Dear Mustafa,

You can use the buffered HF to etch the thin aluminum oxide film with good results. For the details of the etching process please follow the paper in the link: http://onlinelibrary.wiley.com/doi/10.1002/ppap.200932103/pdf

Best wishes

I dont quite understand your question. But I use AZ5214 image reversal too. It works well. My process is: first I define pattern of my device in AZ5214 resist. After that I deposited my material (In2O3, Al or Au) after that I lifted off the unwanted deposited layer in acetone. if you process is same like mine, I can maybe give you some tip. How thick is your Al2O3 that you want to remove and which deposition technique do you use?

Best

Thuy

Thank you for replies. Actually i m trying to etch by using H3PO4, it is a good etchant too. The etching time 10 min however, the photoresist is lifted off before i reach desired etch depth that is what i do not understand.Maybe it occurs because of the weak adhesion of the photoresist to the Al2O3 surface. growth tecnique is ALD and thickness 500 nm.

Moreover, Thuy, how do you do image reversal process?

etching Al2O3 with H3PO4 I do not have experience. I've been using AZ5214E on SiO2 substrate and I do not have any problem with adhision so I guess should be the same with Al2O3 as it is also metal oxide. the lifting off of photoresist, could it be because after you've etched the material a bit at the areas between high ground and low ground the photoresist is exposed to H3PO4 so the etching agent attacks the photoresist underneat? I am not quite sure but just a thought.

Yes I use AZ5214E for image resersal but with my process I've only lifted off layer of up to 250 nm. 500 nm is too thick for my process to work. Did you check the photoresist thickness after image reversal and the resist profile?

up to 300 nm i have not faced any problem but after that suddenly photoresist is lifted off. actually another problem is top thickness is 4 um but bottom thickness 15 um.

i check thickness of photoresist which is 1.6-1.7 um at 4000 rpm spinning.

Do you use HMDS during the photolithography process?

Buffered HF is used routinely in device technology to to etch SiO2. You may etch at room temperature and its etch rate is relatively high such that you can etch your layer in much shorter time giving less chance for the etching solution to attack the photoresist layer.

This is one side of the coin.

The question is that your etching time is too long. Do need all that time? How would you detect that your layer is finished? By observing the color at the surface.

To improve the adhesion of the photoresist you have to start with very clean and dry surface of your wafers. Then you have to postbake the specimens at relatively higher temperatures and for longer time to harden the photo resist.

Best wishes

No I do not use HMDS. Actually I did one experiment coating my substrate with HMDS but that sample when I spin coated AZ5214E the photoresist was not evenly distributed on the wafer. Adhesion was severely affected. HMDS adds nonpolar group (R-group) to the substrate surface and the AZ resist has lot of -OH groups so it as basiccally polar. The two dont like each other and very different to get the resist evenly distributed on the substrate.

The ADL deposition technique is at high temperature right? Do you think it has some effect on the photoresist? Because if it does it might mean that thicker deposition layer the effect is more severe as photoresist is in high temp deposition chamber longer..

Also as Abdehalim said it is important that you clean and dry the substrate properly. But since with the same procedure and you didnt have any problem with thinner layer then it might not be the issue with cleaning. It is more to do with the thicker depostion...

After photolithpgrapy process, i measure photoresist thickness of one of the selected structure by using profilometer. then i dip the sample in H3PO4 for 4 min. After i measure same structure for measuring etch depth. i repeat the process for 4 min and 2 min again to achieve the desired etch depth.

i think there is no problem about the cleaning. Standart 3 solvent cleaning and 5 min at 120 for dehydration. Post bake is 10 min at 120 C for photoresist.

By the way, after the deposition i apply photolitography, the depostion temperature is not important for photoresist coating or lithography process.

ok your process is in opposite order to mine. And I dont think it is due to cleaning either because it works well with thinner layer. if you need to overcome this, maybe can think of other resist alternative such as SU8 for the etching mask as I know lots of people use SU8 masks for dry etching that is your features are not too small. SU8 is more chemically resistant and more mechanically strong as well. But if you want to lift it up after etching then will have to add a layer of omnicoat for easy lifting off. Otherwise after exposure, only O2 plasma or piranhan can remove the resist.