Electron Transport Algorithm Research Articles

Computationally efficient Monte Carlo electron transport algorithm for nanostructured thermoelectric material configurations

Monte Carlo statistical ray-tracing methods are commonly employed to simulate carrier transport in nanostructured materials. In the case of a large degree of nanostructuring and under linear response (small driving fields), these simulations tend to be computationally overly expensive due to the difficulty in gathering the required flux statistics. Here, we present a novel Monte Carlo ray-tracing algorithm with computational efficiency of at least an order of magnitude compared to existing algorithms. Our new method, which is a hybrid of the analytical Boltzmann transport equation and Monte Carlo used a reduced number of ray-tracing particles, avoids current statistical challenges, such as the subtraction of two opposite going fluxes, the application of a driving force altogether, and the large simulation time required for low-energy carriers. We demonstrate the algorithm’s efficiency and power in accurate simulations in large domain nanostructures with multiple defects. We believe that the new method we present is indeed more robust and user friendly compared to common methods and can enable the efficient study of transport in nanostructured materials under low-field steady-state conditions.

A Comprehensive Analysis of Radiosensitization Properties of Metallic Nanoparticles in Brachytherapy of Gastric Adenocarcinoma by I-125 Seed: A Simulation Study by MCNPX and MCNP6 Codes

Purpose: In the current study, we aimed to look into the macroscopic and microscopic dose enhancement effect of metallic nanoparticles in interstitial brachytherapy of gastric adenocarcinoma by Iodine-125 source using a nano-lattice model in MCNPX (2.7) and MCNP6.1 codes. Materials and methods: Based on a nano-lattice simulation model containing a radiation source and a tumor tissue with cellular compartments loaded with 7 mg/g spherical nanoparticles, the microscopic and macroscopic levels of energy deposition by the secondary electrons was estimated. Results: The results show that the values of macroscopic DEF are higher than microscopic DEF values and the macroscopic DEF values decrease by increasing the distance from the surface of brachytherapy source. Accordingly, it could be noted that gold nanoparticles have the highest radiosensitization effect among the other nanoparticles and the related DEF value is close to the resultant DEF values for bismuth nanoparticles. Moreover, the results revealed a remarkable discrepancy between the DEF and secondary electron spectra calculated by MCNPX (2.7) and MCNP6.1 codes, which could be justified by the difference in energy cut-off and electron transport algorithms of two codes. Conclusions: According to the both MCNPX (2.7) and MCNP6.1 outputs, it could be concluded that the presence of metallic nanoparticles in the tumor tissue of gastric adenocarcinoma increases the physical effectiveness of brachytherapy by I-125 source. This study aims to provide recommendations for future preclinical studies. Actually, the results presented herein give a physical view of radiosensitization potential of different metallic nanoparticles and could be considered in design of analytical and experimental radiosensitization studies in tumor regions using various radiotherapy modalities in the presence of heavy nanomaterials.

Adoption of ICRU report 90 recommendations in the Canadian Co-60 air-kerma primary standard

