Challenging Beliefs About How Core-Shell Particles Work

In a new paper to be published in the July issue of LCGC North America, Fabrice Gritti and Georges Guiochon of the Department of Chemistry at the University of Tennessee (Knoxville, Tennessee) present results from their latest studies in their research on the kinetic mechanisms of chromatography columns packed with core–shell particles.

The article you are publishing in the July issue of LCGC North America deals with the performance of columns packed with modern sub-3-µm core–shell particles. How did you become interested in this topic?

Gritti: We were initially motivated to conduct this study because of the exceptional kinetic performance of these new columns with respect to conventional columns packed with fully porous particles. At optimum velocity and for the same average particle size (say 2.7 μm), the literature data showed a significant (30%) increase in column efficiency, a puzzling experimental fact. Two questions naturally surged: Why was that? What new information could we possibly learn from these columns in terms of mass transfer in chromatographic columns?

What stage in your research does this study represent?

Gritti: The article we are publishing in the July issue of LCGC North America addresses the mass transfer mechanism, that is, the quantitative measurement of each individual kinetic event —longitudinal diffusion, eddy dispersion, and solid–liquid mass transfer resistance, better known as the B/u, A, and Cu height equivalent to a theoretical plate (HETP) terms, respectively, in the van Deemter equation — that controls the efficiency of a chromatographic column. The same approach was used for both conventional and core–shell packings and the results were compared. The experimental results were surprising at first glance but unambiguous: First, the B/u HETP term decreased by about 25% due to the presence of the solid impermeable core in core–shell particles. Secondly, the anticipated diminution of the Cu HETP term was unable to explain the high performance of core–shell particles for small molecules. Finally, most of the “magic” behind the exceptionally high efficiency observed stemmed from a 40% diminution of the eddy dispersion HETP term.

What was the most interesting thing you learned while conducting the research for this paper?

Gritti: This study brought up two important thoughts:

Advances in technological sciences can sometimes be unplanned. In this case, the advance just happened and both the column manufacturers and customers are satisfied with it. From a theoretical viewpoint, we learned that the gain in column efficiency for small molecules had nothing to do with the anticipated decrease of the solid–liquid mass transfer resistance HETP term. This argument was erroneously and systematically put forward by the brochures promoting the advantage of columns packed with shell particles over those packed with fully porous particles.

We do not understand much of what is actually going on in modern chromatographic columns (i.d. > 0.5 mm) in terms of mass transfer kinetics. The chromatographic column still remains a “black box.” This is great news because there is plenty of room left for improvement of our understanding of mass transfer phenomena from a theoretical viewpoint. In practice, this could turn out to be huge for the preparation of the next generation of chromatographic columns. The theory behind the kinetic performance of particulate columns (and other types of columns) requires refinement in order to reconcile measurements with expectations. Experimenters and theoreticians could combine their efforts to develop a more accurate description of band broadening along and across packed beds. This could provide new possibilities and new practical solutions for further column improvement within the next decade.

You stated that the investigation led you to challenge beliefs that were both erroneous and widespread throughout the HPLC community. Can you expand on that?

Gritti: One of these misconceptions was already mentioned above: Reducing the average diffusion path across the particles (by using a nonporous silica core) does not necessarily lead to a significant improvement in column efficiency. Take small molecules, for instance: It takes the same negligible amount of time for these analytes to diffuse through either a fully porous or a superficially porous particle. Obviously, this is not true for much larger molecules, such as proteins with molecular weights larger than 10 kDa. For such macromolecules, the diffusion time across the particle increases and can rapidly become the limiting kinetic factor.

A second widespread belief concerns the impact of the particle size distribution (PSD) (or the largest-to-smallest particle diameter ratio) on column efficiency. For some unexplained reason, it is often accepted that narrow PSD leads to more-efficient columns. Since core–shell materials have a much tighter PSD than that of fully porous particles (roughly 5% vs. 20%), this could have easily explained the success of these particles. In fact, this argument is often presented in advertising brochures for columns packed with these core–shell particles. Yet, neither experimental data nor simulation studies have unambiguously supported this explanation.

A third widespread belief concerns the contribution of the eddy dispersion HETP term on the column efficiency. The present study, combined with validated simulation data on mass transport through packed beds, demonstrates that the column performance is primarily dictated by the A term at a high flow rate. Furthermore, it reveals that the eddy dispersion HETP term in chromatographic columns (i.d. > 0.5 mm) is little controlled by the structure of the bulk packing. The confinement of the particles inside a tube (wall effects) and the design of the inlet and outlet frits (or screens) and endfittings appeared as keys for the generation of more-efficient columns. We have yet to develop accurate models to take those important practical issues into proper consideration.

What are the next steps in your research?

Gritti: The next step of this specific research is first to provide accurate and significant data that will illustrate the problem of “trans-column” eddy dispersion in modern HPLC columns. This notion of “trans-column” eddy dispersion is still blurry, so we need to extract the relevant experimental parameters that control this source of band broadening in HPLC. The second step will be to propose practical solutions to the column manufacturers to significantly reduce this inevitable effect in today’s column designs.

What reaction to this article do you expect to receive from the HPLC community?

Gritti: All kinds of reactions from the HPLC community can be expected. In the end, they all will be productive. Positive reactions will encourage us to move forward and plan the next steps of this work. But they can also be a lure because mere compliance rarely serves scientific progresses. “Negative” or more reserved comments are expected and necessary (not to say wanted and even desirable) in any scientific community. Everyone in the field has a different background, experience, and advancement in his or her own career, so we all have different frames of references for our beliefs about the fundamentals and practice of HPLC. These differences will trigger more self-criticism, scientific exchanges, pedagogic efforts, explanations, debates and so forth. This is why we are all doing this type of work.

What is your next big research project?

Gritti: The present work is just the visible part of the iceberg. Our “big” project will consist of elaborating more-accurate models of mass transfer in both monolithic and particulate columns. Measurement of both axial and transverse dispersion coefficients (model parameters) will represent the most challenging experimental part of this project. The long-term goal is to predict trans-column eddy dispersion, simulate practical solutions to minimize its contribution, and test these predictions in the lab in order to approach, as much as possible, the performance of the virtual infinite diameter column (no nefarious effects from the wall and the borders of the column on column efficiency; for example, no confinement effects) with real 0.5–4.6 mm i.d. columns. There is still room for improvement: the reduced plate height of the infinite diameter column is as small as 0.5 for random packed beds, small molecules, and at high speed. On average, today’s 0.5–4.6 mm i.d. columns do not deliver reduced plate heights smaller than 1.4. In the meantime, let us keep in mind that progress in column technology will not happen without a parallel improvement on the instrument front. This is particularly true for small i.d. columns (i.d.

The full article by Gritti and Guiochon will appear in the July 2012 issue of LCGC North America (volume 30, number 7).

Related Information

Justin Cooper and Joel Harris at the University of Utah are using single-molecule fluorescence imaging to study the dynamics of molecular transport within chromatographic stationary phases.

Click here to view the podcast.