Quality control in spectral analysis

Evaluation of the pump-down curves of a glow discharge source

Chemical analysis using modern Glow Discharge Optical Emission Spectrometry (GD-OES) devices are now commonplace in R&D labs for quality control in as well as industry and manufacturing. The reason for this widely spread application is the short net analysis time of two till five minutes between inserting the sample and receiving the results. One way to optimise this technique is to reduce the pumping time. The Steinbeis Transfer Centre for Vacuum Science and Technology and the Leibniz Institute for Solid State and Materials Research Dresden have turned themselves to this challenge - with interesting results.

In a GD-OES device the sample to be analysed is placed in a sample chamber. This is evacuated and filled with a discharge gas (mostly the noble gas argon) up to a certain pressure (< 10-3 mbar). Then the a glow discharge is ignited in the source by applying a voltage between two electrodes. The noble gas ions formed in the discharge collide with the sample surface gradually stripping it by sputtering. The liberated sample fragments enter to the glow discharge and are excited to light emission. This light penetrates through a lens into the analysis chamber, where it is analysed by a spectrometer. Special sensors (photomultipliers) convert the intensities of the different wavelengths into electric signals, which are amplified and automatically analysed by computer. The wave length of the lines in the light spectrum correspond different elements and their intensities are proportional to the quantity of these elements in the sample. Determining these values it is possible to analyse the chemical composition of the sample. In thin film analysis it is possible to determine the concentration profile in dependence of the depth. The total time required for analysis mainly depends on the time required for pumping the source, whereas the reliability of the spectral analysis mainly depends on the purity of both the source gas and the surface of the source chamber – particularly when analysing thin layers (thickness < 100 nm).

The project was a co-operation between the Leibniz Institute for Solid State and Materials Research Dresden and the Steinbeis Transfer Centre Vacuum Science and Technology in Freiberg. Its aim: to improve a commercially available GD-OES device by reducing the pumping time, increasing the reliability of the analytical results, and setting up a system that updates the operator on device status, indicating the reliability of results and possible faults in the vacuum system. We reduced the pumping time by improving the pumping unit and the process control (optimising the used tube sizes and seals and optimising the pre-treatment in the process chamber).

Evacuating the chamber the pressure drop depends on the leak rate, the surface contamination of the source or the sample, or problems with the vacuum pump reducing the effective pumping speed. Therefore, the pressure vs. time curve during evacuation of the source was used to check the reliability of the analysis. This was done by recording the pressure vs. time curve at the device’s source and simultaneously calculating of the same curve for current source volume and effective pumping speed.

To fit the calculated pressure vs. time curve to the measured curve the effective pumping speed, the leak rate, and the outgassing rate must be varied so that the calculated curve matches the measured one as closely as possible. By this the effective pumping speed, the leak rate, and the outgassing rate during the current experiment can be determined.

These values are immediately received before applying the discharge voltage respectively starting the analysis. They can be used to allow the start of the analysis or to warn the operator against any errors. Because analytical equipment of this kind include always powerful computers for the automatic process control and spectra analysis, these computers can be used too to record and evaluate the pressure vs. time curve.  


Prof. i. R. Dr. rer. nat. habil. Christian Edelmann
Steinbeis Transfer Center Vacuum Physics and Technology (Freiberg)

Dipl.-Ing. D. Klemm
Dr. rer. nat. V. Hoffmann

Leibniz Institute for Solid State and Materials Research Dresden (Dresden)

Share this page