If you're looking for a quick-and-dirty method, then dipping your sample in DI H2O is a good way to check if you have oxides on crystalline surfaces. Most crystalline semiconductors I've worked with (Si, GaAs etc) are hydrophobic surfaces due to large surface free-energy. If you drop a few drops of water on the surface, they will form beads, which will quickly run off the surface if you tilt the sample. This is called de-wetting.

If you have any oxide on the surface (even as thin as a native oxide), the water will start to form a more "relaxed film" instead. It will also not easily run off the surface when you tilt the sample. This is because oxides (and a lot of amorphous films) have a much smaller surface free-energy, and hence are hydrophilic.

If you are worried about surface oxides before metal deposition, you should always do a quick native oxide etch before any evaporation step. For Si surfaces, dip the sample for 10 seconds in dilute HF (1:40 H2O), and for most III-V, use dilute HCl. If the native oxide has been successfully removed from the surface, the samples will quickly dewet as you pull them out of the water bath. Immediately load them in the metal chamber and pump it down. It takes only a few minutes in an oxygenated environment for pristine crystalline surfaces to oxidize.

Hope this helps.

Thanks for the great suggestions everyone. The water-drop method was unknown to me and sounds really useful, if a bit qualitative... I think once it has been confirmed independently using one of the other methods it could be a winner. 

We already do a HF (BOE) dip as standard, my concern is that the samples could begin oxidising immediately and a native oxide layer could exist after the short time between surface cleaning and loading for evaporation (mostly a concern for high Al-content samples). If possible I will run trials using XPS for samples left for different time intervals after HF-based cleaning, cross-checking with the water drop method, and post back any interesting results.

Cheers!

Hi Matthew

I also use the water drop method, and maybe this can help, too.

When you bake your Si wafer on the hot plate at higher than 100 degree C (may be 120 C) for several minutes, the surface can be changed to hydrophobic, which means there is less oxidation on the surface.

Vice versa, you can make your Si surface oxygen-rich with ozone treatment, which will make the surface hydrophilic, with oxygen dangling bonds.

These are the simple way to try, I guess.

Young

Dear Matthew,

I) I would suggest you to do Cross sectional TEM.  You will see a contrast between the layers (If there is any oxide layer on the substrate). However for this one you should cut the substrate! 

................................................................................................................................

II) If you are looking for a very easy way just to recognize if the oxide layer is created on the surface or not( NOT for publication just to know before the actual verification)! Then I will suggest you to use visualization. You can  shine a light with a known wavelength (e.g. Withe light) on the surface of your substrate and then see what is the emitted color( just by tilting the sample and looking through the sample surface). If you know the material then  the Band-Gap(electrical or optical)is known you can calculate the absorption spectrum by using equation bellow;

 E=[hc/(lambda)]=1.24/wavelength(nm)  

By calculation you will know what would be the emitted color from the bare semiconductor. Finally, if you saw any different color then you can say Ok it will be based on the Oxidization(As you know Oxidization will increase the Band gab of the material and it means absorption at smaller wavelength will be expected! So a blue shift will occur in absorption/emission spectrum). As a result, again from the emitted color you can calculate the band gap of the oxide layer and then compare it with the theoretical band gap of the oxide! 

Simply doing a EDS analysis with your sample on SEM also does work.

In my opinion the best suited answer is that of Trivedi. Just to clarify, the de-wetting test, easy and quick. is NORMALLY used in Semicond fabs to check if an oxide etch from a Si wafer has been completed.