Charged particle detection

The Microball [1] constructed in the 1990s uses an array of CsI(Tl) detectors arranged in the geometry shown in figure 8.1; the detectors are arranged in a near-spherical geometry with the individual detectors inclined towards the target. The geometry diverges at forward angles to introduce more granularity, reflecting the fact that more particles are emitted in the forward direction in most reactions of interest. Scintillation light is recorded with photodiodes (see section 5.3.2). The CsI(Tl) scintillation light signal is split into its fast and slow components using analogue electronics. The ratio of the fast and slow component can be used to distinguish different particle types, e.g. protons, deuterons, alpha particles, etc (see figure 8.2). In the subsequent data analysis, the number and type of emitted charged particles can be counted on an event-by-event basis. Operated in conjunction with an array of high-purity germanium detectors, e.g. Gammasphere (see figure 6.17), this particle count can be used to select different fusion-evaporation channels (see figure 8.3). A key point in the analysis of such data is the occurrence of 'leakthrough', meaning that if, for example, two proton events are selected then this event selection will also likely contain a significant number of events where three or four protons were emitted but the finite efficiency and solid angle coverage of the Microball means that frequently one or more protons were missed.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The UoYtube [2] was constructed for use as a charged-particle veto detector in conjunction with the JUROGAM II germanium detector array at the University of Jyväskylä Finland. The tube consists of 96 CsI(Tl) detectors arranged around six sides of a box with hexagonal cross section (see figure 8.4). The individual detectors comprise a 20 × 20 × 2 mm square tile of CsI(Tl) coupled to an S3590-08 PIN diode from Hamamatsu. Through a digital data acquisition system, pulse shapes can be explored. Figure 8.5 shows the different pulse shapes obtained for protons, alpha particles, gamma rays and piled-up events. Comparing the fast and slow components of the pulses measured, different particle types such as protons and alpha particles can be distinguished (see figure 8.6).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In some applications with fusion-evaporation reactions, it is desirable simply to veto any event where a charged particle is detected, e.g. when the interest is in the study of exotic proton-rich nuclei which are populated through in fusion-evaporation reactions by neutron emission only. In this case, the pulse shape discrimination afforded by CsI(Tl) is unnecessary. CsI(Tl) is also intrinsically rather slow due to the slow decay component of the light emitted, meaning that arrays of CsI(Tl) suffer pile-up at high rates. This suggests the possibility of replacing CsI(Tl) with plastic scintillator since the signals from plastic scintillator (see section 9.1.2) are very fast compared with CsI(Tl) and can therefore support high detector rates. The JYtube (see figure 8.7)—a veto detector used at the University of Jyväskylä in Finland, uses plastic scintillator elements coupled to 6 mm × 6 mm SensL J-series silicon photomultipiers (the properties of plastic scintillator are described in section 9.1.2). The modern technology of SiPMs (see section 5.3.3) is more suitable in this application than photodiodes as the SiPMs have a large gain leading to signals which do not need preamplification and can be directly digitised; this dispenses with the preamplification stage required when using photodiodes.

