As already discussed in Section 2.2, a large number of theoretical plates can only be achieved by ensuring short diffusion paths in the stationary phase pores;
hence HPLC tends to favour microparticles. A sample molecule cannot penetrate more than 2.5mm into a 5mm particle. The two phenol separation chromatograms in Fig. 7.3 show how performance is increased by using small particle diameters: a column with 10mm material needs a length of 20 cm (below), whereas one with 3mm material is only 6 cm long (above) to give a total of 7000 theoretical plates each. Analysis is completed in 15 min in the former case and takes only one tenth of this time in the latter. This enormous saving of analysis time is due to the short column length and to the fact that the small particle packing reaches its van Deemter minimum at a higher flow-rate than the larger particle variety.
Fig. 7.2 Precolumns.
——————————
1M. R. Buchmeiser,J. Chromatogr. A,918, 233 (2001).
The particle size distribution2should be as narrow as possible (with a ratio of diameters of the smallest and the largest particles of 1:1.5 or 1:2, as already mentioned). This is because the smallest particles determine the column permeability and the largest particles fix the plate number. Small particles produce a high flow resistance and large particles are responsible for a high degree of band broadening. For this reason, a particle size analysis accompanies many products.
Example
d10ẳ4mm,d50ẳ5mm,d90ẳ7mm. This means that 10% of the material is less than 4 mm 50% of the material is less than 5 mm 90% of the material is less than 7 mm
but it says nothing about how large or how small the particles are at the two extremes. This particle size distribution representation may appear unusual, but is based on practical particle size analysis methods such as sieving, air separation, sedimentation, optical methods and the Coulter counter.
Fig. 7.3 Comparison of performance between stationary phases of different particle diameter. Above: 6 cm4.6 mm i.d. column filled with 3mm ODS- Hypersil, flow-rate 2 ml min1. Below: 20 cm4.6 mm i.d. column filled with 10mm ODS-Hypersil, flow-rate 0.7 ml min1. Mobile phase: acetonitrile–water (1 : 1). Sample: 1ml phenol mixture (Hewlett-Packard).
——————————
2R. E. Majors,LC GC Int.,7, 8 (1994).
The properties of the packing materials may vary from batch to batch of the same product, so a large amount of one batch should be bought for series analysis over a long period of time.
Various types of stationary phases are in use: Porous particles, non-porous particles of small diameter, porous layer beads, perfusive particles, and monolithic materials.
Porous Particles
This is the usual type of HPLC stationary phase. These materials are 3, 3.5, 5 or 10mm in size. As a rule of thumb, their performance, i.e. the plate number per unit length, doubles each time from 10 to 5 and 3mm, whereas the pressure drop increases each time by a factor of four. Their internal structure is fully porous and can best be compared with the appearance of a sponge (however, in contrast to a sponge, the structure is very rigid). Within the pores the mobile phase (and the analytes) does not flow but moves only by diffusion.
Non-porous, Small-diameter Particles
If the stationary phase is non-porous, the mass-transfer component of band broadening, the so-called C term of the van Deemter curve (see Sections 2.2 and 8.6) disappears or becomes very small because diffusion within pores does not occur. As a consequence, the curve becomes very flat to the right of its minimum and fast chromatography is possible without a loss in separation performance. An example is presented in Fig. 1.1 with the separation of eight compounds in 70 seconds. In order to maintain a certain sample capacity (i.e. specific surface) it is necessary to use small particles with a diameter of 1–2mm. The capacity is about 50 times lower than compared to conventional porous phases and pressure drop is high. The retention volume of the peaks is low because the pore volume is missing; the fraction of liquid within the column (the column porosity e) is only ca. 0.4 compared to 0.8 of a column with porous packing. Therefore extra-column volumes and time constants need to be kept low. Chromatography of biopolymers seems to be more advantageous on these stationary phases with regard to denaturation and recovery.
Porous Layer Beads (PLB’s)
These are large particles with a diameter in the 30mm range which allows to pack them dry. They consist of a non-porous core (e.g. from glass) which is covered with a 1–3mm thick layer of a chromatographically active material.
PLB’s are rarely used nowadays but can be found in guard columns or as repair material for deteriorated columns with collapsed packing.
Perfusive Particles
In analogy to the non-porous phases the perfusive packings have also been developed for the fast separation of biopolymers. They consist of highly cross- linked styrene–divinylbenzene with two types of pores: very largethroughpores with a diameter of 600–800 nm and narrowdiffusion poresof 80–150 nm. The active stationary phase (e.g. a reversed phase, ion exchanger or affinity phase) fully covers the external and internal surface of the particles. The throughpores are wide enough to allow the mobile phase to flow through, whereas it stagnates in the diffusion pores. The analytes are rapidly transported in and out by the flowing eluent and diffusion paths in the narrow pores are short, therefore separation is fast and the van Deemter curve is very favourable (Fig. 7.4). It is not necessary (and also not possible) to prepare really small particles: the typical diameter of perfusive particles is 20mm. Their flow resistance is very low.
Monolithic Stationary Phases3(Fig. 7.5)
It is possible to synthesize stationary phases which consist of one single piece of porous material such as organic polymers or silica. With this concept, the chromatographic bed is not a packing of particles but a porous rod which totally
Fig. 7.4 Perfusion chromatography. Left: particle of the stationary phase; right:
fast separation of immunoglobulin G from cell culture. (Reproduced with permission from N. B. Afeyan, S. P. Fulton and F. E. Regnier,J. Chromatogr., 544, 267 (1991).) Conditions: column, 10 cm4.6 mm i.d.; stationary phase, Poros M for hydrophobic interaction, 20mm; mobile phase, gradient from 2 to 0 M ammonium sulphate in water in 5 min, 10 ml min1; UV detector, 280 nm.
——————————
3N. Tanakaet al.,Anal. Chem.,73, 420 A (2001); D. Lubdaet al.,LC GC Eur.,14, 730 (2001) or LC GC North Am.,19, 1186 (2001); H. Zouet al.,J. Chromatogr. A,954, 5 (2002).
fills the cylindrical volume of the column. The diameter of the large pores (where the mobile phase flows through) is e.g. 2mm, the mean diameter of the skeleton structure is e.g. 1.6mm, and the diameter of the mesopores is e.g.
12 nm. Such materials have a porosity of more than 0.8, their separation performance is similar to packed beds, and their van Deemter curve is very favourable.
Fig. 7.5 Monolithic stationary phase. (Reproduced by permission of Merck.) Top: Separation of Gamonil and by-products. Conditions: column, 8.3 cm 7.2 mm i.d.; stationary phase, SilicaROD RP 18 e; mobile phase, water with 20 mM phosphoric acid-acetonitrile, combined solvent and flow gradient, 10–50% acetonitrile and 3–9 ml min1; UV detector, 256 nm. Bottom: Scanning electron micrograph of the stationary phase.