This ability to select or reject pulse amplitudes (or “heights”) by electronic means is referred to as Pulse Height Selection (PHS) and where only one voltage (photon energy) range may be considered at any one time is qualified as “single-channel PHS”. The efficiency of PHS methods is obviously dependant upon the energy resolution of the detector and the proximity of other mean pulse amplitudes to be discriminated (resolved).
The single channel PHS system is based upon “anti-coincidence” circuitry. Such circuitry has two inputs and an output pulse will ONLY be generated when the input signals are NOT coincident. Inputs to the anti-coincidence circuit are the outputs from two “discriminators”. Whilst the input signals to the discriminators are variable in amplitude, their output pulses are a fixed shape and size being suitable for triggering the anti-coincidence circuit. A discriminator output pulse will be obtained when an incoming pulse amplitude is greater than a pre-set minimum level. For the sequence of pulses shown, output pulses at time T1 from D1 and D2 discriminators will both be present since the incoming pulse amplitude exceeds their pre-set input minima. No signal will be transmitted to the counting circuitry because the input pulses to the “anti-coincidence” unit are coincident in time. However, at time T2, the anti-coincidence input from D1 is not coincident with any input from D2 thus an output signal to the counting circuitry is produced. The acceptance range is defined by the discriminator levels where D1 would represent the “Lower Level” and D2 the “Upper Level”.
Each time a pulse is digitised, a “count” is added to the appropriate memory location to which the number is assigned - there will be a small range of numbers for any one address/location. Thus each memory location will store counts from a range, albeit small, of pulse amplitudes.
It is arranged so that the number of allocated memory locations cover the energy range of interest. Each memory channel corresponds to a “channel”.
Qualitative and quantitative analysis routines assume analyte peaks to be in their tabulated energy locations. It is therefore necessary to calibrate the system’s energy scale i.e ensure that the memory locations represent the desired amount of energy (in fact an energy range e.g 10eV etc).
An energy spectrum is simply a frequency distribution (histogram) of the pulse amplitude distributions produced by a suitable detector consequent to the conversion of polychromatic incident X-ray photon wavelengths/energies into voltage pulses. It is similar to the familiar Pulse Height Distribution (PHD) obtained in WDXRF methods.
The energy scale, (Y-axis), consists of discrete energy ranges usually referred to as channels. Such channels are typically 10eV wide (i.e. for Si[Li] detectors) but may be 20eV or more depending on the energy resolution properties of the detector. Each channel registers (i.e. it “counts”) the number of occasions an output pulse amplitude falls within its energy range so that a frequency-energy distribution is created.
The shape of an analyte peak approximates to that of a Gaussian distribution.
An EDXRF spectrum is collected simultaneously. The “energies” of incoming X-ray photons to the detector are evaluated and finally presented, via a Multi-Channel Analyser (MCA), as an energy spectrum. The X-axis represents equivalent photon energy and the Y-axis intensity (channel counts). Depending upon the sophistication of the processing electronics (and software), the energy scale is expressed in photon energy e.g. kilo electron volts (keV) or merely as “channel numbers”. The intensity axis is invariably expressed in “counts”. It is necessary to ensure that the measured peaks occur at their correct locations along the energy scale e.g. CaKa at 3.7keV and FeKa at 6.4keV etc. The amplification and processing electronics (and/or possibly detector bias) are adjusted to achieve “energy calibration”. Spectral resolution is defined by the detector’s “energy resolution”. This EDXRF spectrum was obtained from the same sample as for the previous WDXRF example also using a Rh-anode tube at 30kV.
The data contained in this slide is identical to that of the earlier EDXRF example. Here the scan data points have been plotted on an equivalent photon energy scale instead of angle 2-theta. The peak-width for the EDXRF spectrum (slide #8) looks pretty constant over the whole spectrum as do those of the WDXRF 2-theta scan. For this equivalent WDXRF “energy” spectrum, however, the peak-width rapidly increases with increasing energy. It is observed that above ~10keV, spectral resolutions are reasonably equivalent with WDXRF. In fact for energies above ~20keV, the EDXRF system yields superior spectral resolution! However below ~10keV, compared to that of the equivalent EDXRF data, the spectral resolution for the WDXRF “energy” spectrum rapidly becomes superior with decreasing energy.
Consider the situation illustrated above where it is required to analyse Fe in the presence of Mn.
The ROI for FeKa is asymmetric due to the presence of the overlapping MnKb. Gross intensities for Fe will also include a contribution from the overlapping MnKb. In order to obtain net intensities, the contribution due to the overlapping MnKb must first be subtracted. This is the procedure of peak deconvolution.
我的社区
签到
