Energy Measurements with a Si Surface Barrier Detector and a 5.5-MeV 241 Am α Source October 18, 2017 The goals of this experiment are to become familiar with semiconductor detectors, which are widely used in particle and nuclear physics; to understand their principles of operation; and to learn some things in general about energy loss in the propagation of charged particles through matter. Semiconductor detectors are essentially diodes (junctions of p-type and n-type semiconductors) operating under reverse bias. The reverse voltage increases the depletion region around the p-n junction and only a very small leakage current flows. When a charged particle enters the depletion region, it creates carriers of electricity, free electrons and holes, through ionization. This causes current to flow momentarily until all the free carriers are swept away. This small current can then be detected as a pulse, with the help of a preamplifier and an amplifier, by either an oscilloscope or a multi-channel analyser. On average, one electron-hole pair is created for every 3.6 ev of deposited energy. Only the charge that is freed in the depletion region contributes to the size of the pulse. Therefore, the total integrated charge of the pulse is proportional to the energy deposited by the ionizing radiation in the depletion region. When a charged particle loses all its energy and stops within the depletion region, the integrated pulse is proportional to the total kinetic energy of the particle. This is typically the case only for α particles with energies of up to several MeV. In this lab, an 241 Am source, producing α particles with kinetic energy of 5.5MeV, is used, together with a Si Surface Barrier Detector (SSB). Because α particles lose energy at a very high rate, even in air, the source and the detector are placed inside a vacuum chamber. Semiconductor-based detectors are described in many textbooks. Some good examples are Refs. [1-4]. A more detailed, technical article from the early days of semiconductor-detector development that describes the particular type of detector used in this lab, the silicon, surface-barrier detector, is Ref. [5]. Setting up Before you start, make sure that the vacuum pump is running and the vacuum chamber is evacuated. The vacuum also serves to seal the chamber so that it cannot be opened accidentally. Exposing the detector to light while it is under voltage will damage it. The source is not encapsulated; instead, an amount of 241 Am is deposited through electroplating on a small spot in the center of the disk. 1
Great care should be taken to not touch or scratch the source, to avoid contamination. The detector is also extremely sensitive. It consists of a thin layer of SiO 2 (the p-type semiconductor) on the surface of a Si wafer (the n-type). The silicon is protected from further oxidation by an extremely thin layer of gold (so that particles lose only a very small amount of energy before entering the active detector volume); the gold foil also serves as a conductor to provide the bias voltage to the diode. This foil can easily be damaged and should never be touched: even a fingerprint will affect its electrical properties and the detector may be irretrievably lost. Power is provided by an ORTEC Model 428 DETECTOR BIAS SUPPLY, through an ORTEC 428 PREAMPLIFIER. The OUTPUT of the power supply is connected to the BIAS input of the preamplifier, and the INPUT of the preamplifier is connected to the detector: a single cable provides bias voltage to the detector as well as carries the detector signals. The preamplifier has two outputs, labeled E (for Energy measurement) and T (for Timing). It also has a TEST input where signals from a pulser can be sent for anergy calibration. Power to the preamplifier is provided through a permanently attached cable with an Amphenol (multi-pin) connector that plugs into the matching connector at the back of any ORTEC amplifier model. The E output of the preamplifier is sent to the INPUT, set to POS (positive), of an ORTEC Model 570 AMPLIFIER for further amplification and shaping. The OUTPUT of the amplifier can then be sent to the ORTEC EASY-MCA multichannel analyser, which will produce an energy spectrum on a personal computer running the MAESTRO software, connected through a USB cable. Determination of the bias voltage Obtain several α spectra using the EasyMCA and with the 241 Am α source always at the same position, by varying the HV power supply voltage. The amplifier SHAPING TIME should be set to 10 µs and the gain adjusted to provide a maximum output signal of about 5 V when the source is in place and the bias supply voltage is at around 50 V (close to the operating voltage). On the MCA, the peak of the energy spectrum should not be too close to either end of the range. Starting from 0V, with the ORTEC 428 power supply set to positive voltage output, raise the voltage in steps of 5V. At each setting, run long enough that the 5.486MeV peak is clearly visible and can be determined with the help of the cursor. Leave the amplifier settings unchanged. For each run, note the total current flowing through the circuit so that the detector bias voltage can be calculated after correcting for the voltage drop across the 100MΩ protective resistor in the preamplifier. This is accomplished with a voltmeter connected to the CURRENT MONITOR jacks on the front of the bias power supply. The measured voltage can be converted to current knowing that the two points are at the ends of a 1MΩ resistor. Alternatively, one can simply consider that the voltage drop across the preamplifier, 100MΩ, protective resistor will be 100 the drop across the 1MΩ power-supply resistor. Save each spectrum in a file for further analysis. When the peak is no longer moving with increasing supply voltage, use this value as the operating voltage. Note the maximum pulse amplitude in the oscilloscope. 2
