Search results

Heating is a major cause of failure in integrated circuits. The authors have designed thermoreflectance‐based systems operating at various wavelengths in order to obtain temperature images. This paper aims to explore the possibilities of each wavelength range and detail the charge coupled device (CCD)‐based thermal imaging tools dedicated to the high‐resolution inspection of integrated circuits.

Thermoreflectance is a non‐contact optical method using the local reflectivity variations induced by heating to infer temperature mappings, and can be conducted at virtually any wavelength, giving access to different types of information. In the visible, the technique is now well established. It can probe temperatures through several micrometers of transparent encapsulation layers, with sub‐μm spatial resolution and 100 mK thermal resolution.

In the ultraviolet range, dielectric encapsulation layers are opaque and thermoreflectance gives access to the surface temperature. In the near infrared, thermoreflectance is an interesting solution to examine chips turned upside down, since these wavelengths can penetrate through silicon substrates and give access to the temperature of the active layers themselves.

The authors show that the illumination wavelength of thermoreflectance should be chosen with care depending on the region of the integrated circuit (surface, above, or below the substrate) to be investigated.

This set of versatile and sensitive tools makes thermoreflectance an interesting tool for the semiconductor industry, either during prototyping or as a characterization tool after fabrication.

The CCD‐based thermoreflectance approach adopted here allows fast, non‐contact, high‐resolution thermal imaging of integrated circuits.

The evolution of a semi‐open cooling circuit of a nuclear reactor was monitored over a two year period. The work aims to provide orientation elements for preventive procedures against localised corrosion.

The water of the circuit was analysed in stagnation and in circulation, at various sampling points. The rust was analysed by neutron diffraction and the surface quality of the steel was checked by microscopic observations.

The obtained results did not confirm the presence of rust in iron compounds supported by chlorine, such as the Akaganeite, β‐FeOOH. In addition, chemical analysis of water showed that, after two years, the increase of chlorine concentration and water conductivity remained weak. Moreover, the pH was maintained within values favourable rather to the passivation of the steel.

It was deduced through this work that the dosing of the circuit with chlorine was not sufficient that it should require an annual replacement of the water.

The originality of this work resides in the evaluation of a semi‐open coolant circuit in service for ten years and located in an area subjected to seasonal sand winds.