Experimental measurements validate the use of the binary encounter approximation model to accurately compute proton induced dose and radiolysis enhancement from gold nanoparticles

Félicien Hespeels, Stephane Lucas, Tijani Tabarrant, Emanuele Scifoni, Michael Kraemer, Gregoire Chêne, David Strivay, Hoang N Tran, Anne-Catherine Heuskin

Research output: Contribution to journalArticlepeer-review


In protontherapy, it has been suggested that nanoparticles of high-Z material like gold (GNP) could be used as radiosensitizers. The origin of this enhancement phenomenon for proton radiation is not yet well understood and additional mechanistic insights are required. Previous works have highlighted the good capabilities of TRAX to reproduce secondary electron emission from gold material. Therefore, TRAX cross sections obtained with the binary encounter approximation (BEA) model for proton ionization were implemented within Geant4 for gold material. Based on the TRAX cross sections, improved Geant4 simulations have been developed to investigate the energy deposition and radical species production around a spherical gold nanoparticle (5 and 10 nm in diameter) placed in a water volume during proton irradiation. Simulations were performed for incident 2 MeV proton. The dose enhancement factor and the radiolysis enhancement factor were quantified. Results obtained with the BEA model were compared with results obtained with condensed-history models. Experimental irradiation of 200 nm gold films were performed to validate the secondary electron emission reproduction capabilities of physical models used in Monte Carlo (MC) simulations. TRAX simulations reproduced the experimental backscattered electron energy spectrum from gold film with better agreement than Geant4. Results on gold film obtained with the BEA model enabled to estimate the electron emission from GNPs. Results obtained in our study tend to support that the use of the BEA discrete model leads to a significant increase of the dose in the near vicinity of GNPs (<20 nm), while condensed history models used in Geant4 seem to overestimate the dose and the number of chemical species for increasing distances from the GNP. Based on discrete BEA model results, no enhancement effect due to secondary electron emitted from the GNP is expected if the GNP is not in close proximity to key cellular functional elements (DNA, mitochondria…).

Original languageEnglish
Article number065014
Number of pages19
JournalPhysics in Medicine and Biology
Issue number6
Early online date2019
Publication statusPublished - 13 Mar 2019


  • Geant4
  • gold nanoparticle
  • low energy electron
  • Monte Carlo track structure code
  • Proton beam
  • secondary electron emission
  • TRAX

Cite this