M.P. Seah

Channel electron multipliers: quantitative intensity measurement ― efficiency, gain, linearity, and bias effects

Abstract The properties of channel electron multipliers (CEMs) used in measuring electron currents and the factors which affect the recorded output are considered. Previous calibration data are reviewed and found to be inadequate. A general theory is developed which allows the efficiency of the CEM to be predicted as a function of the incident electron energy for the two modes of pulse counting and analog amplification. This description is tested against literature results and is found to be very accurate for both the direct efficiency and the ratio of efficiencies in the two modes. The theory includes predictions for the gain as a function of the applied voltage and this too is found to be very precise. At high count rates non-linearities occur due to pulse pile-up in the counting mode and due to output saturation in both the analog and pulse counting modes. It is recommended that the front of the CEM be fitted with a fine mesh so that a bias of + 200 V may be applied to eliminate the fall in efficiency at low energies and that the output current of the CEM be kept below 5% of the standing wall current to maintain linearity in the analog mode.

Intensity and energy calibration in AES: The effect of analyser modulation

Abstract To a large extent, Auger electron spectra (AES) are recorded and portrayed in the differential mode using the potential-modulation method with a phase-sensitive detector. The precise intensities of the peaks and their energy positions depend on the modulation amplitude in a nonlinear manner. This dependence, in turn, is also a function of the resolution and detailed construction of the analyser. Thus, accurate quantification and accurate assignments of peak energies depend on a detailed knowledge of the effect of analyser modulation in AES. The theoretical response of spectrometers to a Gaussian line is here analysed and compared with measurements for “singlet” peaks using several different commercial spectrometers. This shows that, in certain circumstances, the traditional methods of intensity and energy measurement can be transferred with accuracies as high as 3% and 0.05 eV, respectively. A method for calibrating the effective modulation amplitude, vital for accurate quantification, is presented in the following paper.

Quantitative aes: the problems of the energy dependent phase shift and modulation amplitude and of the non-ideal behaviour of the channel electron multiplier

Abstract In pursuit of the objective to obtain identical Auger electron spectra from different types of instruments, the Varian 10 keV CMA system has been studied. On one particular unit it is shown that the high frequency of 17 kHz, chosen for modulating the spectrometer to produce differential spectra, causes two major errors, each of which affects the relative intensities of peaks in the spectrum. Firstly, it is shown that the phase of the AC signal changes by 56° through the spectrum to 2000 eV, and secondly the amplitude of the AC modulation varies by 100% over the same range. The latter causes low energy peaks to be reduced to half of their true intensity, but the former can cause intensities at either high or low energy to be reduced even more strongly. It is recommended that instruments should be tested and operated, if necessary, at a lower frequency such as 1.7 kHz at which the problems should disappear. A further factor affecting intensities in both the direct and differentiated spectra is the operation of the bias control for the cone of the channel electron multiplier. Field penetration of the cone causes an unwanted behaviour. It is recommended that the cone be fitted with a transparent wire mesh to remove this field penetration and then the bias may be operated at the required potential of at least 200V.

Channel electron multiplier efficiencies: the effect of the pulse height distribution on spectrum shape in auger electron spectroscopy

Abstract Measurements are reported of one factor causing changes in the spectrum shape which occur in AES systems with single channel electron multipliers used in the pulse counting mode. This factor, the voltage across the electron multiplier, is usually set so that the counting rate just reaches its maximum value. The lowest possible voltage is usually chosen to prolong the multiplier life. However, it is shown that the setting chosen in this way depends on the incident electron energy in such a way that a higher voltage is required for lower energy electrons. If too low voltage is chosen, not only is the counting efficiency less than it should be, but the intensity in the low energy part of the spectrum is progressively lost. This arises since the mean output pulse height is a function of both the applied voltage and the incident electron energy. It is concluded that the problem is best avoided by adopting a reference procedure to set the multiplier voltage to just maximise the lowest energy electrons in the spectrum being studied.

Smoothing and the signal-to-noise ratio of peaks in electron spectroscopy

Abstract The smoothing of noisy peaks in electron Spectroscopy by convolutional algorithms is studied further to provide clear guidelines with which the analyst may obtain a rapid optimum result. Following detailed earlier work by Seah, Dench, Gale and Groves we focus on the Savizky and Golay cubic/quadratic, quartic/quintic and Gaussian convolutional functions. It is shown that the most effective smoothing is achieved with a single pass of the quartic/quintic function with a number of points in the smooth equal to 1.7 times the FWHM (N channels) of the peak to be smoothed and that this improves the signal-to-noise ratio by 0.69 N0.5, For the cubic/quadratic the smooth should be of width 1.0 N and the signal-to-noise improvement is 0.66 N0.5. Some points of caution are explained in the meaning of noise reduction, in particular, concerning the use of multiple pass smooths where significant, but hidden, uncertainties still exist in the apparently high quality result.

