Two models for the calculation of ionization cross sections by electron impact on atoms, the Binary-Encouter-Bethe and the Deutsch-Maerk models, have been implemented; they are intended to extend and improve Geant4 simulation capabilities in the energy range below 1 keV. The physics features of the implementation of the models are described, and their differences with respect to the original formulations are discussed. Results of the verification with respect to the original theoretical sources and of extensive validation with respect to experimental data are reported. The validation process also concerns the ionization cross sections included in the Evaluated Electron Data Library used by Geant4 for low energy electron transport. Among the three cross section options, the Deutsch-Maerk model is identified as the most accurate at reproducing experimental data over the energy range subject to test.
V ARIOUS experimental research topics require the capa- bility of simulating electron interactions with matter over a wide range -from the nano-scale to the macroscopic one: some examples are ongoing investigations on nanotechnologybased particle detectors, scintillators and gaseous detectors, radiation effects on semiconductor devices, background effects on X-ray telescopes and biological effects of radiation.
Physics tools for the simulation of electron interactions are available in all Monte Carlo codes based on condensed and mixed transport schemes [1], like EGS [2], [3], FLUKA [4], [5], Geant4 [6], [7], MCNPX [8], Penelope [9] and PHITS [10]. General-purpose Monte Carlo codes based on these transport schemes typically handle particles with energy above 1 keV; Geant4 and Penelope extend their coverage below this limit.
In the lower energy end below 1 keV, so-called “track structure” codes handle particle interactions based on discrete transport schemes; they provide simulation capabilities limited to a single target, or a small number of target materials, and are typically developed for specific application purposes. Some examples of such codes are OREC [11], PARTRAC [12], Grosswendt’s Monte Carlo for nanodosimetry [13], TRAMOS [14], and Geant4 models for microdosimetry simulation in water [15].
The developments described in this paper address the problem of endowing a general purpose, large scale Monte Carlo system for the first time with the capability of simulating electron impact ionisation down to the scale of a few tens of electronvolts for any target element. For this purpose, models of electron impact ionization cross sections suitable to extend Geant4 capabilities in the low energy range have been implemented and validated with respect to a large set of experimental measurements.
The validation process, which involves experimental data pertinent to more than 50 elements, also addresses the ionization cross sections encompassed in the Evaluated Electron Data Library (EEDL) [16], which are used in Geant4 low energy electromagnetic package [17], [18]. To the best of the authors’ knowledge, this is the first time that EEDL is subject to extensive experimental benchmarks below 1 keV.
The Geant4 toolkit provides various implementations of electron ionization based on a condensed-discrete particle transport scheme. Two of them, respectively based on EEDL [19] and on the analytical models originally developed for the Penelope [9] Monte Carlo system, are included in the low energy electromagnetic package; another implementation is available in the standard [20] electromagnetic package. In addition, a specialized ionization model for interactions with thin layers of material, the photoabsorption-ionization (PAI) model [21], is implemented in Geant4.
The EEDL data library tabulates electron ionization cross sections in the energy range between 10 eV and 100 GeV; nevertheless, due to intrinsic limitations of the accuracy of EEDL and its companion Evaluated Photon Data Libray (EPDL) [22] highlighted in the documentation of these compilations, the use of Geant4 low energy models based on them was originally recommended for incident electron energies above 250 eV [19]. This limit of applicability was an “educated guess” rather than a rigorous estimate of validity of the theoretical calculations tabulated in EEDL and EPDL. The lower energy limit of Penelope’s applicability is generically indicated by its authors as “a few hundred electronvolts” [23]. The lower limit of applicability of Geant4 standard electromagnetic package is 1 keV.
The validation of Geant4 models for electron transport based on the EEDL data library and on Penelope-like models is documented in [24] for what concerns the energy deposition in extended media.
Ionization models suitable for microdosimetry simulation, which operate in a discrete particle transport scheme, are available in Geant4 for electron interactions in water [15]; they are applicable for energies down to the electronvolt scale. The cross section models implemented in that context are specific to one material (liquid water); due to lack of pertinent experimental data, their validation is still pending.
The developments described in this paper adopt an iterativeincremental process consistent with the Unified Process [25]. The features and results documented in the following sections correspond to the first cycle of a wider project concerning the development and assessment of models for multi-scale electron transport [26], [27], which is motivated by multi-disciplinary experimental applications. A characterizing feature of the Unified Process, which differentiates it from other widespread software life-cycle models adopting a waterfall [28] approach, is the production of concrete deliverables even at intermediate stages of the project: this development cycle has enabled the validation and comparative evaluation of different physics models, and has produced a data libr
This content is AI-processed based on open access ArXiv data.