The Canadian air-kerma primary standard for 60Co beams is updated according to ICRU report 90 (ICRU-90) recommendations. The effect of these recommendations and a more detailed chamber model on the primary standard is investigated. Dosimetric quantities and corrections required for the realization of air kerma are calculated using the EGSnrc Monte Carlo (MC) simulation system. The validity of Spencer–Attix (SA) cavity theory as function of Δ, the secondary particle production threshold energy, is investigated to assess the accuracy of its current selection criterion. Although individual changes in some correction factors are statistically non-negligible, the overall effect amounts to only a 0.04% increase of the standard. The mean restricted mass electronic stopping power ratio graphite to air, decreases from the previous value of 1.0010 to a value of 0.9949. If this value is used with the recommended W air value of 33.97 eV, the effect on the standard is a reduction of 0.613%. When the recommended value of 33.72 eV is used instead, the change results in a 0.835% reduction in K air. Using SA cavity theory to obtain K air with a Δ value of 20 keV, corresponding to the cavity’s mean chord length, reproduces a direct MC calculation at the 0.03% level. Type B uncertainties due to uncertainty in the photon and electron cross sections for the largest corrections K wall and K comp are estimated to be 0.01% and 0.10%, respectively. No statistically significant effect on the standard is observed due to changes in the correction factors. Fano tests and comparison to single scattering calculations demonstrate EGSnrc’s electron transport algorithm accuracy to be within 0.03% relative to its own cross sections. The latter also validates the criterion for selecting Δ based on the average chord length of the chamber’s cavity. A 0.2% difference in the restricted stopping power ratio obtained in this work and the value resulting from ICRU-90 recommendations has the potential for affecting standards relying on the absolute value of W air, thus further investigation is granted. The current update results in a 0.8% decrease of the Canadian 60Co air-kerma primary standard with a reduced uncertainty of 0.2%. This change is in excellent agreement with the reported change to the BIPM 60Co air-kerma standard.

Evaluation of the electron transport algorithm in magnetic field in EGS5 Monte Carlo code.

To evaluate the accuracy of electron transport in the magnetic field of Electron Gamma Shower version 5 (EGS5) by using the special Fano cavity test. To simulate electron transport in the magnetic field, the trajectory of the electron was reconstructed with a short step length to restrict fractional energy loss, and the maximum user step length (mxustep) was set at 0.01cm or 0.001cm. For the special Fano cavity test, three-layer slab Fano test geometry was used, and uniform and isotropic per unit mass mono-energetic electrons with 0.01, 0.1, 1.0, and 10MeV were permitted from the central axis of geometry in 0.35T and 1.5T. Furthermore, the magnetic field strength was scaled based on the mass density of the material. The relative difference between the calculated dose to gap and the theoretical value was evaluated. Furthermore, the special Fano cavity test was also performed using EGSnrc with the electron-enhanced electric and magnetic field macros under the same conditions, and the results were compared with those of EGS5. Deviations in 0.35T were within 0.3% regardless of the parameter settings. In 1.5T, stable results within 0.3% were obtained using 0.001cm as the mxustep, except for one at 10MeV. Further, the accuracy of EGSnrc was within 0.2%, except for 10MeV for a 0.2-cm gap in 1.5T. EGS5 with the appropriate parameter settings enable electron transport in magnetic fields similar with the accuracy of EGSnrc.

Technical Note: Implications of using EGSnrc instead of EGS4 for extracting electron stopping powers from measured energy spectra.

NRC Report PIRS-0626 (https://doi.org/10.4224/40000364) describes how measured electron energy deposition spectra can be used to determine the electronic stopping power. The stopping power is obtained by comparing measured spectra with spectra calculated using Monte Carlo techniques. The stopping powers reported in PIRS-0626 were obtained using the EGS4 Monte Carlo code. Since then, the EGSnrc code has been released which has more accurate electron transport algorithms. We calculate the effect on the measured stopping powers of using EGSnrc instead of EGS4. The EGS4 spectra calculated in PIRS-0626 were based on primary electron histories. We first show that those spectra, calculated in 1997, are consistent with current EGS4 spectra calculated using histories. EGSnrc spectra are also calculated using histories and these high-precision spectra are compared to extract any energy difference. The energy differences between the spectra are used to estimate the effect on the measured electronic stopping powers. The energy differences depend on the absorber material, the absorber thickness and the beam energy. The improved electron elastic scattering cross section of EGSnrc accounts for only part of the difference between the two codes. The effect on the extracted stopping power is largest for the lowest electron energies and can be as large as 0.9%. The calculated spectra show differences for lower energies, with the EGSnrc spectra having a larger proportion of low-energy electrons. The differences introduced by using EGSnrc instead of EGS4 can affect the estimated stopping power by almost 1% in the worst case but generally the effect is much smaller. We report corrections that can be applied to all the stopping power data in PIRS-0626. An experiment to measure the average energy to create an ion pair in air, , using aluminum detectors will provide an interesting test of the aluminum stopping power data as reported in PIRS-0626 and revised by this work.

