Advances in Porous Layer Open Tubular Columns

Porous layer open tubular (PLOT) columns have been very beneficial for solving application problems, especially for the analysis of volatile compounds. The columns’ high selectivity of adsorbents allows analysis time to be optimized, which is important for laboratory efficiency.

Adsorbents based on molecular sieves, alumina, and porous polymers have been available for some time. Although the separation characteristics are well recognized, PLOT columns are traditionally built by the deposition of small particles. Such particles can be kept in place using techniques such as bonding, gluing, or in situ preparation of adsorbent layers. As a result, column reproducibility on retention times, peak elution order, and especially permeability is not comparable with wall-coated open tubular (WCOT) columns.

Novel procedures have been developed to manufacture PLOT columns with different stabilized, concentric adsorption layers. Compared with state-of-the-art-type PLOT columns, new-generation PLOT columns demonstrate constant flow behavior (permeability) and have significantly improved mechanical stability, resulting in easier operation, better chromatography, reproducible retention times, and longer lifetimes.

Spiking

One of the biggest challenges with PLOT columns is that the layer is built by particles. Any change in gas velocity, pressure, surface stress, or vibration can result in a release of particles or even complete segments of the adsorption layer. Such particles are transported by the carrier gas to the detector or via a backflush to the injection/valve system. When a particle reaches the detector, the detector produces a spike. The release of many particles results in serious flow restriction of the column and contamination of the detection system.

One way to prevent particles from coming in contact with the detectors/valves is to use particle traps. Particle traps consist of a 1–2 m section of a polydimethyl siloxane (PDMS)-coated capillary, and a 0.5-µm film “traps” the particles. The particle trap is usually made of the same i.d. capillary. The siloxane coating acts as a glue that immobilizes the adsorption particle. Particle traps can be connected to all ordinary capillary connectors. Some particle traps are supplied already connected to the columns because the particles are not well stabilized/bonded. Even with this type of particle trap, by emitting many particles, column flow restriction can build up. Sometimes this also happens inside the connector. As a result, most commercial PLOT columns suffer from nonreproducible and nonpredictable flow behavior, because restrictions can be formed randomly. This is an important issue when flow switching techniques are used (i.e., Deans/live), where flow reproducibility is critical for setting up time events.

It is preferable to stabilize PLOT columns so that particle traps are no longer required. The new-generation PLOT columns discussed here are the first of an ongoing development to stabilize PLOT columns. By making concentric coatings and in situ bonding of particles, the adsorption layers of molecular sieve, porous polymers, and alumina have been stabilized, and as a result, the columns have become much more reproducible on flow resistance—from column to column and also when the column is used in practical applications, where pressure changes occur via injection or flow switching.

Flow resistance

Figure 1 - Simulation of flow restrictions in a PLOT column and the impact on the flow restriction factor, F.

When preparing PLOT columns, the stationary phase consists of particles that are deposited with layers up to 50 µm in thickness. These thick layers are difficult to deposit as a homogeneous layer. Typically, there are areas in which the layer is thicker or thinner (see Figure 1). As a result, the positions in which the layer is thicker act as a flow restriction for the whole column. Depending on the number and intensity of these flow restrictions, PLOT columns will show more variation on flow resistance than WCOT capillary columns. PLOT columns with the same dimensions can differ in flow by a factor of 4–6 while operated at the same nominal pressure.

For applications in which flow predictability is essential, like Deans switching, reproducible flow behavior is preferred.

Flow resistance factor

In order to have a value for reproducibility of flow resistance, a new factor, the flow resistance factor (F), was introduced. This factor can be calculated according to Eq. (1). The retention time of an unretained component is used as a measurement. For a fixed column dimension and stationary phase coating, the ratio of retention time relative to the uncoated tubing is a measure of the flow resistance. From this can be calculated the percent flow restriction, according to Eq. (2).

The values for F will always be less than 1, since the coated column always has more flow restriction than the uncoated column. For instance, with molecular sieves in a 0.32-mm capillary, the i.d. is reduced by 60 µm, just because of the 30-µm molecular sieve layer. What is important is how reproducible this value can be in a standard production environment.

Measurements must be made at the same temperature using the same test component and a standard carrier gas. For a certain column dimension, the retention time of an unretained component will be a fixed value.