Zoom In Zoom Out Reset image size

The concept behind a detector telescope arises from considerations of how energy loss $\left(-\frac{{\rm{d}}{E}}{{\rm{d}}{x}}\right)$ and range of light charged particles in silicon differs (see discussion in section 2.2.1). Figure 8.9 shows the strong dependence on Z, where there is a large difference in $-\frac{{\rm{d}}{E}}{{\rm{d}}{x}}$ when comparing, for example, protons and alpha particles for a given energy. In addition, there is a smaller dependence on isotope where, for example, protons, deuterons (${}^{2}H$) and tritons (³H) exhibit differing energy loss curves. These differences in energy loss are naturally reflected in differing ranges for charged particles in silicon; figure 8.10 presents the range in silicon for light ions as a function of their incident energy. It can be seen that a 1 mm thick silicon detector will fully stop alpha particles up to 30 MeV but a 10 MeV proton will 'punch through' meaning that it will exit the detector having not deposited all its energy. This differing range for particles might be seen as an inconvenience but it can be exploited to create a detector telescope which identifies the particle species. A typical telescope arrangement for charged-particle identification comprises two layers of silicon detectors: one thin (∼100 μm) and one thick (∼1 mm), where alpha particles will be fully stopped in the front layer and protons will lose energy in the first layer but deposit most of their energy in the second layer. While the total energy deposited in the two detectors may be the same, discrimination between light charged particle types can be readily obtained by comparing the energy deposited in the front and back detectors (indeed, often, as in the example just discussed, it may be trivial as a particle fully stops in the first layer). Examples of silicon detector arrays that employ the telescope approach are presented in section 8.2.4.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Pulse shapes from silicon detectors are found to vary as a function of the energy and species ($Z,A$) of incident ions. This relates to differences in the depth of penetration, i.e. particle range and the ionisation density profile, i.e. $\frac{{\rm{d}}{E}}{{\rm{d}}{x}}$ along the particle track. The effect of this is subtle in that the ionisation density may be high enough to create plasma conditions in the semiconductor which creates a time delay in the electron–hole separation due to a shielding from the bulk electric field in the semiconductor [9]. Moreover, the electric field in the semiconductor may be distorted by the highly ionised track which has a high conductivity with respect to the bulk material. Various strategies can be applied in terms of the pulse shape analysis. For example, the rise time of the pulse from the charged-sensitive preamplifier can be used to determine the particle type for light ions—note the parallel with the employment of rise time in germanium detectors (see section 6.5.2) for obtaining depth and radius of the point of interaction—but, here, the information obtained is different. Figure 8.11 provides an example of such an analysis, plotting the rise time against the energy of particles to discriminate isotopes up to A = 12 [10]. In this example, rise time is defined as the time for the signal to rise from 20% to 80% of the full height. A parallel approach is to investigate the maximum of the current signal from the detector, achieved by differentiating the output of the charge sensitive preamplifier prior to digitisation. Figure 8.12 shows an implementation of this approach with a test experiment using the ISO-FAZIA system (see reference [11] for full details). Separation of charged particles by Z is achieved up to Z = 36 and isotopic separation is also possible for particles with $Z\lt 20$.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A position sensitive detector (PSD) is an older detector design intended to provide information on the position of interaction [12–14]. The operating principle behind some different designs of PSD is shown in figure 8.13. The most common design for a two-dimensional PSD is to have a resistive electrode and read out signals from contacts on the four corners. The position sensitivity is derived from the ratio of the signals on the four corners by means of resistive charge division. The cathode on the reverse can be standard and then be used to read out a signal with good energy resolution. The chief drawback of such detectors is that the position and energy calibration is significantly more complex than with double-sided silicon strip detectors, which are discussed in section 8.2.4 below; this is because there is no intrinsic pixellation. The standard technique is to apply a mask with small regularly spaced holes in it in order to characterise the position sensitivity using a source. Position distortion occurs due to the ballistic deficit effect [15]; reconstruction of the distorted signal is discussed in [16].

Zoom In Zoom Out Reset image size

Figure 8.14 shows an example of a linear array of position-sensitive silicon detectors used as part of the HELIOS spectrometer. The array is used to detect light charged particles following single-particle transfer reactions within the bore of a solenoidal magnet. The particles follow helical trajectories before striking the linear silicon array; this application was discussed in section 3.4.6.

Zoom In Zoom Out Reset image size

DSSD detectors are commonly used for implantation at the focal plane of a recoil spectrometer; such a spectrometer separates fusion residues from scattered beam using a combination of electric and magnetic fields; this application was briefly introduced in section 3.3.1. The benefit of a silicon detector in this application is that it can record both the implantation of the recoiling nucleus and its subsequent decay, if the decay is via emission of a particle with a short range, e.g. an alpha particle. The intrinsic segmentation of the DSSD into effective pixels allows implantation and subsequent decay to be correlated so long as the decay process is fast relative to the implantation rate per pixel. This technique is known as recoil-decay tagging (RDT). Typically in RDT studies, the heavy ion implanted has a relatively low energy. Given that the energy loss for such a heavy ion is high in silicon compared to light ions such as protons and alpha particles, the heavy ion will implant close to the front surface of the silicon detector. This short range is much less than the thickness of any practical silicon detector. The thickness of the detector used can vary and this is dictated by the characteristics of the subsequent decay of interest:

