CHAPTER I Basic Physical Processes in Radiation Detection 1. Introduction 5 2. Interaction of Heavy Charged Particles 6 2.1 High Velocity Region 2.2 Low Velocity Region 2.3 Medium Velocity Region 2.4 Range-Energy Relations for Heavy Charged Particles - 2.5 Energy and Range Straggling \4. Interaction of Fast Electrons in Matter 15 4, Photon Interactions in Matter 17 4.1 Phidtoeléstiic Effect 4.2 Compton Scattering 4.3 Pair Production 4.4 Linear and Mass Absorption Coefficients of Gamma Rays in Matter 4.5 Distribution of Energy Deposition by Photons in a Finite Medium 5. Mechanism of Charge Production in Detector Media 23 1. INTRODUCTION INTERACTION of nuclear radiation with matter..results-in-a- deposition of . energy in the medium traversed by the radiation. Radiation detection relies on the sensing of this energy deposition through some suitable means. The collection of the free charges released in the medium resulting in a current or a voltage pulse and the detection of the photons emitted during the de- excitation of the excited or ionized atoms and molecules along the path of the incident radiation are two most convenient ways to sense the energy de- position. For a discussion of the fundamental mechanisms involved in the process of radiation detection it is convenient to divide radiations into two broad categories of charged particles and neutral radiations. The energetic charged particles interact primarily with the electrons of the atoms and molecules of a medium via the Coulomb force and lose energy through the excitation and ionization of the atoms of the medium. The heavy charged particles such as protons, alpha particles and the heavy ions deposit their energy in the medium almost cOntiouously through a series of binary colli. ions. But the electrons lose energy in a somewhat discrete manner over a tortuous path. The neutral radiations which are not subject to the Coulomb force, deposit their energy in a medium in a two step process. Energetic photons like x-rays and gamma rays and the neutrons are Scanned with CamScannerthe charged particles and th Covered in Chapter VI. We shalt Specify the zand in the energy loss of only © Photons. Interaction of neutrons will. be Projectile by its mass number m, its atomic number its initial energy £ or velocity »v, ‘The target material will be assumed made of a single material of atomic number Zz, quantity are made in sec. 2.5. 2. INTERACTION OF HEAVY CHARGED PARTICLES Energy loss of heavy charge Scanned with CamScannerBASIC PHYSICAL PROCESSES IN RADIATION DETECTION 7 emergent particle incident z particle M —-| § | —______- N | m,z,e-0€ Ax INTENSITY ENERGY Fig. 1. Schematic diagram of energy loss of a projectile of mass m, atomic number z, and energy E. It loses an energy 4E to the target layer of thickness 4x. The target has atoi number Z, mass number M, mass density e and atomic density N. ionization of the atoms and molecules of the medium traversed by the ions. In this typé of interaction, the maximum energy transfer from low energy liglit ions does not exceed 10 keV; hence ions move in a straight path during slowing down. The energy loss by elastic nuclear collisions is negligibly small in comparison to that from Coulomb interaction except for very low energy ions. Sufficiently energetic charged particles can also interact with the nuclei of the medium by causing nuclear reactions. However, these nuclear reaction processes have'a much smaller cross-section and are rarely of interest in charged particle detection. While discussing the detection of heavy charged particles, we' shall therefore primarily be concerned with their energy loss in _ the detector medium by the processes of excitation and ionization of the atoms and molecules of the medium via Coulomb interaction. ‘ Figure 2 gives a plot of the stopping cross section of protons in silicon covering a very broad range of energies from a fraction of a KeV-to several GeV. An important feature of this curve is the initial increase of the stop- ping cross-section with increasing proton energy up to about 100 keV, anda subsequent decrease beyond this value. Also, in the relativistic region the stopping. cross-section is again seen to slightly increase beyond proton energy oftabout 5 GeV. Variations of the stoppingcross-section with proton Poe a eo that shown in Fig. 2, are also seen for otber target materials. sec! general trend presented above is affected by changing the ee etice ‘in Fig. 3 the stopping cross-section versus projectile n for different projectiles, again for the case of silicon. 