An Advanced Base-Deactivated Capillary Column for Analysis of Volatile Amines, Ammonia, and Alcohols

To analyze basic compounds at nanogram levels using gas chromatography, a basic surface modification is often required to reduce the impact of the acidic fused silica. Additionally, to separate volatile components, retention and efficiency at lower temperatures are required. Basemodified polyethylene glycols have been available for some time, but are not very stable and lose efficiency when used below 60 °C. Siloxanes are more challenging for base modification because the stability of the siloxane polymer should not be compromised. Solutions to this problem exist, but there is still room for improvement. Present phase technologies are not considered to be optimal, and result in short column lifetimes and nonreproducibility in amine response.

Several years ago, a base-deactivation technology incorporating surface deactivation of fused silica was offered in the Rtx-5 amine and Rtx-35 amine columns (Restek Corp., Bellefonte, PA). A similar approach was taken to design a more stable column for volatile amines, in which film thickness was increased and a direct link with the (base) surface deactivation was created. Additionally, the number of crosslinks (bridges) between polymer chains was optimized to ensure that the polymer retains efficiency at temperatures as low as 40 ºC. The higher degree of cross-linking was incorporated to make the column more resistant for amine/water mixtures. Rtx Volatile amines was tested with a series of amine samples and water matrices to demonstrate performance.

Analysis of amines

The analysis of small-chain amines is of great importance in the chemical industry, as well as in pharmaceutical analysis. Volatile amines are building blocks or reactants in the manufacture of many different classes of compounds. The analysis of small-chain amines by capillary gas chromatography is challenging because of the high polarity and the basic nature of the amine compounds. Any activity in the system will immediately impact amine peak symmetry and response.

Amines are generally known to be very difficult to analyze due to their basic character. Because of this, amines will interact with any active site that may be present within the GC system. With decreasing molecular size, the influence of the amine group becomes larger, which results in stronger adsorption characteristics. This effect is most critical with the primary amines. In addition, the amino group introduces a large dipole in the molecule. This dipole is responsible for a strong interaction with silanol groups and siloxane bridges, which often results in nonlinear adsorption effects, appearing as strong tailing peaks in the chromatogram.

There are several GC methods that can be used to perform amine analysis. Packed columns have been widely used with modified and mixed phases. Typical packings include polyethylene glycol (PEG) mixed with potassium hydroxide or sodium hydroxide, which provides excellent peak shapes for amines. Porous polymers with a basic modification can also be used for this separation.

The best way to prevent the interaction of the strong dipole is to derivatize the amine or to deactivate the column in such a way that the interaction is minimized. Derivatization is not a preferred method because it is time consuming and introduces secondary effects such as lower recoveries and change of matrix.

Practical amine analysis: Priming

In amine analysis, systems are often “primed,” which involves several repeated injections of a high-boiling amine compound. The high-boiling amine will cover the active site, resulting in better chromatography. This deactivation is usually only temporary, since the amine will not remain at the active site. Amine priming must be repeated on a regular basis. The advantage of the injection method is that the whole chromatographic system is deactivated, including injection/detection port liners.

Capillary columns for volatile amine analysis: Challenges

Columns for volatile amines must have a high degree of deactivation combined with retention. Additionally, they must be chemically resistant for difficult matrix conditions. Amines are often analyzed together with water and alcohols. Also, ammonia and water have to be quantified.

Figure 1 - Peak shape of short-chain amine on commercial column dedicated for volatile amines. Conditions: see Table 1.

Table 1 - Conditions for small-chain amine testing

Commercially available columns dedicated for volatile amines work well when amines are in pure condition, but as soon as the matrix becomes tough, the chromatography becomes challenging. Figure 1 shows what typically happens when those columns are used for amine–water mixtures. The amine peaks rapidly show poor peak shapes. They split up and elute with a “chair”-type shape. This makes quantification very unreliable and unpredictable, especially in a routine measurement environment.

In order to obtain better stability, a fundamental change was required. The impact of the fused-silica surface as well as the stability of the stationary phase needed to be improved through the use of more robust surface deactivation and stable polymers.

Rtx Volatile amines column

Figure 2 - Impurities in trimethylamine. Conditions: see Table 1.

Figure 3 - Short-chain amines on Rtx Volatile amines. Note symmetrical methanol peak.

Figure 4 - Short-chain amines in water matrix. Overlays of first and fortieth injection. Conditions: see Table 1. Peak identification: 1) methanol, 2) diethylamine, 3) trimethylamine, 4) methylethylamine, 5) dimethylethylamine, 6) diethylamine, 7) methyldiethylamine, 8) triethylamine.

Figure 5 - Monomethylamine (40%) in water, first and fiftieth injection: overlays. Conditions: see Table 1.

A series of columns were prepared using a deactivation technology that was explored when the Rtx-5 and Rtx-35 amine columns were developed. The base deactivation of the surface greatly reduces the interaction of small-chain amines, providing less adsorption and better peak shapes. Additionally, the stationary phase layer was bonded with the surface and was more intensively cross-linked, resulting in better mechanical stability. Remaining reactive silanols in the stationary phase were eliminated; this was done based on results obtained from the development of the Rxi® deactivation technology (Restek).