• Where the decay concerns emission of an alpha particle or proton then a thin detector will be optimal as the thinner the detector, the better the achievable energy resolution. Typical thicknesses employed are in the range 150–700 μm. In some cases involving heavy nuclei, the alpha emission will reach excited states in the daughter nucleus which then decay by emission of conversion electrons. Since both the alpha particle and electrons are emitted from the same implanted atom within the silicon detector, the energy deposited by the alpha particle and the conversion electron may effectively sum which is a feature that needs to be taken account of in the analysis [20].
• Where the decay concerns emission of a beta particle, energy resolution is of no importance because beta particles follow a broad distribution of energies—the Fermi–Kurie distribution (as described in section 1.4). Here, the identification of the beta particle emission is important as a means of correlating events. However, beta particles deposit much less energy in a given range of silicon than alpha particles. In this case, DSSDs with thicknesses up to 1 mm are more suitable so that energy loss by the beta particle is higher in order to identify beta particles above the noise.

An example of an implantation system is the GREAT spectrometer [21] at the University of Jyväskylä which is designed for implantation and decay studies. Two rectangular DSSDs are mounted side by side and cover the focal plane of the RITU recoil separator (see figure 8.16). The photograph also shows a planar germanium detector (see section 6.4.1) which is installed behind the DSSDs and can record low-energy gamma rays or x-rays, as well as beta particles.

Zoom In Zoom Out Reset image size

As an aside, it is worth mentioning innovations using scintillators as replacements to DSSDs (and other applications where silicon detectors are used). Scintillators lack the high energy resolution of silicon detectors but they have a number of advantages such as being radiation hard, and dependent on the scintillator they can have very fast time response allowing fast counting. A good example of a replacement scintillator is YAP:Ce [22, 23] which is a material readily produced in thin sheets—a common application for it is as an electron or x-ray screen. YAP:Ce has a fast decay time of 28 ns. This makes it possible to resolve two successive events within 100 ns of each other. This has initially been applied to a specialist application of searching for two fast successive alpha decays following implantation of a recoiling heavy ion [24]. Figure 8.17 provides a schematic of a novel implantation detector based on YAP:Ce; the system can achieve sub-mm position resolution in x and y which is comparable to that of a silicon-based DSSD [24].

Zoom In Zoom Out Reset image size

Annular detectors are an elegant design for nuclear physics experiments with accelerated beams [25]. With the beam directed down the centre of the hole in the annular detector, the rings will subtend the same angle with respect to the beam direction. An example of such an annular detector is the S3 pattern from Micron Semiconductor Ltd which is shown in figure 8.18. The S3 detector is electrically segmented into 24 rings on the junctions side and 32 sectors (wedges) on the ohmic side. Data is usually collected independently for the two sides of the detector and can be reassociated in the offline analysis to determine more precisely the point of interaction from requiring that the energy for an event recorded on the two sides matches. An exception to this is in the case of charge sharing where the interaction took place at the point of the electrical segmentation and charge is recorded in the two adjoining annular rings or wedges.

Zoom In Zoom Out Reset image size

Silicon detector arrays are ideal for nuclear reaction studies such as single-particle transfer reactions (see section 3.4.6) where they are often combined with an array of high-purity germanium detectors. This technique is increasingly being applied to studies with radioactive beams in inverse kinematics, meaning a high-Z beam on a low-Z target, e.g. deuterated plastic target. Reviews of this topic and the relevant methodologies are given in references [26] and [27]. Annular silicon detectors are very suitable for this application based on their geometry which provides good angular resolution. Measurement of particles as a function of angle with respect to the beam direction is critical to this application. Figure 8.19 illustrates the kinematics of typical light-ion transfer reactions in inverse kinematics; most light ion ejectiles are detected at small angles (up to $\sim 60^\circ$) where $\theta =0$ is the beam direction. A large portion of this angular range can be readily covered with an annular silicon detector (or detector telescope) at forward angles. Moreover, such annular detectors have a hole in the middle, allowing the beam to pass through the silicon detector.

