Although GC has been in use for more than three decades, the simple solution of sealing a hot, pressurized injection system with a slice of silicone is still being practiced. Although simple solutions often work well, this application needs to be revisited because one of the biggest troublemakers in GC is the septum. As the need for trace analysis increases, along with the number of problems in the laboratory caused by septa, this has become even more relevant.
Figure 1 - Septa.
The septum
The septum itself (Figure 1) is made of pure polydimethyl siloxane, which is heavily cross-linked. Sometimes stabilizers (phthalates) are added. Different types are available, all with similar goals: easy penetration, minimal contamination, and a long lifetime. The problem is that, at higher temperatures, the septa will start to bleed. These are cyclic degradation products produced as a homologue series starting with D3 (three siloxane units in a circle) and proceeding up to D10–D12. Because the septum is always at high temperature (since it is on top of a hot split/splitless injection system), degradation will be significant. Sometimes when a septum is replaced, oil can be seen coating its inside. These are the D4–D12 degradation products that are released by the septum.
Septum flush
Instrument manufacturers are aware of this problem, and therefore developed the septum flush. By having a low flow of carrier gas along the septum, usually 4–6 mL/min, the majority of septum bleed products will be flushed away (the majority, but not all). When trace analysis is performed, the bleed products will still produce ghost peaks, which can cause interference (Figure 2). Particularly with techniques such as splitless injection, ghost peaks will appear as impurities.
Figure 2 - Septum degradation (seen as ghost peaks).
Following is a simple test that can be performed in the laboratory: Keep the column cold (30 °C) for 15 min, perform a blank run (no injection—only a temperature program up to 300 °C), and observe the baseline. After that run, cool the column to 30 °C again, keep it at 30 °C for 60 min, and follow up with the same program. If ghost peaks appear that increase linearly with the time that the column is kept at 30 °C, there is contamination in the inlet system.
Figure 3 - Accumulation of impurities on the column using different splitless time intervals during injection.
This test can be made more sensitive by simulating different splitless injection times, varying from 0 to 20 min, as shown in Figure 3. During the splitless time (the time when the split flow is normally closed), all impurities that are produced in the injection system are transported to the column and will be trapped on it. Running a programmed analysis will give a clear indication of the cleanliness of the system.
Figure 4 - Typical septum particles from an injection port liner.
Septum particles
In addition to the bleed produced by septum sealing, an even bigger problem is posed by the septum particles that are sometimes deposited in the liner. When a part of the septum is released (for instance, when using an aged septum or a needle with sharp edges), the septum particle will remain in the insert (Figure 4). These particles are in an area where the temperature is 250–300 °C, and will start to degrade and form bleed products. The bleed products, in turn, are directly introduced into the column and will be trapped and detected as bleed or as ghost peaks.
The process will continue until the septum particle is removed or totally decomposed. Very often, the stationary phase in the column is believed to be the bleed originator; however, it should be realized that, in a typical capillary column, only a few milligrams of polydimethyl siloxane (for example, VF-5ms [Varian BV, Middelburg, The Netherlands]) are present. One septum particle coming from a septum can contain almost 50% of the total amount of stationary phase present in the entire column. This illustrates the serious impact of septum bleed from such particles.
Why is septum bleed a problem?
The degradation products formed by septa are similar to the products that are formed by stationary phase decomposition (Figure 5). Many times when bleed is observed, especially during MS with m/z ions 207 and 281, the column is blamed for producing these ions. However, in the majority of cases, the septum bleed elutes from the column. With the advent of the new generation of low-bleed columns such as FactorFour™ (Varian BV), very little bleed is produced. Other sources of bleed will become dominant and will need to be minimized or eliminated.
Figure 5 - Degradation reaction of stationary phase; formation of cyclic siloxanes.
Minimization of septum bleed
There are several precautions that can be taken to reduce the impact of septum degradation/bleed:
- Avoid the septum and use valve injection. This is not always possible but will prevent the septum problem. A valve and sample loop can be mounted before the splitter, and the coupling with the septum cap or through it can be made of clean materials.
- Use a Merlin MicroSeal. This type of device is available from different manufacturers for many GC types. It can work very well, but care needs to be taken with the alignment of the injection port and autoinjector, since small deviations can cause a failure. On the other hand, it eliminates all septum issues.
- Reduce the risk of septum scoring by using tapered needles, which easily penetrate the septum. In combination with prepierced septa (supplied under the name BTO [Varian BV]), septum scoring is minimized. These septa have a hole in the middle, allowing easier access for the needle while maintaining the sealing characteristics.
- Clean the insert/liner and replace the insert in a timely fashion. Glass wool or the whole insert can be replaced; inserts can be cleaned in an ultrasonic bath or thrown away.
- Replace the septum on time; do not wait until it has begun to leak. A leaking septum is not only a source of septum particles, but will also cause discrimination and permit outside air to enter the system. Once air (oxygen and water) enters a GC system, oxidation and hydrolyzation reactions will start, leading to reduced chromatographic performance. This is seen in higher offset (more bleed), tailing, broadened and lower peaks, and even in changed retention times.
A good way to test for septum lifetime is to use a digital helium leak detector and “sniff” the septum for any potential leakage. Using this technique, one can measure how many injections that particular instrument/method can perform before the septum gives up. As a safety margin, take 75% of that value and replace the septa or perform liner maintenance accordingly.
Dr. de Zeeuw is Field Marketing Specialist Consumable Products, Industrial Sciences, Varian BV, Herculesweg 8, Middelburg, The Netherlands; tel.: +31 118 671279, fax: +31 118 67325; e-mail: [email protected].