Calculation of dose point kernel values for monoenergetic electrons and beta emitting radionuclides: Intercomparison of Monte Carlo codes

Targeted radionuclide therapy (TRT) and beta-emitting seeds brachytherapy (BSBT) exploit the characteristics of energy deposited by beta-emitting radionuclides. Monte Carlo (MC) modelling of electron transport is crucial for calculations of absorbed dose for TRT and BSBT. However, computer codes capable of providing consistent results are still limited. Since experimental validations show several difficulties, the estimation of electron dose point kernel (DPK) is often used to verify the accuracy of different MC codes. In this work, we compared DPK calculations for various point, isotropic and monoenergetic electron sources and several beta-emitting radioisotopes using the codes MCNP, EGSnrc, PENELOPE and TOPAS with different simulation options. The simulations were performed using latest versions of EGSnrc and Penelope, TOPAS version 3.3.1 and MCNP version 6.1 Monte Carlo codes. In our simulations, the geometrical model consists of a point electron source placed at the center of a water sphere emitting isotropically. The water sphere was divided into 28 shells and the energy deposition was scored within these shells. The radius of the outermost shell was 1.2R0, where R0 is the continuous slowing down approximation (CSDA) range. Five monoenergetic beta sources with energies of 0.05, 0.1, 0.5, 1 and 3 MeV were studied. Six beta-emitting radionuclides were also simulated: Lu-177, Sm-153, Ho-166, Sr-89, I-131 and Y-90. Monoenergetic electron simulations showed large deviations among the codes, larger than 13% depending on the electron energy and the distance from the source. In the cases where beta spectra of radionuclides were simulated, all MC codes showed differences from EGSnrc (used as reference value - RV) less than 3% within rE90 range (radius of the sphere in which 90% of the energy of the spectrum electrons would be deposited). TOPAS showed results comparable to EGSnrc and PENELOPE. DPK values for 0.1 MeV monoenergetic electrons, calculated using MCNP6, led to differences higher than ±5% from RV despite our attempts to tune electron transport algorithms and physics parameters.

Fano cavity test for electron Monte Carlo transport algorithms in magnetic fields: comparison between EGSnrc, PENELOPE, MCNP6 and Geant4

A Fano cavity test was performed for four general-purpose Monte Carlo codes, EGSnrc, PENELOPE, MCNP6 and Geant4 to evaluate the accuracy of their electron transport algorithms in magnetic fields. In the simulations, a plane-parallel ionization chamber was modelled as a circular gas disk sandwiched between two circular solid wall disks. It was assumed that an isotropic and uniform line source per unit mass along the central axis of the gas and solid emits mono-energetic electrons with energies 0.01, 0.1, 1.0 and 3.0 MeV at different magnetic field strengths 0, 0.35, 1.0, 1.5 and 3.0 T in the electron transport mode (no Bremsstrahlung). The relative difference between the calculated dose to the gas region and the initial total energy of emitted electrons per unit mass was defined as the accuracy of Monte Carlo codes. In all results, EGSnrc with the enhanced electric and magnetic field (EEMF) macros was not considerably sensitive to the step size parameters and showed accuracy less than 0.18% ± 0.06% with a coverage factor k = 2. The other codes could not achieve competent accuracy with their default settings of step size parameters, compared to EGSnrc with the EEMF macros. With the step size parameters carefully selected, the accuracy of PENELOPE and MCNP6 was within 1.0% and 0.4%, respectively. However, Geant4 showed accuracy within 1.7% except in 3.0 T. EGSnrc with the EEMF macros achieved the best accuracy for the Fano test at the electron energies and the magnetic field strengths investigated in this study and thus, would be recommended to simulate dose responses of ionization chambers in the presence of magnetic fields.