Zoom In Zoom Out Reset image size

Annular silicon detectors can be used to identify which excited states are produced in the transfer reaction and the probability of excitation as a function of angle from the measured intensity of light charged particles, e.g. protons, deuterons, alpha particles detected in the various annular rings. In practice, the achievable energy resolution for the charged particles is limited by the properties of the accelerated radioactive beams employed and the parameters on the targets used since these often need to be thick to maximise the number of reactions observed with low intensity radioactive beams. Such arrays of annular silicon detectors are therefore commonly coupled with high-purity germanium detectors (see chapter 6) or other scintillator-based gamma-ray detectors with good intrinsic energy resolution, e.g. LaBr₃(Ce) detectors (see section 5.2.1). The germanium detectors record γ rays emitted from the excited states populated in the reaction. Selecting coincidences between these characteristic gamma rays and the light charged particles can be used to separate the different reaction channels and effectively recover the energy resolution for charged particles. There are several such systems in operation worldwide; some examples include TIARA [28], SHARC [29] and CAKE [30]. Here, we focus in more detail on a specific example, the STELLA array [31].

STELLA [31] is a dedicated station built to measure the ¹²C+¹²C fusion reaction which is one of the key processes governing the evolution of massive stars [32]. STELLA is designed to measure evaporated charged particles, i.e. protons and alpha particles in coincidence with gamma rays from excited states. A 3D CAD drawing of STELLA is presented in figure 8.20. The evaporated charged particles are measured with three annular silicon strip detectors: two of Micron S3 design and one of Micron S1 design (see figure 8.21). One S1 and S3 form a telescope to discriminate protons from alpha particles. STELLA has been coupled to the FATIMA LaBr₃(Ce) array described in section 5.4.2. This has offered yet another means of discriminating between light ions such as protons and alpha particles using the excellent, fast rise time properties of the LaBr₃(Ce) detectors as a time reference. Figure 8.22 shows the energy of light charged particles detected from ¹²C+¹²C fusion as a function of the time difference between the detection of a coincident gamma ray with FATIMA and a charged particle in the annular silicon DSSD. Two loci are observed—one corresponding to protons and one to alpha particles; the separation achieved is because of the differing rise time in the silicon detector of the proton and alpha particle-induced events. It is worth emphasising that it is the sub-ns timing resolution of the LaBr₃(Ce) detectors that makes this possible. Large germanium detectors have too poor a timing resolution to allow this approach to be applied.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Testing of a silicon detector will commonly take place in a vacuum chamber using an alpha source; alpha particles lose energy in air and so a good vacuum is required to minimise such energy loss. A silicon detector is conventionally supplied with details on its depletion voltage and an I–V curve which plots the leakage current as a function of applied bias. It is a common source of misconception and so worth explaining clearly here that the situation with regard to biasing the detector is rather different to that for a germanium detector. With a germanium detector, a recommended bias voltage is provided and operating above this voltage will be risky. Silicon detectors can often be safely operated significantly above the nominal depletion voltage (sometimes two or three times higher than the nominal value) and their performance can continue to increase as charge collection becomes more rapid in the increased electric field. The key is to monitor the leakage current and observe that it is following the expected diode response. A further difference to germanium detectors is that most silicon detectors will operate at some level even with no bias applied due to the 'built-in' bias potential and the thin depletion region setup by it, albeit with very poor energy resolution.

