How can I get 0.1 nm surface roughness on Si surface?

i. Mr. Sun,

I do not think your requirement is realistic, as the size of an atom is in the range of 0.1 nm. In other words, even if you put all the atoms of a material on an exact plane regardless of crystal structure, you will at least get 0.1 nm surface roughness. The least surface roughness can be achieved by means of chemical mechanical polishing in semiconductor industry. This is a very hot area. Good luck.

Ling

i. Ling is correct that a 0.1 nm surface roughness is state of the art in chemical-mechanical polishing. Wafer fabs spend millions for equipment and procedure development for this purpose. An RMS or Ra roughness of 0.3 to 0.5 nm is a typical goal for high-end polishing on silicon for sample analyzed in our lab.

A. To polish silicon finer than .1nm surface roughness I would final polish on a 64 or 73 pitch lap. Use a colloidal silica that won't attack pitch such as Ultra-Sol 500S ND mixed 1/1 with de-ionized water. Keep pH up around 10-11.

Polishing pads don't bridge surface texture as well as a pitch lap will. You should be able to get your surface down to less than 5 angstroms RMS surface roughness using this process.

A. Hi, Michael. Silicon wafers for laser mirrors must be far smoother than mirror polished stainless sheet. 0.667 microinch is 16.9 nanometers.

Regards,

A. This does not answer your question directly. A mechanical surface finish is a direct result of pressure and abrasive. When I say abrasive, particle size and hardness are major factors. If you include lubricants, this could have an effect on how the abrasive performs. So you may want to look for a different size media or abrasive hardness, or you can play with the pressure.

Thanks for the answer, more pressure seemed to do the trick.
Need to run some more tests to be sure, but I'm down to 0,4 0,5 nm at the moment.

A. This thread intrigues me. I know very little about the surface of silicon, but I do know that 0.1 nm is about the diameter of an atom, maybe even less than that, depending on the atom. Are we seriously talking about a surface where a couple of atoms out of place will put the surface out of specification? How do you measure these surfaces? I have used an AFM in my time, but do not consider it to be a piece of production control equipment.
If I am understanding the thread, once again, surface engineering is leading the way in surface sciences - we are now working at one thousandth the official definition of nanotechnology!
Maybe I need a course on silicon surface technology!