An atomic standard to calibrate analyser modulation in AES

Abstract For the comparison of differentiated AES spectral data produced using the potential-modulation technique and recorded under different conditions of modulation, or for the transfer of data from laboratory to laboratory, or for the quantification of spectra, the magnitude of the effective energy modulation of the analyser must be known. Even for simple analysers the modulation may differ from the generally accepted value obtained as the product of the measured modulation and the spectrometer constant. A rapid and simple method is therefore presented using the shape of the Ag M 4,5 NN doublet which enables the true analyser energy modulation to be determined to within 3%. The method is tested on a range of instruments and shows that, whilst for some CMAs the generally accepted approach works, for the Varian 10 keV beam CMA, the true modulation deduced from the atomic standard gives a value which is lower by a factor of 1.2 than that assumed previously. The method may also be used to show that the true energy modulation is 1.6 times higher than indicated throughout McGuires Reference Manual.

ESCA microscope: a new approach for imaging in XPS

Abstract This work demonstrates the feasibility of a novel yet simple method of obtaining XPS images using a conventional X-ray photoelectron spectrometer with a minimum of modification. A spherical sector electron energy analyser is operated in the selected area mode (small input aperture, moderate lens magnification) and deflection plates are introduced between the input lens and the sample, allowing the virtual image of the input aperture to be raster-scanned across the sample surface. The image is then formed by using the customary spectrometer output to z -modulate a synchronously raster-scanned video monitor. At present, the resolution is 150 μ but improvements to 10 μ should eventually be possible. Examples are given of studies on electronic components with insulating and conducting areas.

Signal-to-noise ratio assessment and measurement in spectroscopies with particular reference to Auger and X-ray photoelectron spectroscopies

Abstract An analysis is made of some of the problems associated with the measurement of signal-to-noise ratios in electron spectroscopy. We develop here the assessment of the signal-to-noise ratio for counting systems, in particular where the statistics are Poissonian. In order to measure the noise statistics accurately it is shown that at least 135 measures of the counts at a given setting are required. These may help to confirm or disprove the Poissonian nature of the count. At high counting rates the counting system dead time causes deviations from the Poissonian character. However, if the counting system gives Poissonian statistics, simple equations may be used to give the signal-to-noise values from spectral measurements. In systems with more than one detector the allocation of the data from the different detectors to different spectral energy channels may provide some effective smoothing so that the noise appearing in the spectrum may be up to 25% too low compared with a Poissonian estimate. In this case the true signal-to-noise ratio may only be observed from the detector outputs directly, prior to data reduction by the computer. In practical situations the dominance of the Poissonian counting statistics should be exhibited over most of the effective measurement range. However, when searching at the limits of spatial resolution in AES (very low beam currents) or at the highest sensitivity (small signals on a large background) in AES or XPS, other noise sources become important and should be considered.

Versailles project on advanced materials and standards study of intensity stability of cylindrical mirror analyser-based Auger electron spectrometers

Abstract Following an earlier study to define the intensity-energy calibration of Auger electron spectrometers, this study determines the changes in calibration that occur after two further years of use. Five respondees, using cylindrical mirror analysers from the earlier interlaboratory study, have remeasured spectra for the Cu, Ag and Au reference materials, SCAA 87, under standard conditions. The results show that two of the instruments perform as before with no effective change in calibration, two have 10% lower intensities at energies below 100 eV and one has 100% higher intensities at low energy, when the spectra are all normalised at 1000 eV. These changes arise from ageing of the channel electron multiplier (CEM) detectors and the effects of stray magnetic fields. Calculations are provided to show the improvements in stability achieved by biasing the front of the CEM by +250 V.

A verification of the relativistic correction for electro-static electron spectrometers

Abstract The relativistic predictions for the non-linear calibration of a spherical sector electron spectrometer, operated with retardation in the constant δ E / E mode, is verified to an accuracy of ± 0.05 eV over the energy range 0–2500 eV. The verification, involving the comparison of the energy deficit measured between the energy of peaks in the constant δ E mode with that in the constant δ E / E mode, is obtained by the use of an elastically reflected electron beam at 100 eV intervals over the full energy range. The energy deficit, for a VG Scientific ESCA 3 Mk II with a retard ratio of 1.656, increases from zero at 0 eV to 2.28 eV at 2500 eV. It is concluded that the spectrometer is very well behaved and that the established relativity theory is accurately verified in this investigation.