Table 2 - Abbreviations and retention times

The resulting column was tested with a series of mixtures (results shown in Figures 2–5). Figures 2 and 3 show the analysis of gas samples of amine compounds. These samples are the simplest, and the chromatography looks good. Also, alcohols elute as sharp peaks. Figure 4 shows the same components, but in a water matrix. Usually when water is involved there are different matrix interactions, which affect peak shape. Figure 4 shows an overlay of amines, 200–1000 ppm in water, of the first and the fortieth injections. Peak shapes are almost identical. This was a very promising result. Figure 5 shows the analysis of a 40% solution of monomethylamine in water. This mixture is known to be very difficult and is a perfect indicator of column stability. The figure shows the first and the fiftieth analysis. There is a small impact on peak shape, but it is minimal compared with the former chromatography solutions available. Table 2 provides the typical retention times for a number of volatile amines.

Figure 6 - Impurities in pyridine, diethylamine, triethylamine, and isopropylamine. Conditions: see Table 3.

Table 3 - Conditions for impurity analysis

Figure 7 - Ammonia and water. Column: 60 m × 0.32 mm Rtx Volatile amines; oven: 45 ºC; injection: split, 1:10; carrier: helium, 150 kPa; GC: HP 5890 (Agilent Technologies, Santa Clara, CA); detection: μ-TCD.

Figure 8 - Carbon dioxide and ammonia. Conditions: see Figure 7.

Figure 9 - Water impurity in ethanol. Conditions: see Figure 7, with the exception of oven temperature at 60 ºC.

Rtx Volatile amines performs very well for polar amines, not only in terms of inertness and water tolerance, but also loadability. Figure 6 illustrates several applications for impurity analysis in pure amine compounds. Figures 7–9 show the separation of gases and water using thermal conductivity detection (TCD). Ammonia and water elute as sharp peaks. The column can also be used for measuring trace amounts of water in solvents, replacing Karl Fischer titration.

Considerations for sample introduction of amines

Injection of amines is always critical since these compounds are generally very viscous and the injection device needs to provide sufficient heat capacity to evaporate the sample. The presence of alkaline groups (salts) in the sample should be prevented as much as possible. If the solvent is water, the pH should be as neutral as possible. The combination of water and alkaline will very rapidly destroy the siloxane phase.

Syringe injection

If a syringe is used for sample introduction, it must be flushed frequently with solvent to remove any residual sample. The amines adhere strongly to any surface, whether it is glass, fused silica, or stainless steel, and, as a result, can cause a reaction on the surface of the needle whereby the needle’s plunger becomes blocked. (During injection, the needle becomes very hot, causing polymerization/pyrolysis reactions.) As a result, discrimination and memory effects are very common.

Amines in water

Amines in water can be injected using headspace sampling if the amines are relatively volatile. At low concentrations, the sensitivity is not high enough, but with extraction techniques, the concentration can be increased. Also, a more suitable solvent can be chosen since water is very difficult to inject due to its high boiling point and polarity. As a result, recoveries are low in general and calibration depends strongly on the sample matrix.

Direct injection of aqueous samples onto wide-bore columns is a possibility, but the conditions for injection must be well controlled, and frequent maintenance is necessary to avoid adsorption in the injection port.

Increasing heat capacity

There are several ways to increase the heat capacity of the injection system. The application of inserts with special geometry has become popular. However, the more parts there are in the insert, the higher the risk of adsorption of amines, and deactivation becomes more critical.

Another solution is to fill the insert with a base-deactivated packing material. This material offers the high heat capacity needed for efficient evaporation. The retention of the packing material can be reduced by using a high injection port temperature. If the column temperature is kept at a low value during injection (40– 50 °C), the injected amines will be focused on the column inlet as a narrow injection band. As a result, the injection should be very reproducible. This works quite well as long as there is no condensation of the solvent (water). If condensation is suspected, a higher column temperature must be chosen (70–90 °C). If retention is too low for some of the components in the sample, a higher-retention column could be used (a thicker film of the same stationary phase).

Summary

A stable bonded stationary phase for amine application based on a nonpolar stabilized polysiloxane has been developed. Because of the surface bonding and base deactivation of the fused-silica surface, Rtx Volatile amines can handle direct liquid injections of water with a high level of short-chain amines such as monomethylamine, dimethylamine, and trimethylamine. The phase shows a very good peak shape for amines, alcohols, ammonia, CO2, and water, and can be used at temperatures as high as 290 °C.

The authors are with Restek Corp., 110 Benner Cir., Bellefonte, PA 16823, U.S.A.; tel.: 814-353-1300; fax: 814-353-1309; e-mail: jaap.dezeeuw@restek.com. The authors would like to give special thanks to Gilbert Baele, Taminco, Antwerp, Belgium, for testing the application and limitation of the base-deactivation technology.

Comments