10 Scanned with CamScanner§ NUCLEAR RADIATION DETECTORS High @ 10 } velocity velocity region region Relativistic Region 0 10 10 1 «1 WW 10° Proton Energy (MeV) Fig. 2. Stopping cross-section of protons versus proton energy in silicon. ‘The different velocity regions indicated in the ‘fgure reflect that somewhat different theoretical approach- ‘ce are valid for these regions (data taken from (1))- the stopping cross-sections have been normalised by dividing tile atomic number while the abscissa repré- eats the projectile energies divided by the projectile mass number. The data are presented in this form because, as seen later, the stopping cross-section is primarily a function of the velocity » or £/m of the projectile rather than its energy E. Similarly, as will be seen later, in the asymptotic limit of high ‘Em values, the stopping cross-section for different projectiles of the same velocity differ only in being proportional to the square of the projectile Momic number. Some notable features of Fig. 3 are as follows: (i) At high projectile velocities, the different curves of «/z? versus Ejm tend to converge jhile at lower values of E/m there are considerable diffe toa single curve wI rences between different curves which are seen to run nearly parallel to each this figure, them by the square of the project wey € 3 3 = 0 on 10 10° we 1 1 10 10° E/m (Mev/u) Fig. 3. Family of stopping cross-sections for heavy ions in silicon ‘Figure based on semi theoretical calculations of (2, 3D. Scanned with CamScannerBASIC PHYSICAL PROCESSES IN RADIATION DETECTION 9 other. (ii) With increasing projectile atomic number z, the peak value of «/2 shifts to higher velocities. (iii) At very low velocities, a new contribu- tion to the stopping cross-section is seen which increases with the mass of the projectile. In the following we give the energy loss expressions valid for different velo~ city regions. 2.1 High Velocity Region (/m > 1 MeV/u) When light projectiles such as hydrogen/helium and other light ions move with a sufficiently high velocity (E/m 2 1 MeV/u), their orbital electrons are totally stripped. The stopping cross-section of a fully ionized light particle by excitation and ionization is given by the Bethe equation [3, 4, 5}. 4n2ZAT 2mow v) 2c 8 = ESE lin OP —in(i- 3)-S+F-F] © Where mis the electron mass, ze is the projectile charge and ¢ is the velocity of light. is the mean excitation and ionization energy which is a property of thé target atom. It is a suitably weighted average of all thie” possible “excita- tion and ionization processes for a given atom. Since quantum mechanical calculations of this quantity are rather complex, reliable values of J are bet- ter determiried empirically from the measurements of the energy loss [6, 7]- Bloch [8] has given an approximate relation for estimating J for heavy ele- ments, as I = Z Ip where J ~ 14 eV: Oscillations. of 1/Z around this value have also been observed, The term C/Z accounts for the shell corrections arising from the nonparticipation of the inner shell electrons in the stopping process and is important only at the lower energies. Methods for treating shell corrections have been described in literature (9, 10]. Empirical esti- mates of C/Z can also be obtained from measurements of the energy loss. The term 8/2 arising. from the, density effect is important only at the high etiétgy @tid [11]. The two, remaining correction terms in Eq. (1) originate from the relativistic correction. ‘Neglecting smaller correction terms, the specific energy loss expression can be written in a simplified form for the non-relativistic region (v < c), as follows: dE _ dwztet 2rmg-0 oer a The observed energy dependence: of the stopping cross-section in the in- termediate and high energy regions of Figs. 2 and 3 is in general agreement with predictions of Eq. (1) with appropriate value of J [2}. 