Joint evaluation of performance and environmental impacts of a microfabricated device : VO2-based device

Details

Supervisors

Raskin, Jean-Pierre ; Bidoul, Noémie ; Pip, Alex

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil électricien, à finalité spécialisée

Abstract

VO2-based devices are particularly intriguing due to their capacity to undergo a phase change from insulator to metal in response to temperature variations. The potential applications for such devices are extensive and involves switches, memories, photodetectors, smart windows, sensors, oscillators etc. In this work, a cradle-to-gate approach was employed to evaluate the environmental impact of the microfabrication of the VO2-based device, from the extraction of raw materials to the exit of the WINFAB laboratory. The objective was to identify the hotspots and propose energy-efficient alternatives, while also assessing the impact of these alternatives on device performance. The process involved a standard cleaning step, followed by wet thermal oxidation to create a SiO2 layer. Subsequently, a VO2 sputtering step and VO2 annealing step were conducted to trigger crystallization. This was followed by lithography, metallization, and lift-off. The Primary Energy Demand (PED) in [MJ/cm²] was chosen as the indicator for the Life Cycle Assessment (LCA). Several assumptions were made to obtain conclusive results, with one of the major assumptions being the utilization of Boyd’s work as source for the Energy Intensity (EI) values, except for the metals where the EduPack software was used. Real-time power consumption measurements revealed that the equipment consumed less power than stated in the datasheet, which was also taken into consideration and slightly reduced the final result. The overall PED of the process was determined to be 1.37 · 102 [MJ/cm²]. The most significant impact was observed during the metallization step, primarily due to gold deposition and its associated EI, representing 99.27% of the total PED. The methodology used in this work and the results obtained were discussed with the research laboratory, CEA-Leti. An alternative approach proposed replacing the gold metal layer with copper, resulting in a 99.25% reduction in the total PED to 1.03 [MJ/cm²]. In this alternative process, the VO2 sputtering step has the highest impact due to the EI of vanadium. Following the reduction in impact, a performance analysis was conducted, considering the Rcontact and the Rins/Rmet ratio for both gold and copper evaporation. The use of copper evaporation exhibited a lower Rcontact, indicating smoother current flow across the junction. Moreover, the greater Rins/Rmet ratio for copper evaporation resulted in a larger negative differential resistance (NDR) zone, representing the wanted region for oscillations, that we desire maximum. A proposed virtual process suggests substituting the gold deposit with aluminum, along with the utilization of different equipment, and replacing wet thermal oxidation with PECVD. The implementation of this virtual process results in a total PED of 7.07 · 10−1 [MJ/cm²], leading to a remarkable reduction in the overall impact by 99.48%. The biggest challenge faced during this work revolves around obtaining the necessary data for EI. This factor significantly impacts the ultimate outcome. Collaboration between industry, researchers, and other stakeholders is necessary to facilitate the sharing of such information.