Advances in the Treatment of the Electromagnetic Cascade in the TRIPOLI-4 Monte Carlo Code

TRIPOLI-4 is a Monte Carlo particle transport code developed at CEA Saclay (France) that is employed in the domains of nuclear reactor physics, criticality safety, shielding/radiation protection, and nuclear instrumentation. The goal of this paper is to report on current developments, validation, and verification made in TRIPOLI-4 in the treatment of electron/positron/photon transport. The new capabilities and improvements concern refinements to the electron transport algorithm, the introduction of a charge-deposition score, the new thick-target bremsstrahlung option, the upgrade of the bremsstrahlung model, and the improvement of electron angular straggling at low energy. The importance of each of the developments mentioned above is illustrated by comparisons with calculations performed with other codes and with experimental data.

Advances in Monte-Carlo code TRIPOLI-4®’s treatment of the electromagnetic cascade

TRIPOLI-4® is a Monte-Carlo particle-transport code developed at CEA-Saclay (France) that is employed in the domains of nuclear-reactor physics, criticality-safety, shielding/radiation protection and nuclear instrumentation. The goal of this paper is to report on current developments, validation and verification made in TRIPOLI-4 in the electron/positron/photon sector. The new capabilities and improvements concern refinements to the electron transport algorithm, the introduction of a charge-deposition score, the new thick-target bremsstrahlung option, the upgrade of the bremsstrahlung model and the improvement of electron angular straggling at low energy. The importance of each of the developments above is illustrated by comparisons with calculations performed with other codes and with experimental data.

Evaluation of the new electron-transport algorithm in MCNP6.1 for the simulation of dose point kernel in water

Abstract In this work, the calculation of specific dose distribution in water is evaluated in MCNP6.1 with the regular condensed history algorithm the “detailed electron energy-loss straggling logic” and the new electrons transport algorithm proposed the “single event algorithm”. Dose Point Kernel (DPK) is calculated with monoenergetic electrons of 50, 100, 500, 1000 and 3000 keV for different scoring cells dimensions. A comparison between MCNP6 results and well-validated codes for electron-dosimetry, i.e., EGSnrc or Penelope, is performed. When the detailed electron energy-loss straggling logic is used with default setting (down to the cut-off energy 1 keV), we infer that the depth of the dose peak increases with decreasing thickness of the scoring cell, largely due to combined step-size and boundary crossing artifacts. This finding is less prominent for 500 keV, 1 MeV and 3 MeV dose profile. With an appropriate number of sub-steps (ESTEP value in MCNP6), the dose-peak shift is almost complete absent to 50 keV and 100 keV electrons. However, the dose-peak is more prominent compared to EGSnrc and the absorbed dose tends to be underestimated at greater depths, meaning that boundaries crossing artifact are still occurring while step-size artifacts are greatly reduced. When the single-event mode is used for the whole transport, we observe the good agreement of reference and calculated profile for 50 and 100 keV electrons. Remaining artifacts are fully vanished, showing a possible transport treatment for energies less than a hundred of keV and accordance with reference for whatever scoring cell dimension, even if the single event method initially intended to support electron transport at energies below 1 keV. Conversely, results for 500 keV, 1 MeV and 3 MeV undergo a dramatic discrepancy with reference curves. These poor results and so the current unreliability of the method is for a part due to inappropriate elastic cross section treatment from the ENDF/B-VI.8 library in those energy ranges. Accordingly, special care has to be taken in setting choice for calculating electron dose distribution with MCNP6, in particular with regards to dosimetry or nuclear medicine applications.

