(Image from Kromek)
Cadmium zinc telluride CZT is a room temperature semiconductor that directly converts x-ray or gamma photons into electrons and holes.
It is a unique semiconductor compared with silicon and germanium detectors, in that cadmium zinc telluride CZT operates at room temperature and can process >100 million photons / second /mm2. Additionally, CZT’s spectroscopic resolution clearly outperforms any commercially available scintillator. The unique combination of spectroscopy and very high count rate capability at room temperature makes CZT an ideal detector solution for many applications.
The detector we used is the SDP 310/Z/20 built by Ritec. The Spectrometric Detection Probes Model 310 (SDP310) are miniature portable gamma-ray detection devices used for measurement of gamma ray spectra in a range of an energy registered more then 50 keV.
The SDP310 contains a single CZT detector, charge sensitive preamplifier and stainless steel housing of 0.8 cm external diameter connected in a watertight manner to a cable of 0.6 cm in diameter. Small external diameter of the SDP310 allows make measurements in out-of-the-way places. Small probe design also allows the detector to be well shielded and collimated for application in strong radiation fields.
The technical data of the detector are the followings :
- Bias Voltage = 200 V
- Energy Resolution (FWHM) =16.9 KeV (for Cs137 662 KeV)
- Polarity Pulse : Negative
- Detector Volume = 20 mm3
- Detector Type = CdZnTe quasi-hemispherical detector
- Efficiency (for Cs137 662 KeV) = 0.3 %
- Preamplifier integrated, preamplifier supply +12 V, -12 V
- Radiation window : 0,25 mm stainless steel
The image below shows the detector.
Our setup comprises the detector, +12 V, -12 V supply, the HV detector bias voltage supply and the shaper amplifier. The HV bias is produced by the precision regulated power supply HV80A from AiT Instruments. That component is capable to produce a voltage in the range 10 – 80 V, so less than the recommended value of 150 – 200 V. With that bias the resolution will be little worse than expected, anyway we are able to get good results even with this bias voltage.
The shaper amplifier has been already described in the post PMT Pulse Processing (ENG). The image below shows the whole setup (too many wires …).
With our setup we checked the pulses generated by the detector and the pulses coming out of our shaper amplifier. The scope traces show, in blue, the pulse from the detector and, in yellow, the pulse from the shaper amplifier. The detector pulse, with negative polarity, is amplified with the integrated charge preamplifier, it lasts for around 300 – 400 μs with exponential decay, and its amplitude is around 20 mV, The output pulse has a gaussian shape with duration of about 100 μs and an amplitude of around 200 mV. With these features this kind of pulses can be easily acquired with our MCA : MCA Theremino.
Subsequently we improved the setup, using a low-ripple HV generator capable of delivering a voltage of 150 – 200 V, therefore in line with the suggested value. All the connections and the shaping amplifier were inserted inside a metal box, as shown in the figure below:
Inside the box we have placed the sensor connections and the shaping amplifier already described in the post PMT Pulse Processing (ENG)
The sensor has been placed on a wooden support:
With the setup described above we have acquired the gamma spectra of some sample sources. The detector has a rather small CZT crystal inside and this results in a low detection efficiency, especially with increasing gamma radiation energy. For this reason it is advisable to place the sensor directly in contact with the source and carry out the acquisition for a sufficient time to accumulate a high number of pulses.
- 1 μCi Am241
- 0,25 μCi Cs137
- 1 μCi Na22
In addition to these we also acquired the gamma spectrum of an uraninite sample (uranium ore) and the gamma spectrum of a sample of Monazite sand (thorium ore).
The results obtained, despite the somewhat improvised setup, are very good. The energy resolution, in the medium energy range, is between 2% and 3%, therefore in line with expectations. The detector is also sensitive to relatively low energies, in fact it responds well to the 60 KeV of the Americium. The linearity of the sensor is excellent so it is sufficient to calibrate it in correspondence with a single sample energy (for example the 662 KeV of the Cs137), this characteristic can be appreciated in the uranium ore spectrum (in logarithmic scale) in which all energy peaks are placed in the correct position.
The only little problem is the low efficiency which, in the case of weak sources, requires long measuring times.