The state-of-the-art column packing technique is very critical in production of reproducible high efficiency packed columns. Such techniques have been known to be variable and not completely released yet. It seems that the major manufacturing companies of HPLC columns keep their own protocols.

The conventional HPLC columns are usually purchased from the manufacturing companies in most of the chromatography related laboratories. However, capillary columns are frequently prepared in many laboratories over the world, and details of the preparation protocols have been well released in the literature. The very high separation efficiency of the resultant columns packed with particles of 2 um or less up to 500,000 plates /m has been reported in several occasions although the actual column length has been mostly limited to 20 cm or less. Now the technique has been further elaborated to make 1 m long packed capillary columns whose actual separation efficiency is ca 500,000 plates /column (Journal of Chromatography A, 1462 (2016) 165–169). The tips of this advanced protocol are use of denser slurry (200 mg/mL), sonication of the column during the packing, and quick application of very high pressure (2070 bar) after packing only 2 cm under an initial pressure of 150 bar. The key concept is quick packing with sonication under a very high pressure.

Concerning the packing techniques of conventional HPLC columns, some tips of column packing were released from an industrial source in 2006 (Journal of Chromatography A, 1126 (2006) 50–57) and summarized as follows.

• The packing particles should be spherical and have a narrow particle size distribution.
• The column loading apparatus needs to be designed to permit suitable flow patterns to maintain the same internal diameter from the bottom outlet of the slurry reservoir into the column inlet.
• The slurry reservoir must be of sufficient volume to permit slurry concentrations in the 7–15% solids range.
• Column tubing must have mirror-finish walls to minimize column wall effects and to avoid the formation of fines during column loading.
• The packing retainer devices (frits, etc.) should be thin and homogeneous.
• Column loading pressure should be as high as possible for forming homogeneous packed beds of high stability.
• The choice of slurry liquid is a critical factor in column loading. For best results, the liquid should be of low viscosity and a good energy-match for the chemical surface of the packing particles.
• For column bed stability, it is important to close a freshly packed column quickly. This is to prevent the column from totally relieving the packed bed compression that is initially located at one end of the column but which quickly distributes throughout the packed bed after closing both ends of the column.
• Retaining some internal packed bed compression created during column filling is important since it helps to compensate for later stresses imparted by column operation at high pressures and elevated temperatures.

There has been a caution against too high a pressure for possible fracture of packing particles, and gradual increase of packing pressure has been recommended in general. There has been no special comment on vibration of the system. The above descriptions are not enough and some of them are misleading to actually pack high quality HPLC columns.

We have been involved in packing columns with newly developed stationary phases. From our experience, it is not helpful in improvement of column packing quality to increase the packing pressure slowly. It is not necessary to limit the packing pressure below 10,000 psi for fear of fracture of particles. Only fast (high throughput flow) packing with severe vibration is required. It is O.K. to apply the initial packing pressure even over 20,000 psi if the stainless steel frit is installed in the bottom union. The reason of application of rather low initial pressure until achievement of the bed length of 2 cm in the capillary packing described above is to protect the fragile silica-sintered bottom frit. The packing pressure is spread over the packed bed length, and the actual pressure on each particle is not so significant. The packing bed of 5 cm length can be instantly formed (within a second), and the whole column bed is fabricated in its initial format in a few seconds. In the initial format, there may be some minute (but anomalous, see below) void spaces most of which can be easily removed by severe vibration. The bed packing quality is determined depending upon the packing speed and vibration strength during the initial 1 min even though more packing time (ca 30 min) with vibration is required to further remove the anomalous voids.
The external porosity in the packed bed with spheres is 0.26 if the bed is packed in the face-centered-cubic close packing format. The actual external porosity is ca 0.38-0.42 for conventional packed columns of good packing quality. It means that there should be some void spaces in the conventional packed columns. The critical factor is the dispersity of the voids. The packing quality is poor if large voids are scattered in the bed. The average size of voids should be small and the voids should be well spread over the whole packed bed to secure good packing quality and consequent high column separation efficiency. It is better if the small voids are interconnected to form some through-flow-channels to increase mass transfer kinetics.
Heavy duty slurry packers with a pressure limit of 20,000 psi or higher are not commercially available at present. Such slurry packers may be home-built with some efforts. The critical point is how to ensure heavy duty fittings for such high pressure. Fortunately, now the UPLC fitting parts are commercially available, and those fittings may be used instead of home-made fittings prepared laboriously otherwise.

Posting of opinions on this article or experiences on column packing is welcome.