A first step would be to verify the supplied I–V curve while monitoring the output signal on an oscilloscope, increasing the bias voltage in small steps to allow the system to settle. This should verify that the detector is performing according to specification. Energy calibration of a silicon detector is conventionally carried out with a triple-alpha source (see section 3.1.1) to ensure linearity. In the same procedure, the energy resolution can be determined from fitting the peak widths. Given that each of the three main alpha peaks has a small satellite peak below it, being able to resolve these satellites visually in a histogram immediately implies the energy resolution is good. There are some subtleties to energy calibration of silicon detectors that must be borne in mind. Firstly, silicon detectors frequently have a dead layer (SiO₂) and metallisation, e.g. aluminium on their surfaces. An alpha particle from a source loses energy continuously along its track and so will already have lost some energy before it enters the active volume of the silicon detector. This needs to be corrected for otherwise there would be a discrepancy in the suggested energy of alpha particles emitted from a heavy ion implanted into a DSSD, for example. Such energy losses can be calculated and removed from the calibration using, e.g. SRIM (see section 2.2.8). A second important effect is that the pulse height in a silicon detector generated by a charged ion of a given energy differs depending on the ion species [35]. It has been suggested that this relates to the energy required to generate an electron–hole pair in silicon which depends on the charge of the ion (see figure 8.24). The correction is not particularly large—of order 1% between protons and alpha particles. However, such a correction needs to be applied if, for example, a silicon detector is calibrated, as is usual, with an alpha particle source and then latterly used to measure protons.

Zoom In Zoom Out Reset image size

Silicon detectors suffer from radiation damage after being exposed to a high fluence of charged particles, leading to them often being considered a consumable item in experimental nuclear physics. Radiation damage of a silicon detector will lead to decreased pulse height and degradation of energy resolution which continues to the point where the detector is no longer useful for spectroscopic measurements. Damage effects are most severe in a flux of heavy ions [36] and are signalled by a strong (and irreversible) increase in leakage current. There are other ways in which leakage current can increase which are unrelated to damage. For example, if the detector has been insufficiently grounded then it can become charged up in an in-beam experiment by the spray of delta electrons from the target which can appear to greatly increase the leakage current. This leakage current will disappear if the detector is turned off and allowed to recover.

Radiation damage in silicon is associated with the generation of Frenkel defects in the crystal which act as charge-trapping centres leading to electron–hole recombination. For DSSDs, damage to the interstrip region can lower its resistance leading to incomplete charge collection and cross-talk between strips [36]. One of the issues with radiation damage is that it can be highly localised, particularly if the flux of particles is high in a specific angular range, for example. Nevertheless, even if the damage is local, it can spoil the properties of the detector as a whole. There are strategies which can mitigate radiation damage; an example applied to an annular DSSD is discussed in reference [37]. The left-hand side of figure 8.25 shows the impact of radiation damage on an annular DSSD. In each case, the pulse height is plotted for alpha source events in a given sector (or ring) as a function of ring (or sector) on the other side of the DSSD. Since the damage is localised in the inner rings (high ring number), the pulse height is significantly degraded for events registering in the inner rings. The approach taken to remedy this was to consider the DSSD as a large number of pixels based on each ring/sector combination and apply an algorithm to correct the pulse height on a pixel by pixel basis. This largely recovers the resolution of the DSSD as shown on the right-hand side of figure 8.25.

Zoom In Zoom Out Reset image size

Diamond detectors have been used to a minor extent in nuclear physics [38]. They are more widely employed in high-energy particle physics, principally as beam monitors. At very high energies, e.g. for $\gt 0.1\,\mathrm{GeV}$ protons, the radiation hardness of diamond detectors can be ten times higher. However, diamond detectors can suffer radiation damage more easily than silicon detectors at low energies—the more typical domain of nuclear physics [39, 40]. Silicon carbide detectors have also been discussed as a radiation-hard alternative to conventional silicon [41].