Global modeling of thermospheric airglow in the far ultraviolet

AbstractThe Global Airglow (GLOW) model has been updated and extended to calculate thermospheric emissions in the far ultraviolet, including sources from daytime photoelectron‐driven processes, nighttime recombination radiation, and auroral excitation. It can be run using inputs from empirical models of the neutral atmosphere and ionosphere or from numerical general circulation models of the coupled ionosphere‐thermosphere system. It uses a solar flux module, photoelectron generation routine, and the Nagy‐Banks two‐stream electron transport algorithm to simultaneously handle energetic electron distributions from photon and auroral electron sources. It contains an ion‐neutral chemistry module that calculates excited and ionized species densities and the resulting airglow volume emission rates. This paper describes the inputs, algorithms, and code structure of the model and demonstrates example outputs for daytime and auroral cases. Simulations of far ultraviolet emissions by the atomic oxygen doublet at 135.6 nm and the molecular nitrogen Lyman‐Birge‐Hopfield bands, as viewed from geostationary orbit, are shown, and model calculations are compared to limb‐scan observations by the Global Ultraviolet Imager on the TIMED satellite. The GLOW model code is provided to the community through an open‐source academic research license.

TU-AB-BRC-04: Commissioning of a New MLC Model for the GEPTS Monte Carlo System: A Model Based On the Leaf and Interleaf Effective Density

Purpose: To commission a new MLC model for the GEPTS Monte Carlo system. The model is based on the concept of leaves and interleaves effective densities Methods: GEPTS is a Monte Carlo system to be used for external beam planning verification. GEPTS incorporates detailed photon and electron transport algorithms (Med.Phys. 29, 2002, 835). A new GEPTS model for the Varian Millennium MLC is presented. The model accounts for: 1) thick (1 cm) and thin (0.5 cm) leaves, 2) tongue-and-groove design, 3) High-Transmission (HT) and Low-Transmission (LT) interleaves, and 4) rounded leaf end. Leaf (and interleaf) height is set equal to 6 cm. Instead of modeling air gaps, screw holes, and complex leaf heads, “effective densities” are assigned to: 1) thin leaves, 2) thick leaves, 3) HT-, and 4) LT-interleaves. Results: The new MLC model is used to calculate dose profiles for Closed-MLC and Tongue-and-Groove fields at 5 cm depth for 6, 10 and 15 MV Varian beams. Calculations are compared with 1) Pin-point ionization chamber transmission ratios and 2) EBT3 Radiochromic films. Pinpoint readings were acquired beneath thick and thin leaves, and HT and LT interleaves. The best fit of measured dose profiles was obtained for the following parameters: Thick-leaf density = 16.1 g/cc, Thin-leaf density = 17.2 g/cc; HT Interleaf density = 12.4 g/cc, LT Interleaf density = 14.3 g/cc; Interleaf thickness = 1.1 mm. Attached figures show comparison of calculated and measured transmission ratios for the 3 energies. Note this is the only study where transmission profiles are compared with measurements for 3 different energies. Conclusion: The new MLC model reproduces transmission measurements within 0.1%. The next step is to implement the MLC model for real plans and quantify the improvement in dose calculation accuracy gained using this model for IMRT plans with high modulation factors.

Behaviors of the percentage depth dose curves along the beam axis of a phantom filled with different clinical PTO objects, a Monte Carlo Geant4 study

Abstract The aim of this work is to assess the capabilities of Monte Carlo Geant4 code to reproduce the real percentage depth dose (PDD) curves generated in phantoms which mimic three important clinical treatment situations that include lung slab, bone slab, bone-lung slab geometries. It is hoped that this work will lead us to a better understanding of dose distributions in an inhomogeneous medium, and to identify any limitations of dose calculation algorithm implemented in the Geant4 code. For this purpose, the PDD dosimetric functions associated to the three clinical situations described above, were compared to one produced in a homogeneous water phantom. Our results show, firstly, that the Geant4 simulation shows potential mistakes on the shape of the calculated PDD curve of the first physical test object (PTO), and it is obviously not able to successfully predict dose values in regions near to the boundaries between two different materials. This is, surely due to the electron transport algorithm and it is well-known as the artifacts at interface phenomenon. To deal with this issue, we have added and optimized the StepMax parameter to the dose calculation program; consequently the artifacts due to the electron transport were quasi disappeared. However, the Geant4 simulation becomes painfully slow when we attempt to completely resolve the electron artifact problems by considering a smaller value of an electron StepMax parameter. After electron transport optimization, our results demonstrate the medium-level capabilities of the Geant4 code to modeling dose distribution in clinical PTO objects.

