What makes your column a good High Temperature GC column?

In the last few years, the request for HT-columns (High temperature GC columns) has increased and a lot of promises have been made from different vendors. In fact, the possibility of running a GC column near or above 400°C would increase the possibilities of the GC technique quite significantly. In the petroleum industry, for example, the amount of so-called "dark crudes", these are crude oils with a high content of high boiling compounds, is increasing. As well, in the food industry and in the synthesis of organic compounds high boiling compounds are an increasing challenge for the GC analyst.

When thinking about HT columns, one first needs to define what does that mean, an HT column? Nearly every column supplier deals with the following figures: temperature range for a wax column: somewhere between 40°C/80°C and 260°C/270°C, and for a polysiloxane column somewhere between -60°C/-40°C and 350°C/360°C.

The lower temperature limit is easily to be explained. Below its "glass transition temperature," the liquid stationary phase cannot be used for gas/liquid distribution chromatography. Because this occurs over a relatively wide temperature range (i.e. the liquid stationary phase becomes solid) and this temperature range is not well defined, chromatographic results below the given lowest temperature limit are usually not reproducible.

But how is the upper temperature limit determined? Every GC column supplier defines the highest tolerable amount of bleed according to their internal specifications. These internal specifications are determined by running a specific temperature program and observing the bleed of the column of interest. The acceptable bleed level depends on several criteria, such as column length and phase thickness. To compare different columns from different manufacturers in regards to temperature stability, therefore, is not easy, because often the internal specifications are unknown. In the end, every user has to compare different columns with his own idea of the highest acceptable bleed level to determine the most suitable column for his application.

The most interesting news out of this is that a column will not be immediately destroyed by running at a higher temperature than the specified limit; only bleed will be higher and column life time will be shorter.

But now we have recognized the first possible restriction of the upper temperature limit: degradation of the stationary liquid phase. Depending on temperature, oxygen content and humidity of the carrier gas, and the presence of acids and bases or other aggressive compounds, separation phases will depolymerize more or less depending on the temperature used. This is chemistry, and can only be minimized by using highly purified carrier gas and a distinct strategy of bonding and crossbonding the stationary phase. Polysiloxanes, for example, will break down after their polymer chain is broken once and terminal silanol groups are formed.

The other restriction is well known by all lab people working with High Temperature GC. The base of every column is the fused silica tubing. During the tubing production process, a thick walled fused silica tube is melted to its softening point and the right dimension is drawn by having two shafts running with different speeds.

Fused Silica is glass like, and if not protected, within a short time it can re-crystallize under influence of oxygen and cations and become brittle. Under stress it will also become brittle, like glass. To avoid this, the fused silica is coated on the outside with a polymer which has to fulfill the following characteristics: it has to be flexible enough to protect the fused silica, it must have a similar temperature expansion coefficient, it must be temperature stable and have a long life time in an air environment.  Polyimide polymers have been found to be a good choice.

Unfortunately polyimide also depolymerizes under high temperature conditions. This is why fused silica columns become dark brown and eventually black when used at high temperatures. The influence of high temperature can be seen by the figures, given by a fused silica supplier on his homepage (http://www.microquartz.de )

Specification (under GC conditions):

–80 h at 400°C –40 h at 420°C –10 h at 450°C

These figures may be read only as an indication, but they show immediately that lifetime of the polyimide coating is highly dependant on the temperature used. At the end, the polyimide will still depolymerize in time when exposed to (even higher) temperatures. And if the coated polyimide is burned, your fused silica column will become brittle.

As a result of this short introduction one perception will last. A capillary column is a consumable supply which can be easily compared with the tires of your car. If used under perfect conditions on a very clean and smooth road, they will last longer than if used under rough conditions like a farm track. Also, if driven with ambient speed, they will last longer than if driven with high speed.

What does that mean regarding to HT columns? If you do not know anything about the highest tolerable amount of bleed, and if you do not know anything about the dependency of the polyimide coating lifetime on temperature effects, you don't know anything about the HT behavior of your GC column.

Some months ago, LECO EUROPE was investigating an interesting new method for a GCxGC-TOF application. This method required that a clear and sharp C60 peak needed to elute within a suitable time. The challenge was to develop this method as a “reversed” GCxGC method with the polar column used in the first dimension.

A Restek Rxi-17 Sil MS column (17m, 0.18 mm ID, 0.18 µm dF) was found to be best to solve the analytical challenge in combination with an Rtx-1 MS column (1.2 m, 0.18 mm ID, 0.18 µm dF) as second dimension, but the temperature program had to be ramped up to 390°C (Table 1).

Table 1: Temperature program Reversed GCxGC Challenge

The surface plot of a reversed GCxGC-TOF experiment shows the good results of this method.

But at least a 13 minute hold time at the 390°C end of the temperature program is needed to elute the C60 peak.  As it may be seen from the cutout of this surface plot, the Rxi-17 Sil MS and the Rtx-1 MS column combination withstands these rough conditions quite well, although Restek states for the Rxi-17Sil MS phase a Tmax of 360°C.

Gaining these results in one shot, this would have been a very academic approach. But it was the intention of the application chemists at LECO to develop a robust ready to use routine method.

To prove the robustness of the method, it was run 400 times with the chosen column combination. These cycles where repeated with 4 different column stets, coming from different batches. As a result of these tests a life time of at least 350 cycles can be observed. This means a life time of 76 hours at 390°C.

From the comparison contour plots (cycle 1 and cycle 372), it can be seen that the limitation of life time comes from phase degradation of the very thin (0.18 µm dF) liquid stationary phase and from the matrix of “real” test samples, the used Fused Silica was working well.

Contour Plot of Cycle 1

Contour Plot of Cycle 372

At cycle 372, high boiling alkanes are discriminated and the middle section starts to tail as a result of some matrix effects and stationary phase loss. The given data show that the Rxi-17 Sil MS column is for sure the most polar temperature stable bonded phase that is commercially available.

Restek Corp. wants to thank Petra Gerhards, European Field Market Development Manager at LECO and Dr. Wibke Peters, Application Chemist Separation Science at LECO, for providing the data and for the good results achieved with the developed method.