In the examples discussed above, we have seen the routes to spectroscopic measurements of alpha particles and heavy ions and how position sensitivity can be used to infer the angle of emission of particles in nuclear reaction studies. Here, we discuss a fascinating example of where it is possible to 'image', albeit indirectly, the tracks of alpha particles (and other heavy ions) in a gas volume. This can be important for study of break-up reactions where, for example, a high energy photon beam interacts with ¹⁶O and breaks it up into ¹²C + ⁴He [42]. The three-dimensional imaging is achieved with an Optical Readout Time Projection Chamber (O-TPC) (see figure 8.26). Within the gas volume, ionisation occurs and electrons drift, through a Frisch grid, into an amplification volume. The resulting electron avalanches and ionisation generate optical photons which can be imaged using CCD cameras leading to a two-dimensional projection. The third dimension is obtained from the projected time-structure recorded by photomultiplier tubes (PMTs). Figure 8.27 shows an example of the image reconstruction of ¹²C dissociated into three alpha particles using the O-TPC.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

An application where fission fragment detection with silicon detectors is appropriate is in the study of beta-delayed fission where low fission event rates are expected and so radiation damage is not of high concern [46]. Figure 8.29 illustrates such an implementation where silicon detectors are used for fission fragment detection. A low-energy radioactive beam of ¹⁸⁰Tl, at the very low rate of 150 atoms s⁻¹, is implanted into thin carbon foils on a 'windmill' system which rotates stepwise. ¹⁸⁰Tl beta decays to ¹⁸⁰Hg. A small fraction of this decay branch goes to highly excited states which spontaneously fission. The fission fragments are detected in two annular silicon detectors mounted either side of the implantation foil. MINIBALL germanium cluster detectors are mounted behind one of the silicon detectors to measure coincident gamma rays. The system allows the measurement of the individual and total fission fragment kinetic energies which affords insight into the mass asymmetry in the fissioning system.

Zoom In Zoom Out Reset image size

Figure 8.30 shows a setup for measuring fission fragments following a multi-nucleon transfer (MNT) reaction where a number of nucleons are transferred onto a heavy system such as ²³⁸U to produce a heavier, excited nucleus which then fissions. Information on the number of nucleons transferred can be measured using an array of silicon ΔE–E telescopes (see section 8.2.2) to identify the beam-like residual nucleus and, hence, the fissioning system. The fission fragments are recorded in four multi-wire proportional counters (MWPCs). MWPCs [47] are a well-established variant of the proportional counter discussed in section 4.3.1. The gas-filled proportional counter contains a crossed array of anode wires. When the fission fragment enters the MWPC volume, it causes ionisation and the charges are collected on the neighbouring wires. This provides information on the position of interaction within the gas volume. MWPCs generate a fast timing signal (∼1 ns) which supports time-of-flight measurements ³ and high rate capability; they are also intrinsically radiation hard.

Zoom In Zoom Out Reset image size

The decommissioning of nuclear sites requires reliable techniques for the detection, identification, and quantification of radioactive substances. Discriminating between disposable and contaminated waste has a significant impact on the cost of decommissioning. In particular, there is no simple, low-cost solution available for assessing the presence of radioactive sources that are essentially alpha-only emitters, such as ²³⁹Pu, on the inner walls of pipework. The range of alpha particles in media is intrinsically short, and as such, alpha-emitters are undetectable from the outside, as any radiation occurring on the inside of the pipe is absorbed by the wall (even gamma radiation from ²³⁵U excited states). The emitting isotope may even be embedded in tiny cracks in the pipe wall, meaning that alpha particles may lose energy as they emerge. Consequently, alpha-particle detectors need to be positioned as closely as possible to the inner walls in order to achieve the greatest detection efficiency and be able to detect very low radiation doses.

