Statistical experimental design to optimize RF-sputtered NiTiO3 thin films

Influence of sputtering conditions and annealing parameters on structure and morphology of NiTiO3 ilmenite thin films

“…At 950 °C, NiTiO3 crystallites with (006) oriented planes are formed in addition to the other orientations. At 1200 °C, cristobalite SiO2 (100) plane from oxidation of silicon substrate is detected [34], as also observed by Bellam et al [21]. Therefore, annealing of NiTi films is conducted at temperatures ranging from 700 °C to 900 °C to crystallize NiTiO3 phase without significant modification of silicon substrate.…”

“…In most of the above-mentioned technologies, materials must be designed in the form of thin films to improve their life span under operating conditions or to reduce their dimensions for 3 miniaturization constrains. NiTiO3 thin films were grown by several techniques such as Atomic Layer Deposition [17,18], Pulsed Laser Deposition [19] and by Plasma Sputtering [6,15,20,21]. Among all these techniques, magnetron sputtering is a versatile one, suitable for the deposition of high-quality and well-adhered layers of a wide range of materials at commercially useful scales.…”

“…Recently, Bellam et al have proposed a statistical design of experiments to improve the features of NiTiO3 thin films deposited by a Radio-Frequency magnetron co-sputtering process [21]. This statistical approach provides interesting information to evaluate, with a reduced number of experiments, the influence of large number of processing parameters (9 in their study) on the NiTiO3 properties.…”

Using Data-Driven Learning to Predict and Control the Outcomes of Inorganic Materials Synthesis

“…The basic ideas behind statistical modeling and DoE date back to as early as 1935 . DoE has since demonstrated its utility in industry, including manufacturing and engineering, − and its application to problems in chemistry has seen a rapid increase in the last two decades. , Particularly, there has been a sharp rise in examples of its utility in optimizing materials syntheses in academic literature, including a recent tutorial outlining the use of DoE for materials chemistry. ,,− The increased use of DoE can be attributed to its effectiveness at achieving synthetic control of both dependent and independent variables when used in conjunction with RSM, which is an optimization technique that provides insight throughout a high-dimensional parameter space to formulate a predictive model by using a minimal number of experiments.…”