. The way this is conveyed is that the bigger the molecule, the more significant the benefit. Yet I have seen no comparisons made in the inventors' literature which is where I'd expect to see claims pop up first.

On the other, it is my understanding that large molecules traditionally require wider-pore packings to prevent exclusion. The fused-core material comes only in the regular size of 90 angstrom.

In asking the question re: C-term vs. pore size, I'm curious as to whether performance drops as we open up the pores. If not, why are the 100 angstrom columns so prolific, for example, maybe they simply more mechanically stable.

[Tom Waeghe]

But remember, the really bid advantages seem to be for the large (really large) molecules, where the C-term becomes really important.

Strange that there are no comparisons for [really] large molecules in either their Pittcon poster or the brochure offered via the AMT website. Also, this packing material is "only" 90 angstrom.

Another question, is the C-term dependent on the size of the pores? If "no", can I get the same small-molecule performance out of a 300 and 100 angstrom packing?

I would say that you cannot get the same performance from a 100 Angstrom pore size material as you can from a 300 Angstrom material, except for molecules that are small enough to access both pore sizes.

Publications in the open literature have shown that the advantage for fused-core particles at higher velocities (C term) is really only significant for molecules above about 1000 Da (Journal of Chromatographic Science, Vol. 46, March 2008). The advantages gained for those sizes of molecules and larger ones end at the point where the 90 Angstrom pores cannot be fully accessed. I'm not sure at which MW range that occurs, but would guess that it is in the 4000-7000 Da range (Insulin with MW ~5700 Da shows a significant advantage as described by Gritti et al., J. Chromatogr. A, 1157: 289–303 [2007]). The advantages of the 2.7 um HALO particles versus totally porous particles of the same size are the former’s higher bulk density and much narrower particle size distribution, which allows more uniform flow paths through the column bed. As a result, less eddy diffusion occurs, (much smaller A term in van Deemter), and the efficiency gain is analogous to that reported by Desmet for highly ordered pillar arrays recently (Anal. Chem. 2003, vol. 75, 6244-6250; Journal of Chromatography A, Volume 1073, Issues 1-2, 6 May 2005, Pages 43-51). The shorter diffusion distance of 0.5 microns helps to give the fused-core particles mass transfer characteristics which approach those of 1 micron totally porous particles.

[Uwe Neue]

Again, I had wanted to stay out of this, but the poster picture is an excellent example how one can create incorrect impressions by manipulating graphs.

Let us first look at the scale of the Y-axis. It does not start at 0, but at 1.5. Therefore a value at the minimum of 1.6 looks MUCH smaller than the value of 2.5 for the classical particle. If the Y-axis would start at 0, the difference between 1.6 and 2.5 would be much less impressive.

Well, but this is only the start. You see a plot of reduced plate height versus velocity. This is not normal… It should either be a plot of the actual plate height versus the actual velocity, or better of the reduced plate height versus the reduced velocity. Why was this normal plot not chosen? If this would have been done, the ascending branch of the 3.1 micron Ascentis column would have been flatter. One would not have seen any advantage of the mass transfer term of the Halo particle compared to the fully porous particle, at least not for this molecule, which has a molecular weight of 574.

Finally, and this is a bit harder to see, the B-term on the left side of the graph appears to be somewhat larger for the fully porous packing compared to the Halo packing. This is expected, since the retention factor on the Halo particle is smaller than on a fully porous packing (this also answers the question by Merlin). However, this contributes to the location of the minimum of the curve for the fully porous particle: the minimum of the plot is higher for the fully porous particle. Thus part of the difference between both particles is simply due to the choice of the retention factor, and not due to any superiority of the superficially porous particle.