Comparison between EGSnrc, Geant4, MCNP5 and Penelope for mono-energetic electron beams

A simple geometry is chosen to highlight similarities and differences of current electron transport algorithms implemented in four Monte Carlo codes commonly used in radiation physics. Energy deposited in a water-filled sphere by mono-energetic electron beams was calculated using EGSnrc, Geant4, MCNP5 and Penelope as the radius of the sphere varied from 0.25 cm to 4.5 cm for beam energies of 0.5 MeV, 1.0 MeV and 5.0 MeV. The calculations were performed in single-scattering mode (where applicable) and in condensed history mode. A good agreement is found for the single-scattering calculations except for the in-air case at 0.5 MeV where differences increase with decreasing radius up to 5% between EGSnrc and Penelope. Differences between results calculated with the default user settings when compared to their own single-scattering modes are under 5% for all codes when the sphere is surrounded by vacuum, however, large differences occur for Geant4, MCNP5 and Penelope when air is introduced around the sphere. Finally, the parameters associated with the multiple scattering algorithms were tuned reducing these differences below 10% for these codes at the expense of increased computation time.

A fast, parameterized model of upper atmospheric ionization rates, chemistry, and conductivity

AbstractRapid specification of ionization rates and ion densities in the upper atmosphere is essential when many evaluations of the atmospheric state must be performed, as in global studies or analyses of on‐orbit satellite data. Though many models of the upper atmosphere perform the necessary specification, none provide the flexibility of computational efficiency, high accuracy, and complete specification. We introduce a parameterized, updated, and extended version of the GLobal AirglOW (GLOW) model, called GLOWfast, that significantly reduces computation time and provides comparable accuracy in upper atmospheric ionization, densities, and conductivity. We extend GLOW capabilities by (1) implementing the nitric oxide empirical model, (2) providing a new model component to calculate height‐dependent conductivity profiles from first principles for the 80–200 km region, and (3) reducing computation time. The computational improvement is achieved by replacing the full, two‐stream electron transport algorithm with two parameterizations: (1) photoionization (QRJ from Solomon and Qian ()) and (2) electron impact ionization (F0810 from Fang et al. (, )). We find that GLOWfast accurately reproduces ionization rates, ion and electron densities, and Pedersen and Hall conductivities independent of the background atmospheric state and input solar and auroral activity. Our results suggest that GLOWfast may be even more appropriate for low characteristic energy auroral conditions. We demonstrate in a suite of 3028 case studies that GLOWfast can be used to rapidly calculate the ionization of the upper atmosphere with few limitations on background and input conditions. We support these results through comparisons with electron density profiles from COSMIC.

Monte Carlo radiation transport in external beam radiotherapy

The use of Monte Carlo in radiation transport is an effective way to predict absorbed dose distributions. Monte Carlo modeling has contributed to a better understanding of photon and electron transport by radiotherapy physicists. The aim of this review is to introduce Monte Carlo as a powerful radiation transport tool. In this review, photon and electron transport algorithms for Monte Carlo techniques are investigated and a clinical linear accelerator model is studied for external beam radiotherapy. The statistical uncertainties and variance reduction techniques for Monte Carlo simulation are also discussed.

