The Cambridge Microsystems Savannah 200 atomic layer deposition (ALD) system uses a two-step chemical process that alternates pulses of water vapor and trimethylaluminum (TMA) to build a layer of Al2O3 on a silicon substate. Because of the two-step nature of the process, the film is forced to grow by successive monolayers, each with a thickness…
Karl and I have been working to measure separations using SEPCONs with particles ranging in size from 20 nm to 1 µm. We started out naively enough by slightly diluting some flourospheres we found in the refrigerator, putting them in the TECAN’s nanoquant plate and pressing run. I know. Ha ha. The first point we should…
Using a 6″ wafer, we start with 50 nm of RV SiN, then deposit 40 nm a-Si followed by 20 nm of SiO2. RTP of 1050 C, 100 C/s, 60s. JP patterned the wafer to make standard SEPCON chips. The pores were transferred to the SiN using RIE. The RIE etches faster at the perimeter…
We created a 25-40-25 OSO stack with a SEPCON pattern. RTP at 1050 C 100 C/s 60 s. The following images show a comparison between the STEM detector and the SEM detector, and then a couple more images at 100 kx and 50 kx. I calculated a porosity of ~6.5 %, and…
Below are some SEM images of a burst 30-50-30 NSN membrane in a SEPCON pattern. Many similar chips failed at lower pressures, but this one survived to ~5-7 psi. (didn’t keep close track of the chips after they were burst, so I’m not sure which pressure this burst at, but it was within the 5-7…
Still trying to reproduce Joe’s large pore material on SEPCONs using NSN stacks. This post has data from two wafers with 30/50/30 and 30/100/30 NSN. RTP is 1000C 50C/s for 1min. Nitride is etched off by placing samples in 50% HF for 3.5 minutes. (~30s after that the films start to lift off) I start…
We are trying to reproduce the very small pores Joe created with 30/100/30 OSO and large pores created with 30/100/30 NSN using a SEPCON pattern. (Joe’s previous post.) We had a problem with several chips releasing from the center of the wafer prior to completely clearing all the features. In the future we will use…
What we’d like to have accompany the Sepcon filters is a general way to tune the pore sizes for any given separation. If, for instance, someone wanted to have a filter cutoff of 12nm, we could create a custom wafer with a mean pore size of roughly 17nm. But it would be much easier (and…