No abstract

The recent discovery of ferroelectricity in thin film HfO2 materials renewed the interest in ferroelectric FET (FeFET) as an emerging nonvolatile memory providing a potential high speed and low power Flash alternative. Here, we report more insight into FeFET performance by integrating two types of ferroelectric (FE) materials and varying their properties. By varying the material type [HfO2 (HSO) versus hafnium zirconium oxide (HZO)], optimum content (Si doping/mixture ratio), and film thickness, a material relation to FeFET device physics is concluded. As for the material type, an improved FeFET performance is observed for HZO integration with memory window (MW) comparable to theoretical values. For different Si contents, the HSO based FeFET exhibited a MW trend with different stabilized phases. Similarly, the HZO FeFET shows MW dependence on the Hf:Zr mixture ratio. A maximized MW is obtained with cycle ratios of 16:1 (HfO2:Si) and 1:1 (Hf:Zr) as measured on HSO and HZO based FeFETs, respectively. The thickness variation shows a trend of increasing MW with the increased FE layer thickness confirming early theoretical predictions. The FeFET material aspects and stack physics are discussed with insight into the interplay factors, while optimum FE material parameters are outlined in relation to performance.

No abstract

No abstract

Nanostructure-initiator mass spectrometry (NIMS) is a laser desorption/ionization analysis technique based on the vaporization of a nanostructure-trapped liquid "initiator" phase. Here we report an intriguing relationship between NIMS surface morphology and analyte selectivity. Scanning electron microscopy and spectroscopic ellipsometry were used to characterize the surface morphologies of a series of NIMS substrates generated by anodic electrochemical etching. Mass spectrometry imaging was applied to compare NIMS sensitivity of these various surfaces toward the analysis of diverse analytes. The porosity of NIMS surfaces was found to increase linearly with etching time where the pore size ranged from 4 to 12 nm with corresponding porosities estimated to be 7-70%. Surface morphology was found to significantly and selectively alter NIMS sensitivity. The small molecule (<2k Da) sensitivity was found to increase with increased porosity, whereas low porosity had the highest sensitivity for the largest molecules examined. Estimation of molecular sizes showed that this transition occurs when the pore size is <3× the maximum of molecular dimensions. While the origins of selectivity are unclear, increased signal from small molecules with increased surface area is consistent with a surface area restructuring-driven desorption/ionization process where signal intensity increases with porosity. In contrast, large molecules show highest signal for the low-porosity and small-pore-size surfaces. We attribute this to strong interactions between the initiator-coated pore structures and large molecules that hinder desorption/ionization by trapping large molecules. This finding may enable us to design NIMS surfaces with increased specificity to molecules of interest.