THE SCINTILLATION DETECTOR AS A SPECTROMETER

We have seen that a scintillation detector will generate a pulse that is proportional to the energy deposited in the crystal by each event recorded. It follows that sorting of the pulses by pulse height will lead to determination of the frequency of occurrence of the pulses as a function of the energy deposited in the crystal. The pulse height analyzer performs this function by establishing a series of "bins", each representing a small increment of the available pulse height (energy) range of the instrument. At the start of a measurement, each bin is empty. As pulses are received, pulse height is checked, and the count stored in the corresponding bin is increased by one. At the end of the count, we have the equivalent of a histogram: Counts recorded per increment of energy range vs. energy.

It is important to recognize that the system does not perform the sorting by source, or by incident gamma energy. Energy deposition in the crystal is the quantity measured. Even under the best of conditions, a monoenergetic gamma source will produce a complex spectrum: superposition of several incident gammas requires an understanding of the processes that take place in the crystal.

For the purpose of discussion an idealized spectrum is shown for a 2 MeV incident gamma flux. We see that there are several areas that require discussion.

Let's look at each of these areas in turn.

Ideally, a single narrow peak would result from measurement of a monoenergetic gamma emitter. Such a spectrum would be obtained only if:

• none of the energy of an incident photon were lost from the crystal through scattering,
• the scintillation intensities produced in the crystal were strictly proportional to the incident energy dissipated in the fluor,
• the number of photons reaching the photocathode were strictly proportional to the number of photons produced in a scintillation,
• the number of electrons emitted from the photocathode were proportional to the number of photons striking it, and
• perfect proportionality were obtained throughout the electronic section of the pulse-height analyzer.

An important aspect of a practical instrument is its resolution. In the case of a scintillation system, resolution is calculated as the ratio of the width of a photopeak (measured at a point equal to one-half the amplitude of the peak) to its energy:

The dispersion of activities about the maximum of a peak is approximately Gaussian.

Compton scattering permits some of the incident energy to leave the crystal with the scattered gamma. It accounts for most of the activity observed at energies below the photopeak. The sharp edge observed is a function of the maximum energy which may be deposited in the crystal by a single-scatter event: it is defined by a 180^o scatter of the gamma prior to its leaving the crystal.

The expression for the energy of the gamma which has been scattered 180^o is derived from the energy-momentum analysis of the Compton effect:

Where:

• E = energy of the incident gamma, MeV
• E' = energy of the scattered gamma, MeV
• theta = gamma scatter angle

Thus, the maximum energy which may be deposited by a single scatter event is E - E' for = 180^o.

The plateau at lower energies is then seen to be due to those scatter events which scatter the gamma at less than 180^o.

The "filling-in" of the region between the Compton edge and the photopeak is due to multiple-scatter events in the crystal.

For gamma-rays over 1.022 MeV, pair production occurs in the crystal, with the probability increasing with the gamma energy. The pair formed interacts locally, with the positron ultimately producing annihilation radiation. If one of the .511 MeV gammas formed manages to escape the crystal, the pulse generated will be .511 MeV less than the original gamma energy. This creates the first escape peak.

If the second gamma also escapes, the pulse forms the second escape peak at 1.022 MeV less than the incident gamma. To find the escape peaks, identify the photopeaks over 1.022 MeV and calculate the [ E_o-.511 and E_o-1.022 ] values and check for a match to lower energy peaks.

A backscatter peak will often appear in a spectrum. It is caused by radiation which passes completely through the crystal without interaction and which is then scattered through 180^o by the shielding or by parts of the detecting equipment, back into the crystal where it is detected. The theoretical location (energy) of the backscatter peak can also be calculated by means of the energy-momentum analysis of the Compton effect, given above.

In this way it is usually possible to identify a backscatter peak from a second photopeak.

When two gammas enter the crystal and deposit their energy simultaneously, the resulting pulse will represent the sum of the energies of the two components.

Sum peaks can be identified by adding combinations of low-energy photopeaks and matching the result to the higher-energy peaks.