2.2 Low Velocity Region (Z/m < 10 keV/u) Eq. (1) is based on the assumption that the projectile ions are fully stripped of electrons and a bare ion is interacting with the electrons of the target atoms. However when the ion velocity becomes small, its equilibrium ionic charge decreases as it starts capturing-an appreciable number of electrons. At very low velocities (E/m < 10 keV/u), the charge neutralisation of the y Scanned with CamScanner10 NUCLEAR RADIATION DETECTORS projectile ion is nearly complete and the collisions between the projectile and the electrons of the target atoms are almost elasticin a reference frame mov- ing with the ion. Lindhard, Scharff and Schiott (LSS) [3] and Firsov [12, 13, 14] have given theoretical descriptions for this energy region where the electronic energy loss is found to be nearly proportional to the velocity of the projectile. Another contribution to the energy loss arises from the Rutherford colli- | sion of the projectile with the target nuclei. This contribution, known as the nuclear contribution, is in general quite smail as compared to the electronic contribution. For light ions in most detector media, the electronic energy loss dominates the total energy loss even down to the energy region of the order of a few keV. However, for heavy ions, the nuclearcontribution starts dominating at low projectile velocities as evident from the bump at low values of E/m seen in Fig. 3. 2.3 Medium Velocity Region (10 keV/u < Elm < 1 MeV/u) The Bethe expression (Eq. 1) is valid only when the projectile is totally ionized. This requires that the Velocity of the: projectile is: much larger than 2 zo where v is the Bohr velocity. For example, this criterion is fulfilled for hydrogen atoms with energies greater than 100 keV, helium atoms with energies greater than 1.6 MeV and, in general, heavy atoms with energies greater than 0.1 mz? MeV. At lower energies, the projectiles are not fully ionized and during slowing down there is a continuousionization and recap- ture of electrons resulting in an “effective charge” of the projectile at any given instant. Bohr has given [15] the following estimation of the effective charge at low velocities eee SRS @) A general approximation for the stopping cross-section for an ion with | nuclear charge z is given as = <2?) Get (4) where Vz?) = yz, and ¢,-1 can be obtained from Eq. (1). The parameter 7 depends on the ion velocity. Valués of the parameter y can be derived from experimental measurements of the specific energy loss for different heavy ions. These values follow semiempirical expressions of the general type y= 1-4 exp(22 2) es where v9 is the orbital velocity of the electron in the first Bohr orbit of the hydrogen atom and A, B, c are constants, The value of ¢ would be 2/3 if y is determined by the ratio of the ion velocity to the Thomas-Fermi electron velocity. Forster et al. [16] have proposed the following values for the con- stants A and B based on measurements of the energy loss of five different ions in six different metal targets. Scanned with CamScanner(6) BASIC PHYSICAL PROCESSES IN RADIATION DETECTION A = 1.035~0.4 exp (—0.162) B= 0.875 The energy loss behaviour seen in Fig. 3 can be satisfactorily accounted on the basis of Eqs. (1) to (6). An energetic charged particle slowing down in a medium spans all the velocity regions—high, low and medium—discussed above. In the beginning of its path, if the projectile is fully stripped, the specific energy loss is pro- Portional to the square of the projectile charge and increases with- decreas- ing projectile velocity roughly as 1/v? as given by Eq. (1). However, as’the Projectile slows down and electron pick up by the projectile becomes impor- tant, the effective charge of the projectile decreases and the specific energy loss starts falling off after going through a maximum. Since projectiles witli a larger atomic number begin to pick up electrons early in their slowing down process, both the maximum specific energy loss and the projéctile energy at which this maximum is reached depend on the mass and charge of the pro- jectile as seen in Fig. 3: Figure 4 shows typical plots of. the specific energy loss along the track of the projectile, known as Bragg curves, for some light and heavy ions in a gaseous medium. It has been shown [17] that the energy lost by the particle in the last portion of its range is more sensitive to the particle type and hence can be utilized with advantage for particle identi- fication. A number of compilations with tables of interpolated and extrapolated. energy loss values can be found in the literature [1, 11, 18-22]. 