Mass Analysis

The choice of the mass spectrometer in Py-MS is dictated by the pyrolysis conditions, the analysis speed required, the mass range and resolution desired, the preferred ionisation mode, the cost of the instrument and the need for computerisation. Fast heating rates generally demand either instruments with simultaneous ion detection capability, i.e. incorporating photoplate or channelplate detectors, or fast scanning instruments such as quadrupoles, time-of-flight spectrometers or voltage-scanned magnetic sector instruments. However, the requirements of high analysis speed and sample throughput may rule out the use of photoplate systems. Also, cost factors may weigh against the use of photoplate instruments. Mass range may become a problem with time-of-flight instruments, especially if coupled with the requirement of unit resolution at higher mass ranges. If ionisation techniques other than EI are required, e.g. FI or CI, only magnetic sector instruments and quadrupoles come into consideration. Finally, ease of computerisation provides a strong argument in favour of quadrupoles.

Most of the Curie-point Py-MS work reported has been performed with quadrupole instruments. The only real disadvantages of quadrupoles for Py-MS studies are the limited ion transmission at higher mass ranges and the inability to perform high resolution studies. If further major breakthroughs were to be made in the development of Fourier transform ion cyclotron resonance mass spectrometry for routine analytical applications (ref. 117), this technique might well become the preferred approach for Py-MS studies. Alternatively, the recent introduction of commercial magnetic sector instruments featuring fast voltage scanning over several mass decades (ref. 118) might also become an attractive alternative to quadrupole systems, especially if highresolution versions of these magnetic sector instruments could be developed.

When using quadrupoles in a Curie-point Py-MS configuration, as shown in Figure 18, typical operating conditions are: electron energy 14 eV (open cross-beam ioniser), ion energy 5-10 eV, mass range m/z 16 (lowest organic molecular ion is CH^) to m/z 200, scanning speed 8-10 spectra/s and total scanning time 10-30 s. Whereas the expansion chamber is heated to 150-200°C, the ion source is not specially heated and the mass spectrometer housing is usually at room temperature. The liquid nitrogen-cooled shroud around the ion source acts as a very efficient pump for organic molecules, which traps most pyrolysis products escaping ionisation. This prevents them from reentering the ion source after multiple collisions with the poorly defined surfaces of the vacuum envelope or quadrupole head. Apart from being a very efficient pump, the liquid nitrogen-cooled screen is also a pump with a highly constant performance, mainly defined by elementary parameters such as temperature and surface area. The constant pumping characteristics of the cold screen are an important factor in the achievement of long-terni reproducibility in the analysis of multicomponent samples. The other solution to this problem, diffusion-limited pumping (ref. 119), does not achieve the same pumping efficiency as the cold screen by far. To prevent the accumulation of excessive amounts of trapped products, the screen is not permanently cooled but is allowed to heat up to room temperature at the end of each day. The molecules released during the heating up period are then pumped off by the main vacuum pump of

Figure 20. Schematic diagram of art automated Curie-point Py-MS system. For a detailed description, see reference 43. Note the 36-position turntable; the liquid nitrogen-cooled screen completely surrounding the ion source and the highspeed ion counting channel. Typical time needed to analyse one batch of 36 samples is 1 hour.

Figure 20. Schematic diagram of art automated Curie-point Py-MS system. For a detailed description, see reference 43. Note the 36-position turntable; the liquid nitrogen-cooled screen completely surrounding the ion source and the highspeed ion counting channel. Typical time needed to analyse one batch of 36 samples is 1 hour.

the system, which may either be a diffusion pump or a turbomolecular pump. Finally,

Was this article helpful?

0 0

Post a comment