Energy loss of α particles in aluminum Do a long run at the operating bias voltage to allow an attempt to separate the 5.486 and 5.443MeV α peaks by doing a two-peak fit. The peaks are not symmetric, so you will have to experiment with a few other functions until you find one that results in a satisfactory fit. Then turn off the voltage, isolate the vacuum chamber from the vacuum pump, vent and open the chamber, and place a 3mg/cm 2 Al absorber over the source. Close the chamber, evacuate, and then slowly turn on the voltage. Take another long run with the absorber in place. The changed position of the peak will allow you to calculate the energy loss of the α particles in aluminum. Energy calibration The energy scale can be calibrated with the help of a pulser. With the detector at its nominal operating voltage and no source, turn on the pulser, connected to the TEST input of the preamplifier. Turn the dial to 549, corresponding to the 5.486MeV energy of the alpha particles. Observing the output signal on the oscilloscope, adjust the attenuation until the pulse height is similar in size to the maximum pulse height that was produced by the alpha source. Start data acquisition and note the peak position in the spectrum. Then, using a small screwdriver adjust the calibration until the peak appears at the same channel as the peak from the alpha source at the nominal voltage and with no absorber. Now the pulser is calibrated so that a dial reading of 549 corresponds to E = 5.486MeV. Save the plot for later analysis and take several more pulser spectra corresponding to various energies. (For example, a setting of 200 should correspond to the pulse height that would have been produced by alphas depositing 2.000 MeV of energy in the silicon.) The peak position versus energy (actually, pulser dial setting) will give the calibration of the MCA channels. The best calibration will be obtained by doing a linear fit of the points, after confirming that the relationship is indeed linear. Analysis Calibration Do a linear fit of the MCA channel vs. pulser pulse-height setting, which corresponds to gamma energy, using either the FitYK software or the analytic expressions (another method can be substituted as long as it also returns errors on the fit parameters). This can then be inverted and used in the following analysis to convert channel numbers to energy. Depletion depth from energy deposited in silicon The pulse height produced by monoenergetic alphas is proportional to the energy deposited in the active layer of the silicon. The thickness of this active layer (depletion depth) depends on the reverse-bias voltage V b supplied to the detector. Plot the position of the 5.486MeV, 241 Am peak versus the bias voltage, which is the supply voltage corrected for the voltage drop across the 3
protective, 100-MΩ resistor in the preamplifier. The suggested procedure to convert this energy to depletion depth is as follows: 1. Use the Stopping Power and Range Tables from NIST, the National Institute of Standards and Technology, available as an online calculator (ASTAR) at http://physics.nist.gov/physrefdata/star/text/astar.html to calculate the total range of a 5.486-MeV α particle in Si (use the CSDA, or Continuous Slowing Down Approximation values); call this distance D. In general, some of this will be within the active region and some in the silicon beyond. Only energy deposited within the active region of thickness d will result in a signal being produced by the detector. 2. The calibrated energy spectrum for a given bias voltage gives the amount of energy E deposited in the active thickness d. The energy deposited in the inactive region is then E = E 0 E, where E 0 = 5.486MeV is the total α energy. 3. To find how far the particles penetrated into the inactive region, use the ASTAR calculator again to find the range of an α of energy E in silicon; call this distance d. 4. The depth of the active region is then simply d = D d. This is the depletion depth in the detector when the bias voltage is V b. Plot d vs. V b and compare to the theoretical expectation that the depletion depth varies as d = k V b +V 0, where V 0 is the intrinsic voltage in a p n junction ( voltage build-up ) when the bias voltage is zero; for Si, this is in the range of 0.3 0.6V and including it as an additional parameter may improve the quality of the fit. Do a fit using the above theoretical functional form and comment on how well this describes the data. At what bias voltage do you expect to have fully depleted Si? (The total thickness for this detector can be found on the provided detector data sheet.) Hint 1: When the depth of the depletion region increases beyond the total range of the particles the above procedure does not apply (d would be negative). Only use the data points in the range where d appears to be increasing with increasing V b (the energy peak position keeps shifting to the right). Hint 2: Plotting d 2 rather than d vs. V b results in a linear function that is easier to fit (and also to identify visually the region where the data follow the theoretical expectation). Energy loss in aluminum From the obtained peak position with the 241 Am source placed under the thin Al absorber, calculate the energy lost in this layer of Al and the broadening of the peak due to the statistical nature of energy loss; since there is no good functional form to use for fitting those peaks, you may use the FWHM of the peak, determined by placing the cursor at two channels lower and higher than the peak where the count drops to half the maximum value, as a measure of the resolution. Express the resolution both as an absolute value (kev) and as a percentage of the peak energy. 4
Additional questions 1. How much of the energy resolution (peak width) can be attributed to the detector, as opposed to the electronics? Compare with the width of the peaks obtained with the pulser. 2. How does the bias voltage affect the energy resolution? 3. For this type of detector, the theoretical expectation for the depletion depth is 2ε d = (V b +V 0 ). qn d where ε = ε 0 ε Si = 1.054pF/cm, q is the electron charge, and N d is the donor (impurity) concentration in the n-type silicon. Estimate N d. References 1. A.C. Melissinos and J. Napolitano, Experiments in Modern Physics, 2nd ed.; Academic Press, 2003 (Chapters 2.4 and 8.5) 2. W.R. Leo, Techniques for Nuclear and Particle Physics Experiments, A How-to Approach, 2nd ed.; Springer-Verlag, 1994 (Chapter 10) 3. C. Leroy and P.-G. Rancoita, Principles of Radiation Interaction in Matter and Detection, 3rd ed.; World Scientific, 2012 (Chapter 6) 4. G.F. Knoll, Radiation Detection and Measurement, 4th ed.; Wiley, 2010 (Chapter 11) 5. G. Dearnaley and A.B. Whitehead, The Semiconductor Surface Barrier for Nuclear Particle Detection, Nucl. Instr. and Meth. 12 (1961) 205 5