Department of Pure and Applied Chemistry/CPACT

Notes prepared by

Professor David Littlejohn Department of Pure and Applied Chemistry/CPACT University of Strathclyde Cathedral Street Glasgow G1 1XL UK

Revised October 2007 8. Process Gas Chromatography

Gas chromatography is currently the most commonly used chromatographic technique in process control, in both at-line and on-line analysis. The technique is primarily used to monitor the composition of gas streams or volatile low molecular weight samples, and is widely used in e.g. ethylene plants and other petrochemical applications. Unlike laboratory instruments, process gas chromatographs must be designed to operate in harsh environments or hazardous areas, giving continuous operation with better than 97% up-time. This means that the instrument will be on-line providing successive analyses around the clock while requiring less than eleven days a year to calibrate and maintain the unit! Sample conditioning and transportation are important factors to consider when using a process gas chromatograph on- line. Although the basic components of the instrument have the same function as those in laboratory-based gas chromatographs, many differences in design are necessary to adhere to safety requirements.

8.1 Instrument details

The heating mechanism often chosen is air-bath heating, since a plant air source is almost always available. Air-bath heating offers several advantages, including rapid and uniform heating of the components, and is easier to control than other types of heating schemes. Whatever heating mechanism is employed, it must be capable of operation in hazardous environments.

Sample Injection

Typical injection volumes for a process chromatograph are on the order of 200 µl for a vapor sample and less than 2 µl for liquid samples. The majority of process chromatography valves are remotely actuated pneumatic actuators and require between 30 and 60 psig to operate. Stainless steel valves are the most common, but other materials are often used when corrosive samples or environments are present. In those cases, valves constructed of Teflon, hastelloy or monel are used.

Detectors

Although there are a variety of gas chromatograph detectors available for laboratory GCs, in practice, only three are used extensively in process gas chromatographs. The thermal conductivity detector (TCD) is currently the most popular detector used in process GCs, owing to its universal response to compounds and its simple construction and operation. Compounds vary in their ability to conduct heat, with lower molecular weight components generally having much better heat transfer capabilities. For this reason, carriers such as helium and hydrogen are generally used with a TCD. When a

2 component is present in the carrier as it passes through the detector, the thermal conductivity will decrease and a signal is generated.

The flame ionization detector (FID) is often the detector of choice when levels of hydrocarbon are to be measured. The FID utilizes an oxygen-rich hydrogen flame to combust organic solutes as they elute from a column. The current conducted, is roughly proportional to the number of carbon atoms present in the flame. The FID is quite sensitive, allowing measurement of low ppb level of solutes. Only carbon-containing components solicit a response, with carbon oxides (CO, CO2) producing little or no signal.

The flame photometric detector (FPD) is used to measure trace quantities of sulfur and, in very rare instances, phosphorous-containing compounds. The FPD is similar to the FID in that a flame is used to generate the particular species to be monitored. Sulfur-containing compounds are converted to an excited state of S2*, which quickly loses a photon (chemiluminesces). The released energy is collected using a lens and, transmitted via a fiber optic bundle to a photomultiplier tube equipped with a narrow band-pass (394 nm) filter.

Carrier Gas

For process gas chromatography, the particular choice of carrier is the result of consideration of cost, the detector to be used and, in some applications, the effect of the carrier on column efficiency and pressure drop. The carriers typically chosen are helium, hydrogen and nitrogen. Helium is commonly chosen as it is readily available, non-flammable and relatively inexpensive. For thermal conductivity detectors, helium and hydrogen are the most popular due to their high thermal conductivities.

Columns

Modern process gas chromatographs utilize either packed columns or open tubular (capillary) columns, the differences generally being the column internal diameter and the method of suspending the stationary phase. Technological advancements have permitted the construction of capillary columns using metal capillary tubing that has been rendered inert through various processes. These columns are much more durable than their fused silica counterparts and, as such, lend themselves well to the process environment. It is expected that the majority of applications will be performed with capillary columns. Typically, a particular stationary phase is chosen based upon its ability to effect widely differing partition coefficients between the various components to be separated in a sample. The partition of a component between the two phases depends upon interactions that occur between individual component molecules and stationary phase molecules. This interaction can be based on differences in molecular size, polarity or boiling point. The stationary phase will selectively retard components of the mixture based on these differences. Most stationary phases rely on differences in compound polarity to achieve separation. For example, olefins, which contain a polarizable double bond, can be easily separated from saturated

3 hydrocarbons, which tend to exhibit non-polar characteristics. The majority of stationary phases exist in the liquid phase under chromatograph conditions, and sample components become partially dissolved (or partitioned). Other phases, such as molecular sieves, remain a solid and rely on surface adsorption differences for separation. Molecular sieves are often used for the separation of oxygen, argon, nitrogen, hydrogen and methane.

Other factors that can affect the separation of two peaks are column temperature, column pressure, carrier flow rate and amount of liquid phase (or liquid phase film thickness). For column temperature, theory predicts that lower column temperatures favour better separation. In general, this is what is found in practice and one chooses the lowest possible temperature that is appropriate for the application. Most columns, however, have a lower temperature limit that must not be exceeded.

A procedure that is often used is programmed temperature GC (PTGC). Here, the temperature of the column is increased linearly with time during the analysis, after some initial time period at isothermal conditions. Increasing the film thickness results in increased column sample capability. Generally, greater film thicknesses serve to improve the resolution of similarly eluting components, increase analysis times and, in some cases can help reduce the need for sub-ambient cooling in programmed temperature GC applications.