A detector system has been developed [51] based on flexible crystalline silicon that can conform to the inner diameter of pipework for the direct detection of alpha-particles. This solution not only offers the potential for 360° inner diameter coverage but can also be adapted to pipes of different diameters. Being as close as possible to the inner pipe walls, maximum sensitivity and uniform efficiency can be achieved, even for low energy alpha-particles, where other approaches may struggle. As a proof-of-principle, it was demonstrated that a sheet of flexible crystalline silicon of 50 μm thickness, as shown in figure 8.34, forms the basis for a high-performance in-pipe alpha-particle detector (see figure 8.35). The sensor can be more generally adapted to any curved surface, such as the exterior of barrels filled with radioactive waste. It can also address the requirements of some nuclear physics experiments that require non-planar detectors.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In section 7.6, we described aspects of gamma-ray astronomy and the strong synergies between the techniques of experimental nuclear physics, in particular the application of high-purity germanium detectors and anti-coincidence shields. Here, we describe an example relating to charged-particle detection in the area of Space Science where the techniques and technology employed also have strong connections with nuclear physics. The example is the Low-Energy Telescope (LET) which is one of four sensors that make up the Solar Energetic Particle (SEP) instrument of the IMPACT investigation for NASA's STEREO mission [52] (figure 8.36). The LET is designed to measure the elemental composition, energy spectra, angular distributions, and arrival times of charged particles from the Sun: protons and helium ions over the energy range ∼1.5–13 MeV/u and heavier ions up to nickel over the energy range from ∼2 to ∼30 MeV/u. This is a very similar energy range to that encountered in low energy nuclear physics although the directionality of the particles is clearly different. The detection and identification of ions with LET is achieved with an array of silicon detector telescopes (see section 8.2.2). The detectors making up the telescopes, L1–L3, have differing thicknesses: L1 detectors are 24 μm thick, L2 are 50 μm thick and L3 are 1000 μm thick. The principle here is the same as in the nuclear physics example: an ion with incident energy, E, will transit one (or more) of the thin detectors and deposit an energy ${\rm{\Delta }}E$ in a thickness, L. In the case where the particle is not normally incident to the detector surface, L can be adjusted to take account of the angle of incidence. The stopping detector records the residual energy, $E^{\prime}$, where $E=E^{\prime} +{\rm{\Delta }}E$. The particle's corresponding atomic number, Z, can be obtained from [52]

$$\begin{eqnarray}&&Z\approx {\left(\frac{k}{L(2+\varepsilon {)}^{a-1}}\right)}^{\frac{1}{(a+1)}}({E}^{a}-E{^{\prime} }^{a}{)}^{\frac{1}{(a+1)}},\end{eqnarray} \tag{ 8.3 }$$

where $2+\varepsilon$ is the mass to charge ratio of the nuclei under consideration and k and a are constants which model the range-energy relationship for heavy ions in silicon (see section 2.2.1). Figure 8.37 shows Monte Carlo simulations of the energy deposited in the different elements of the detector telescopes which shows interesting parallels with some of the nuclear physics examples discussed above. The readout of the LET system illustrates interesting aspects of digital electronics. There are 54 individual silicon elements. These are readout by custom-made Pulse Height Analysis System Integrated Circuits (PHASICs)—see discussion of ASICs in section 10.3. A Minimum Instruction Set Computer (MISC) automatically translates these measured energies into particle identification for ∼12 energy intervals at event rates around 1000 events s⁻¹ [52].

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The first example is a stack of nine DSSDs (Micron Model W1-1000) used at RIKEN-RIBF in Japan, where each DSSD was segmented into 16 strips in the horizontal and vertical axes as a position-sensitive β-counting system (see figure 8.38). Incoming heavy ions are degraded in energy and then implanted into one of the DSSDs in the stack. The subsequent beta decay can be correlated with the implantation by detecting the beta particle in the same DSSD pixel. The beta particle will not be fully stopped in the DSSD but will lose energy as it passes through the DSSD and where that energy exceeds the threshold set on the detector it can be registered as a beta particle. Depending on the implantation depth of the heavy ion within the DSSD, the efficiency for β particle detection varies from 40% to 80% [53].

Zoom In Zoom Out Reset image size

An alternative concept uses a scintillator-based, pixellated implantation-decay detector constructed from the inorganic scintillator, YSO. Figure 8.39 provides a schematic of such a device [54]. YSO is chosen for its high effective atomic number (Z = 35) and density (4.4 g cm⁻³) which can stop beta particles in a relatively short range. YSO is radiation hard and so will not be damaged as rapidly as a more conventional silicon implantation detector. Moreover, it can support high counting rates due to its fast decay time (42 ns). In the example shown, the YSO is pixellated into sticks of $1\times 1\times 5\,\mathrm{mm}$.

