As the title states, I partially etched some different material (Wafer 2238, 100 nm NPN) to create nanofeatures with blown out holes. With a thicker membrane, we can blow out the pores more (etch longer) without compromising the strength of the material (with regard to templating purposes). These membranes were fractured and coated with 6 nm of platinum. These images are only the NPN; not coated with gold or templated with MgF2.
Equivalent pore diameter histograms for partially etched substrates.
Minor Axis histograms of partial etch (0, 20, 40, 60, 80s) of Wafer 2238 (from top to bottom). Average length: 42, 48, 53, 56, 72 nm. Based on estimates of etch rate, the thicknesses of these membranes are (100, 80, 60 40, 20 nm, from top to bottom). Pores merge considerably below 40nm.
Zoom-in of wafer 2238. The scaly structures are residual pieces (siloxynitride?) used as a mask to create the thicker NPN materials. there also appears to be some conical shape to the pores, instead of anisotropic sidewalls.
Viewed off-normal, the thickness of the siloxynitride is appreciable.
40s etch crossection, 64 nm thick. (vibration hinders some resolution). The pores have a conical crossection.
80s etch aerial image. Residue pattern from the silicoxynitride is transferred through the nanomembrane’s volume.80s etch crossection. 23nm thick. The silicoxynitride acts as an etch mask (with poor selectivity) and is causing the texture at the surface.80s crossection. Shows the scalloped merged pores with sharp peaks.
I think the 2238-80s material is too weak to hold up to MgF2 processing, but there is hope for the 60s material.
The graph below is air permeabilities for four samples from wafers 344,338,342,345 without and with 15 nm thick deposited oxide on the front side. All of this samples don’t have pinholes.Look at the data for samples 2 and 4 on the graph. The approximate porosities of these samples are 2.7 and 4.6% which is the…
Mike and I tried to pass 3 um beads mixed with rhodamine through the stir cell to prove that, if the membrane we intact, only the rhodamine would pass through and not the beads. As seen in the pictures below, only the rhodamine passed through, not the nanoparticles, since these particles are 100x larger than…
This past year I have been collecting burst pressure data on pnc-Si and NPN. The data for 30nm pnc-Si came from my previous post, “Wafer 1059: Burst Pressure Analysis”, which also includes a wafer map. This wafer had 7% porosity and an average pore diameter of 20nm. I tested square window sizes with side lengths of 0.1mm,…
Due to the discrepancies between the 1D model and experimental runs with rhodamine, I have switched over to using COMSOL to solve for a 3D model. In this system, 200×200 um membrane squares with a 45 degree etch separate the two wells. The top well is set up with .5 mol/m3 of rhodamine, and the bottom…
Today I deposited two 30 nm thick Si films and annealed one at 1000 C for 1 min and the other at 1000 C for 5 min. My running hypothesis is that performing an Ar sputter after the silicon deposition creates “broken” bonds that seed crystallization. Therefore, the crystallization should nucleate at the interface and…
UPDATE: Some data presented here is out of date. For up-to-date data, see my next post here. — Lately I’ve been performing comparison experiments between our NPN membranes (wafer 1132, 3 slot SEPCON, 40-48nm pores and 13-20% porosity) and commercial cellulose triacetate filtration membranes, with respect to the degree to which they hinder the diffusion…
