KAHIBARO
37.2. Scintillation
Table of Contents
Light production
In a scintillation material, a charged particle such as an electron, positron, or ion deposits part of its kinetic energy in the medium. At the microscopic level, this energy excites and ionizes atoms or molecules. When these excitations relax back to lower energy states, they can emit photons in the optical or near ultraviolet range. This process is what GATE and Geant4 refer to as scintillation light production.
In GATE, you do not implement the microscopic processes yourself. Instead, you activate optical physics models that already exist in Geant4. When these models are enabled and the material has scintillation properties, every energy deposit in that material can generate a number of optical photons. The optical photons are then transported through the geometry according to their optical properties, such as refractive index and absorption length, which are configured separately in the optical physics and optical properties chapters.
The main idea is that energy deposition acts as the “input” to light production. For each step where a particle deposits energy in a scintillating volume, GATE calculates how many scintillation photons are produced and at what times they are emitted. These photons are new particles of type “optical photon” and they are handled like any other particle in the simulation, but with optical rather than ionizing interactions.
Scintillation light emission is not instantaneous. Real scintillators have characteristic decay times that describe how quickly the light output falls off after the energy has been deposited. Many scintillators are described by a combination of a fast and a slow component, each with its own decay constant. In GATE, this time behavior can be configured through material properties that specify the scintillation time constants and the relative contributions of different components. The simulation then samples emission times from these distributions, so the time structure of the optical photon population resembles the physical light pulse shape.
The wavelength distribution of scintillation photons is also material dependent. A scintillator emits light with a characteristic emission spectrum, not at a single wavelength. This spectrum is represented by a probability distribution as a function of photon energy or wavelength, and GATE uses this distribution to sample the energy of each emitted optical photon. The definition of this emission spectrum and of other wavelength dependent properties belongs to the optical properties configuration and is handled elsewhere in the course.
From the perspective of a scintillation detector, the combination of yield, spectrum, and decay times determines the detector signal. A higher number of photons typically improves energy resolution, because more photons reduce the relative statistical fluctuations. Faster decay times can improve timing performance, because the detector can more precisely measure the arrival time of the light pulse. GATE lets you explore these effects by modifying scintillation parameters and observing how the simulated detector response changes.
To ensure scintillation light is actually produced in a simulation, three elements must be present. First, optical physics processes must be enabled in the physics configuration, especially the scintillation process. Second, the material must be given scintillation properties, which tell GATE that it is a scintillator and specify its yield, time constants, and emission spectrum. Third, particles need to deposit energy in that material. If any of these are missing, the simulation will not generate optical photons even if the geometry contains a scintillator.
The microscopic details of how optical photons propagate, reflect, refract, and are detected by photodetectors are covered in other chapters. In this chapter the focus is on the link between deposited energy and light production, which is summarized by the concept of scintillation yield.
Scintillation yield
Scintillation yield expresses how efficiently a scintillator converts deposited energy into optical photons. It is typically defined as the average number of scintillation photons produced per unit of deposited energy. In many data sheets and textbooks, the yield is given in units of photons per MeV. For example, a common value for LYSO is of the order of tens of thousands of photons per MeV, while plastic scintillators often have lower yields.
In a GATE simulation, the expected number of photons produced by an energy deposit $E_{\text{dep}}$ is computed from the scintillation yield $Y$ by a linear relation. If $Y$ is given in photons per MeV and $E_{\text{dep}}$ in MeV, the mean number of photons $N$ is
$$
N = Y \cdot E_{\text{dep}}.
$$
This relation is implemented at the step level. For each step where a charged particle deposits energy in a scintillating material, the simulation calculates the mean $N$ from the local energy loss and then samples the actual number of produced photons from a stochastic distribution, usually Poissonian. This introduces statistical fluctuations in the light yield, which are essential for reproducing realistic detector energy resolution and pulse height spectra.
A useful rule to remember is:
$$
\boxed{N_{\text{photons}} = Y \cdot E_{\text{dep}}}
$$
where $Y$ is the scintillation yield in photons per MeV and $E_{\text{dep}}$ is the energy deposited in MeV. The simulation will then sample the actual number of photons around this mean, introducing intrinsic statistical fluctuations in the light signal.
In practice, scintillation yield can depend on several factors, such as the type of particle, the energy of the particle, and the local energy loss density. Some materials exhibit quenching effects, which reduce the light yield at high ionization densities. In detailed simulations and experimental modeling, these effects may be described by more complex models, such as Birks’ law, which introduces a non linear relationship between energy deposition and light output. In GATE these more advanced behaviors are controlled by additional material properties and physics options that extend the basic linear yield model.
Another important aspect is that the scintillation yield used in simulation must be consistent with the detector modeling and with the intended application. For example, if you want to reproduce a measured energy spectrum from a gamma detector, you may need to adjust not only the scintillation yield but also detection efficiency, optical transport, and photodetector response. The yield sets the starting point by defining how many photons are produced, but how many of those photons reach the photodetector and create an electronic signal depends on many other factors handled in the optical properties and photodetector chapters.
For beginners, it is useful to think of scintillation yield as a simple proportionality factor. Doubling the yield will, on average, double the number of photons produced for the same deposited energy. This can improve the statistical precision of the signal, but it must remain within physically reasonable values for the chosen material. When exploring simulations, you can change the yield parameter to study how detector performance metrics, such as energy resolution or timing precision, respond to changes in light output. This is a common way to perform sensitivity studies or to understand how much the intrinsic light yield of a scintillator affects overall system performance.
In GATE, scintillation yield is not set directly inside the geometry description, but as part of the material properties. You define a material as scintillating by assigning it a yield and associated timing and spectral properties. Once that is done and the optical scintillation process is active, every energy deposit in that material will automatically produce optical photons according to the specified yield. This separation between material definition and physics processes helps keep your simulation organized and makes it easier to reuse the same scintillator definition in different detector configurations.
Views: 10