Zoom In Zoom Out Reset image size

The third example to be discussed (see figure 8.40) is an innovative design using a rotating highly-pixellated cylindrical array of plastic scintillators ⁴ called CAITEN which has also been used at RIKEN-RIBF in Japan [53]. There are 40 000 pixels of plastic scintillator each with dimensions of $6\times 6\times 20$ mm³. Energetic radioactive ions are implanted into a region of the detector. The detector is rotating at a rate between 0 and 60 rpm to reduce the density of non-correlated decay events in a given pixel; this is the same philosophy as that behind earlier tape-transport systems where radioactive ions are stopped on a tape and transported to detectors where the decay is subsequently observed. The scintillation light is recorded with a cylindrical array of 24 position-sensitive photomultiplier tubes (PSPMTs). Each PSPMT has a segmented light guide mounted in front of it. The light is transported across an air gap of around 3–4 mm between the scintillator barrel and the light guides.

Zoom In Zoom Out Reset image size

Since many experimental studies populate highly excited nuclear states which emit many coincident gamma rays (and conversion electrons in competition with gamma ray emission), there is high interest in detecting gamma ray-conversion electron coincidences particularly for studies of high-Z nuclei where the probability for conversion electron emission is significantly higher (see section 1.7.3). The SAGE spectrometer [60, 61] is an evolution of the earlier SACRED design (see above) and uniquely allows coincident detection of gamma rays and conversion electrons in in-beam studies. Figure 8.44 shows the design of how the electron spectrometer component of SAGE interfaces with a germanium detector array while figure 8.45 shows an expanded view of the electron spectrometer itself. A societal example of electron-gamma ray coincident detection for measuring atmospheric Radioxenon was already provided in section 7.3.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Radioisotopes for positron emission tomography (PET) are produced at cyclotron facilities and subsequently prepared as various molecules in solution such as ¹⁸F-FDG. A standard task in the assay of such materials is to establish their activity and their purity. Measurements of activity are conventionally based on detection of annihilation photons (511 keV) following the ${\beta }^{+}$ decay of the radioisotope by positron emission. This is carried out using a standard gamma-ray detector, e.g. based on a NaI(Tl) scintillator. There is increasing interest in the PET community in 'dose-on-demand' systems where single doses of a specific radiotracer would be produced on site when required. Allied to this is the desire to miniaturise the apparatus involved in handling the radiotracer using microfluidic chips. Recent advances in photodetector technology such as SiPMs offer a route to a compact detector which can directly detect the positrons emitted from the radiotracer [64, 65]. Figure 8.48 shows a prototype detector system based on plastic scintillator (see section 9.1.2) with a thickness of 1 mm and a 6 mm square Sensl C-series SiPM (see section 5.3.3). At this thickness, the background from 511 keV annihilation photons is negligible as the probability for gamma rays to interact with plastic scintillator is very low (the main mode of interaction for plastic scintillator is Compton scattering). This system has been tested by injecting a plug of radiotracer solution sequentially through a conventional NaI(Tl) annihilation gamma-ray detection system and then into the microfluidic chip with its integrated positron detector. Figure 8.49 shows that both detector systems are capable of recording the radiotracer activity.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As discussed in section 3.1.2, ⁹⁰Sr is a common beta-only emitter used as a radioactive source. This radioisotope is commonly found in nuclear decommissioning sites as it is a long-lived isotope (${t}_{\frac{1}{2}}=28.8$ years) produced in nuclear fission [66]. Leaks and spills of material can lead to it being found in groundwater at nuclear installations. Conventionally, in order to assess such contamination, samples of water are obtained, manipulated and processed in a laboratory. It would be highly advantageous to have a direct in situ measurement. Figure 8.50 shows a concept of an in situ detector using a gallium arsenide (GaAs) PIN diode [66]. GaAs has a large bandgap (1.42 eV) meaning that it can operate at room temperature without suffering appreciably from thermal noise; it is also a radiation-hard material.

Zoom In Zoom Out Reset image size