2.4 Range-Energy Relations for Heavy Charged Particles Another quantity of practical interest related to the differential stopping power of a heavy charged particle in matter is its range in the stopping material. Because of a large difference in the mass of the projectile and the deviation of the projectile from its initial direction .of motion electro: is negligible and one can define the range of the projectile as follows: = dE R= | marae co The above integration can be carried out in general only numerically, since, as discussed earlier, no simple analytical expression exists for the differential energy loss which is valid for the entire energy range of the projectile. Con- sequently, several empirical range-energy relations for various charged part cles in materials bave been proposed. Also scaling laws to relate the ranges for different projectiles, projectile velocities and stopping materials have en Proposed. For the non-relativistic case with vc, where the energy loss expression can is i eee be approximated by Eq. (2), the particle range can be * SBarenz fam ; @) ie sees Hs ealJ2_ NUCLEAR RADIATION DETECTORS xe (12 MeV/u) \ 120) 14x) — MeV + mg sem? 2 © 3.8 “aot 3 MeV/ u) Sy (ed Wocumev/u) 20 720 «140 «160 0 20 40 60 80 100 PATH LENGTH IN cms IN METHANE Fig. 4. Bragg curves for heavy ions in methane. Gas density = 0.1314 mg/cm? (data taken from (21). “Thus for the energy region in which Eq. (2) is valid, the particle ranges vary jn.a simple manner with respect to projectile mass, charge and velocity. For ‘xample, for two particles which have different values of m/z? such as *He gad ‘Hee, but the same initial velocity, the ranges are proportional to m/z. More accurate range comparisons can be made between particles of diffe- rent z by taking into account the different effective charges of the projectiles stow velocities. Substitation of the effective charge (Bd 4) with values of + less than unity for smaller velocities encountered during slowing down of the particle would result in a smaller stopping power and a larger range and tnisborrection would be more for particles of large z._ Thus it is found that the range of protons in air is shorter than the range of alpha particles of the sree initial velocity; the following expression provides a good estimate of this ratio — (ml?) CR 7 1 = ES Ra Fe) Om where subscript a refers to alpha pé of any mass, 1, 2 or 3. Ro isa suitable constant. ranges of a particle in different stopping materials. A well known appro: mation is the Bragg-Kleeman rule which provides a quick but approximate estimate of relative ranges of particles jn different substances. According to this rule, the ratio of the ranges in two different materials is given by articles and H refers to a hydrogen ion ‘One can also relate the Ri_ vA 10) Ra pda 2 | Scanned with CamScannerBASIC PHYSICAL PROCESSES IN RADIATION DETECTION 13 where p is the density and A is the atomic weight, and suffix 1, 2 refer tothe vio media. Compounds and mixtures follow the same empirical rule pro- sided one defines an effective atomic weight as Age Yuma ay 4= Sn TAs where mare the atomic fractions. Figures 5 and 6 show the range-energy curves in silicon for the light charged particles and some heavy ions. The form of the energy range curves suggests a power law of the form R = aE* to adequately represent the ranges seers wide range of particles and energies. The constant @ depends on the identity of the particle (~ wre) and to a lesser extent on the absorb- bhasa value near 1.7 but varies slowly with is simple relation has proven useful in identi- urement of the partial energy loss 4E It ing material, while the constant energy and type of particle, Th fying types of particles from [Link] in a transmission detector and the residual energy in a stop detector. can be shown that z = (E+ 4EP-B the transmission detector., Equation (12) gives ” which depends only on the identity of the (12) where Tis the thickness of the value of the parameter “‘a projectile. z Energy( MeV ) orurenon w rd 10 a Range in microns of silicon Fig. 5. Range-energy curves for hydi is for hydrogen and helium isotopes in silicon (igure reproduced with permission from EG and G ORTEC catalogue). Scanned with CamScanner