Study, analysis and correlation of the failure of electronic devices with the manufacturing process

Every company intrinsically has the goal of increasing the throughput, maximizing efficiency of the supply chain and reliability of their outcome. Considering the recent shortage of silicon-based electronics and attention to the climate issue, it is nowadays more and more important for companies to avoid imperfect devices losing precious material and wasting energy spent during the production process while, at the same time, shortening production and worker's time. Therefore, to preserve the environment and increase productivity, it is critical to develop a research branch focused on making sure that semiconductor devices maintain their nominal parameters when harsh electrical and environmental testing conditions are applied. Some of these conditions can be: high humidity, high current and high temperature that last for a very long time. Many buyers require such difficult test by the producer, as an example, one of the most demanding market is the automotive. This thesis, carried out at the Polytechnic of Turin is focused on the understanding of the mechanism behind some damaging effects caused by the reliability test H3TRB on a commercialized p-n-n+ silicon-based photodiode device produced by Vishay and, in particular, on the passivation layers that causes an increase of the dark current. The introduction chapter is devoted to the presentation of the starting situation, focusing on the description of the structure, the package, the characteristic of the test, the presentation of the problem under study and the different hypotheses and supposed failure that can justify the drift of the dark current. In particular, the company has already found that apparently the increase is related to some defects called bubbles found on the passivation on the device that seems to be lifted by a "seed". The main hypotheses that will be analyzed in this thesis are: thermal stress due to several annealing inside the production line; chemical attack due to possible unwanted residual, water corrosion, due to the separation of H+ and OH- activated by the high bias. The instruments, the reasons behind the decision to perform different analyses and the advantages and disadvantages of the tools used in this work are discussed in chapter number two. In the third and main chapter all the characterizations are chronologically shown as performed, starting from the cleaning procedure that is needed to be sure not to confuse dust with defects, passing through a basic optical inspection and defects classification and positioning, concluding with more advanced tools as scanning electron microscopy and atomic force microscopy that can provide a higher magnification, a compositional characterization and a morphological analysis. All the results are summarized and objectively discussed in chapter four. As we will see the position of the defects plays an important. Differently from the information provided by Vishay no "seed" has been found, however, in its place some holes are present. Apparently, these bubbles seem to form in two separate times, first a nitride hole and then a bubble etch. As a conclusion the last chapter shows which assumptions have been discarded and which are still valid, in addition, some possible solutions to avoid these discovered damages will be presented. Finally, in order to better understand this mechanism some further analyses are proposed; for instance to move away the black mold that partially covers the active area or to check at each production step the quality of the nitride layer.