Packed columns have higher sample capacity than capillary columns, although the difference has been greatly reduced by the large-bore 530µm capillaries invented by Agilent. Improvements in detector sensitivity have also reduced the need for large samples. The one area in which packed columns may have an advantage is in analysis of gas samples.

For almost all other samples, capillaries provide much better efficiency (narrow peaks) which leads to greatly improved peak separation. In fact, the separating power is so great that many analyses can be done on surprisingly short columns in very brief runs. This time saving translates directly into reduced turnaround time and increased sample throughput.

For new or updated methods, we recommend capillary columns unless there is some overwhelming reason for using packed columns.

It must be as inert as possible, particularly for trace analysis work or for compounds that tend to tail badly, for example active compounds such as mercaptans or alike. For capillaries, fused silica is the material of choice.

There are two basic types of fused silica capillary columns: The Wall Coated Open Tubular or WCOT columns and the Porous Layer Open Tubular or PLOT columns. The stationary phase in WCOT columns is a liquid film coated to the deactivated wall of the column. These are the most widely used columns in gas chromatography. In the PLOT columns the stationary phase is a solid substance that is coated to the column wall.

Packed columns may be glass or metal, usually stainless steel. Metal, although inherently active, is durable and suitable for nonpolar materials. But if samples with polar components are to be analyzed, select glass. If even this is too active (peak tailing, sample loss), try a deactivation treatment.

When selecting capillary columns the first decision to be made is whether a PLOT column is needed. Here are the typical application areas for the three types of PLOT columns:

If none of the above mentioned applications is what you are interested in then you will be able to use a WCOT type column.

When faced with an unknown sample, first try the column that is presently in your GC. If that does not give satisfactory results, consider what you know about the sample. The basic principle is that analytes like to interact with stationary phases of similar chemical nature. This means that the more you know about your sample the easier it is to find the optimum separation phase.

The most important step is to consider the polar character of your analytes:

• Nonpolar Molecules - generally composed of only carbon and hydrogen exhibit no dipole moment. Straight-chained hydrocarbons (n-alkanes) are common examples of nonpolar compounds.

• Polar Molecules - primarily composed of carbon and hydrogen but also contain atoms of nitrogen, oxygen, phosphorus, sulfur, or a halogen. Examples include alcohols, amines, thiols, ketones, nitriles, organo- halides, etc.

• Polarizable Molecules - primarily composed of carbon and hydrogen, but also contain unsaturated bonds. Examples include alkenes, alkynes and aromatic compounds.

Hewlett-Packard offers you the right stationary phase for your specific separtion needs: Is your sample a mixture of non polar components of the same chemical type, such as hydrocarbons in most petroleum fractions? Try a nonpolar column such as HP-1 which separates them in (approximate) boiling point order. Perhaps you suspect some aromatic components; try a column such as HP-5 or HP-35 with phenyl groups.

Samples with polar or polarizable compounds often resolve well on the more polar and/or polarizable stationary phases that contain phenyl groups and alike. Examples are the HP-210 or HP-225 columns. If even more polar phases are required consider the polyethylene glycol (PEG) phases, also often called the wax phases.

Please see the phase selection chart for stationary phase suggestions that are based on application and analyte polarity.

Bonding creates chemical bonds between the phase and the column tubing. Crosslinking polymerizes the phase in place to increase its molecular weight. Both processes are happening simultaneously during the manufacturing process of bonded/crosslinked columns and have the desirable effects of increasing temperature stability and reducing column bleed. Bonded/cross-linked columns can be rinsed to remove contamination that might build up over time and allow larger volume injections. Where there is a choice, we recommend the bonded/crosslinked phases over the standard coated version.

The general rule is that thin films elute components sooner with better peak resolution and at lower temperatures than thick films. This indicates that they are well suited to samples with high-boiling components, closely-spaced components, or temperature sensitive components.

The "standard" film thickness is 0.25 to 0.5 m m. These work well for most samples (including waxes, triglycerides, and steroids) eluting up to 300 ° C. For components eluting at higher temperatures, thin films (0.1 m m) are available.

While standard or thin films are appropriate for high-boiling components, thicker films are needed to resolve low-boiling materials. Film of 1 to 1.5 m m work well for components eluting between 100 and 200 ° C. Extremely thick films (3 to 5 m m) are needed for gases, solvents, and purgeables to increase their interaction with the stationary phase.

Another reason for using a thicker than normal film is to maintain resolution and retention times when changing to a wider bore column. For this reason, wide bore columns tend to be available only with thicker films.

Thick films mean more material in the column and therefore more bleed. Temperature limits must be lowered as film thickness rises.

As a general practice, 15m columns are used for fast screening, simple mixtures, or very high molecular weight compounds. The 30m length has become the most popular one for most analyses. Very long columns (50, 60 and 105m) are for extremely complex samples.

Column length is not a very strong parameter in column performance. For example, doubling column length doubles isothermal analysis time but increases peak resolution by only about 40%. If an analysis is almost but not quite good enough, there are better ways than length to improve it. Consider a thinner film, optimizing the carrier flow through the column, and using temperature programming if you are not already doing so.

One special situation is the analysis of samples with extremely active components. These will tail severely if they contact the column material. Relatively short columns with thick films reduce the chance of interaction by having less column material and smothering it with stationary phase to conceal active sites.

Increased diameter means more stationary phase, even with the same thickness, for greater sample capacity. It also means reduced resolving power and greater bleed.

Narrow columns provide the resolution needed for complex samples, but typically require a split injection because of the low sample capacity. Wider columns avoid this if the loss of resolution can be tolerated. When sample capacity is a major consideration, as with gases, very volatile samples, and purge and trap or headspace sampling, large id or even PLOT columns may be appropriate.

Also consider the limitations and needs of your instrumentation. An adapted packed column inlet can use the larger bore capillary column but not the narrow ones. Inlets designed specifically for capillary columns generally handle the entire id range. GC/MS and MSD with direct coupling may require narrow columns because the vacuum pumps cannot handle the high flows used with larger columns. Look at your entire system to discover which parts limit your choice of column diameter.

Need help from friendly and experience chromatographers?
Contact Agilent Technical support