Evaluation of the accuracy of the electron transport algorithms of the MCNP6 and EGSnrc Monte Carlo codes using the Fano theorem

Evaluation of the accuracy of the electron transport algorithms of the MCNP6 and EGSnrc Monte Carlo codes using the Fano theorem

An evaluation of calculation parameters in the EGSnrc/BEAMnrc Monte Carlo codes and their effect on surface dose calculation

The Monte Carlo (MC) method has proven invaluable for radiation transport simulations to accurately determine radiation doses and is widely considered a reliable computational measure that can substitute a physical experiment where direct measurements are not possible or feasible. In the EGSnrc/BEAMnrc MC codes, there are several user-specified parameters and customized transport algorithms, which may affect the calculation results. In order to fully utilize the MC methods available in these codes, it is essential to understand all these options and to use them appropriately. In this study, the effects of the electron transport algorithms in EGSnrc/BEAMnrc, which are often a trade-off between calculation accuracy and efficiency, were investigated in the buildup region of a homogeneous water phantom and also in a heterogeneous phantom using the DOSRZnrc user code. The algorithms and parameters investigated include: boundary crossing algorithm (BCA), skin depth, electron step algorithm (ESA), global electron cutoff energy (ECUT) and electron production cutoff energy (AE). The variations in calculated buildup doses were found to be larger than 10% for different user-specified transport parameters. We found that using BCA = EXACT gave the best results in terms of accuracy and efficiency in calculating buildup doses using DOSRZnrc. In addition, using the ESA = PRESTA-I option was found to be the best way of reducing the total calculation time without losing accuracy in the results at high energies (few keV ∼ MeV). We also found that although choosing a higher ECUT/AE value in the beam modelling can dramatically improve computation efficiency, there is a significant trade-off in surface dose uncertainty. Our study demonstrates that a careful choice of user-specified transport parameters is required when conducting similar MC calculations.

PEBSI – A Monte Carlo simulator for bremsstrahlung arising from electrons colliding with thin solid-state targets

We present a Monte Carlo code dedicated to the simulation of bremsstrahlung arising in collisions of polarized electrons with thin target foils. The program consists of an electron transport algorithm taking into account elastic electron-nucleus scattering and inelastic collisions with target electrons as well as a treatment of polarized-electron bremsstrahlung emission. Good agreement is found between the predictions of the electron transport code and data stemming from other simulation programs and experiments. In addition, we present first results from the bremsstrahlung simulation which indicate a significant decrease in the degree of linear polarization of bremsstrahlung even for the thinnest gold targets considered.

Effects of physics change in Monte Carlo code on electron pencil beam dose distributions

Pencil beam algorithms used in computerized electron beam dose planning are usually described using the small angle multiple scattering theory. Alternatively, the pencil beams can be generated by Monte Carlo simulation of electron transport. In a previous work, the 4th version of the Electron Gamma Shower (EGS) Monte Carlo code was used to obtain dose distributions from monoenergetic electron pencil beam, with incident energy between 1 MeV and 50 MeV, interacting at the surface of a large cylindrical homogeneous water phantom. In 2000, a new version of this Monte Carlo code has been made available by the National Research Council of Canada (NRC), which includes various improvements in its electron-transport algorithms. In the present work, we were interested to see if the new physics in this version produces pencil beam dose distributions very different from those calculated with oldest one. The purpose of this study is to quantify as well as to understand these differences. We have compared a series of pencil beam dose distributions scored in cylindrical geometry, for electron energies between 1 MeV and 50 MeV calculated with two versions of the Electron Gamma Shower Monte Carlo Code. Data calculated and compared include isodose distributions, radial dose distributions and fractions of energy deposition. Our results for radial dose distributions show agreement within 10% between doses calculated by the two codes for voxels closer to the pencil beam central axis, while the differences are up to 30% for longer distances. For fractions of energy deposition, the results of the EGS4 are in good agreement (within 2%) with those calculated by EGSnrc at shallow depths for all energies, whereas a slightly worse agreement (15%) is observed at deeper distances. These differences may be mainly attributed to the different multiple scattering for electron transport adopted in these two codes and the inclusion of spin effect, which produces an increase of the effective range of electrons.