I am looking for a way to make holes, close to 500 micron in diameter, through a 4" Silicon(100) wafer (525 micron thick) for microfluidic applications and was wondering if metal-assisted chemical etching (MACE) or magnetically-guided MACE could be applied for this purpose using HF/H2O2 etching mixture. I am also looking into applying deep reactive ion etching or KOH etching for this purpose, but that is a different subject.
Looking through the literature, the main focus of HF/H2O2 based MACE research seems to be on obtaining micro- or nano-structures (such as gratings) with rather high aspect ratios and low wall-roughness etc., and I realise that this is most likely a delicate process of optimising different parameters such as etchant concentrations, temperature, thickness and morphology of catalysers etc. There seems to some discrepancy on what concentrations of HF and H2O2 give the best results depending on if magnetic guiding is applied or not, and on the dimension of structure to be obtained. In some cases [HF]>[H2O2] seems to be applied while in other cases [HF]
The metal assisted etching process is very slow (typically 0.1 um/ min)....you can try KOH etching with oxide or nitride mask... alternatively you can also try HF HNO3 Acetic acid mixture with Cr-Au metal mask
Deep etch ideas here...
https://sst.semiconductor-digest.com/2005/04/through-wafer-via-etching/
https://sst.semiconductor-digest.com/2006/03/si-drie-for-through-wafer-via-fabrication/
If I were trying this I would not try metal assisted, I would focus on RIE. You was to dig a hole... I would use RIE over wet etch in this case. You seem to want directional control... RIE is best overall approach..
Dear Bjödn,
Few things we should keep in mind while MACE:
1. Try Pulsed DC Magnetron sputtering for the better adhsion of Au with Si.
2. Deposited Au film should be porous. For this, deposit the Au film with different deposition rates (A/s).
3. After Lithography or just before the Au deposition use the BHF solution for removing the native oxides from the Si surface.
4. Lift off process may cause the high roughness and lateral etching of the vertical walls.
5. Avoid seed layer for the adhesion (Cr or Ti forms the oxides which are the culprits for MACE). Since, Direct contact between metal catalyst and Si is required for the MACE process.
6. Variation in the HF/H2O2 concentations leads to the grass formation and porous defects on the surface. These values need to be optimized.
I hope this will help you.
Best regards,
Varun
Varun Pratap Sharma: Point 5: It seems to me that this point objects a lot of published research which said that adhesion layer helped improving the MaCE process. Could you please show me a (some) proof for this point? Thank you alot.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Polymers