4 8.2 Multidimensional chromatography

Attempts to achieve baseline separation with a single column and simple hardware configuration may not allow the analyst to achieve the analysis or cycle time desired for the application. In addition, sample component concentrations may vary widely in a particular process, creating the necessity of large dynamic range capabilities for the application. These problems, as well as others, have resulted in the development of techniques utilizing a number of columns in various configurations and switching between these columns during analysis to address particular requirements. This technique of column switching is known as multicolumn or multidimensional chromatography. Column switching techniques provide a means to achieve several chromatography functions that are particularly important in process analysis.

 Eliminate undesirable components from the analysis – Rapid analysis cycle times are often very critical for process control. In a process chromatography application, rapid analysis of only the components of interest is of vital importance. Various column-switching techniques are available which allow venting and eliminate measurement of undesirable components.

 Shorten analysis cycle times – Various switching techniques are often used to shorten the analysis time. Rather than wait for all components to elute from a single column, very often, different separating columns are applied to different groups of components in the same analysis.

 Simplify an analysis – Although column-switching techniques generally add to the complexity of the analysis hardware, and as such are not necessarily desirable when they can be avoided, the vast majority of applications in process chromatography use some form of column switching procedure. Many separations are not achievable by a single injection onto a single column. Use of separate columns, optimized for different portions of the analysis, can result in an easier solution to the analysis problem, while providing the desired speed of analysis.

 Column cleaning – Most process streams will contain more components than are desired to measure. Quite often, these components are more strongly adsorbed and, if not removed between each analysis, can accumulate and modify the characteristics of the column. In other cases, these components may elute during a later analysis cycle and interfere with the measurement.

Two of the most widely used column switching techniques are backflush and heartcut. Back flushing is accomplished by reversing the direction of carrier

5 gas flow through the column during the analysis. Any components that were still in the column will be eluted in the opposite direction from the injection. Back flushed components can then be: (i) taken to vent for discarding, (ii) taken to the detector for measurement, (iii) taken to a second column for separation, or (iv) taken to some combination of these three configurations.

Among the many uses of back flushing are ensuring that the columns do not accumulate impurities, reducing analysis time by flushing undesired components to vent, or separating components in a second column, which are not separated on the first column. Very often, back flushing is not used alone, but rather with some other column switching technique.

An example of a use of backflush is illustrated in Figure 34. In the first arrangement, the carrier gas takes the gaseous samples through column 1, where some unwanted components remain on the stationary phase at the start of the column. The analyte components are not retained in column 1 and move with the carrier gas to column 2 where they are separated. Then the flow is switched to the second arrangement, with the direction of flow reversed through column 1 to flush the unwanted components to waste. If the components trapped at the beginning of column 1 were not “unwanted” but were analytes that did not separate on column 1, it is easy to imagine that instead of flushing these components to waste, they could be passed into another column where separation could take place. In all of the arrangements, multiple valves, columns and detectors are required in order to achieve the required separations in a reasonable length of time.

Figure 34 Example of a back-flush system: the upper diagram shows the initial flow path; the lower diagram represents the reversed flow path to backflush compounds from column 1.

The hardware configuration used for the heart-cut column switching technique is depicted in Figure 35. Generally, heart cutting is used to separate

6 a minor component for measurement from a major component, such as a solvent peak. Two columns are used for heart cutting: the heart cut column and the analysis column. The heart cut valve can be used to direct the eluent from the heart cut to either vent or to the second column for separation and quantitation. In the example shown below, the chromatograms on the left are moving in time from right to left. The analyte peak is eluting on the tail of a very much larger interferent compound peak. By switching the effluent flow from the first column to the second at a time just before the analyte peak is eluted, easier separation is achieved in the second column as the ratio of interferent to analyte is reduced from 1000:1 to 10:1. This is illustrated by the sketch on the right of the diagram below, where time on the chromatogram is moving from left to right.

Figure 35 An example of a heart cut system in gas chromatography

8.3 Comments on the attributes of at-line gas chromatography

On-line sampling systems can cost as much as the analyser instrument itself and lack flexibility. At-line instruments have now been developed with a high degree of automation and offer greater flexibility. As the following indicate, at-line deployment of GC is often a useful alternative to on-line applications of the technique (refer to table 9 below).

 Samples are collected manually in sample bottles.  The analyser can be easily located on the plant, usually in the plant control room building.  They are capable of multi-stream analysis over a wide range of concentration ranges.  They use sample trays that allow a number of samples from different points in the process stream to be loaded into the instrument.  Sample bottles from different points are colour coded and so receive a different analysis.  The analysis is performed automatically.

7  These instruments are easy to use and can be operated by manufacturing technicians without the need for experience in analytical measurement.

Table 9 – Comparison of At-line and On-line chromatography.

At-line On-line Sampling Manual Automated Response Time 15-30 minutes Minutes Flexibility High Low Installation Cost Low High Running Cost Moderate Low

By the end of this section, you should be able to:

 Understand the differences in hardware required for plant gas chromatographs as opposed to laboratory instruments.  State the factors that affect choice of carrier gas and column.  Show how complex separations can be achieved through the use of multiple columns.  Give the relative benefits of at-line and on-line gas chromatography.

8.4 Review Questions

1. Describe the possible uses of column switching procedures in process GC.

8 2. Explain the purpose of backflush and heart-cut procedures.

8.5 References

The following references provide additional information on specific topics:

General Background

1. R. Annino Process Gas Chromatography: Fundamentals and Applications Instrument Society of America, Research Triangle, NC, USA, 1992.

2. K. J. Clevett Process Analyser Technology John Wiley and Sons, New York, 1986.

3. R. Sacks, et al. Analyst, 1991, 116, 1313.

4. 4 G. Lee, C. Ray R. Siemers and R. Moore Am. Lab., 1989, 8, 110.

Detectors

9 5. W. C. Askew Anal. Chem., 1972, 44, 633.

6 S. O. Farwell and C. J. Barinaga J. Chromatogr. Sci., 1986, 2