[en] A proposed fabrication route for ITER first wall components implies a diffusion welding step of Be tiles onto a Cu-based substrate. However, Be has a tendency to form particularly brittle intermetallics with Cu and a lot of other elements. Insertion of interlayers may be a solution to increase bond quality. Applying traditional analyses to this study can be problematic because of Be toxicity and low atomic number Z. Ion Beam Analysis methods have thus been considered together with scanning electron microscopy (SEM) and electron back-scattering diffraction (EBSD) as complementary techniques. The following work aims at demonstrating how such techniques (used in micro-beam mode), and in particular NRA (Nuclear Reaction Analysis) and PIXE (Particle Induced X-ray Emission) techniques, coupled with SEM/EBSD data, can bring valuable information in this area. Quantification of data allow to obtain concentration values (provided the hypotheses on the initial junction composition are valuable), then phase diagrams give clues about the composition and structure of the junction. SEM retro-diffused electrons chemical contrast images and EBSD allow to characterize the presence of the awaited intermetallics, and finally confirm or refine the conclusions of Ion Beam Analysis data quantification. A series of reference first wall mock-ups have been analysed. Interlayer-free mock-ups reveal intermetallics which are mainly BeCu (apparently mixed with lower quantities of BeCu₂ compound). While Cr or Ti interlayers seem to behave as good Be diffusion barriers in the sense that they prevent the formation of BeCu, they strongly interact with Cu to form CuTi₂ or Cr₂Ti intermetallics. In the case of Cr, Be seems to be incorporated into the Cr layer. PIXE analysis has however been unable to characterize Al-based interlayers (Z=13, close to the lower PIXE sensibility limit) and emphasizes one limitation of Ion Beam Analysis methods for lighter metals, justifying the use of other complementary techniques like SEM and EBSD. Next steps to this work will be to apply the analysis procedure described in this study on the soon coming tested mock-ups, in order to characterize the impact of the different tests (high heat flux tests and thermal fatigue tests) on the evolution and performance of the Be/Cu junctions. (author)