I usually use very thin Cr layer (upto 5 nm) in organic transistor to improve the adhesion. It improve the device performance . Using thin Cr or Ti layer by co-deposition is easy and best way to improve the adhesion. There are some other methods like O 2 plasma treatment but that need very high temperature.i think using thin Cr or Ti layer is better.

Thanks, I will try both Cr and Ti and characterize the two samples to determine which works better for my intended application.

I think that Ti is better, and i advice you to keep far away from any technique using mid or hi temperature, because it will oxidize your Ag layer.

Heating your sample to 200-250C during deposition may help.

Thanks Beihai. We thought of that, but we do not have the capabilities since we are relying on the Frederick e-beam deposition system. Annealing the film after may also results in it cracking. Is there a way of doing this using our system?

I'm not familiar with your system configurations. You may consider of adding a halogen lamp inside the chamber to heat your substrate. Or, you may also try annealing your sample under vacuum.

Thanks, I will explore all possibilities.

You can try to clean the sample with ion beam before Ag coating if your system has a ion source..

Dear Jephias, what (vacuum) conditions do you work under and how do you hydrogenate and otherwise treat the film before Ag growth?

Quite a few years ago (ca. 15 yrs ?) I remember to have seen a talk where STM people looked at the growth of Ag on Si substrates (in the group of Güntherrodt (Basel, Switzerland)). The results were not terribly good.

Jin-Cherng, unfortunately the system does not have an ion source. We however have a separate ICP-RIE, argon and O2 systems for etching. Can I use any of these and the load my sample into the e-beam system for deposition?

Kai, we grow the a-Si:H film on glass substrate normally using the standard PECVD process. The vacuum in the Frederick for the e-beam deposition is in the 1.0x10^-7 Torr. I am thinking of putting a very thin base of ITO first and find if it affects the optical properties of the overall device.

My suggestion would be the addition of an adhesion layer. Ti/Ni/Ag & Cr/Ni/Ag are popular in Semiconducor backside contacts. The Ti or Cr for adhesion and the Ni for a barier since the Ti and Cr will create silicides and be consumed over time.

I used to produce very sturdy Au films adsorbed on glass by using an extremely thin Cr interlayer (of the order of 1 or 2 nm). Since we were not using an UHV system, Cr was most likely present as an oxide. It is unlikely that such a thin oxide film may alter the optical properties of you sample (except, possibly, in the UV, slightly above the Ag plasmon energy).

You may try to used silane self-assembly technique to enhance the adhesion, sometimes, its works really well due to improved kinetic of depostion

@ Vladimir. I m not sure if this can work since I have already grown the a-Si:H film on all my samples. I only need to add an Ag layer of nano-structures. I think what you are suggesting works if I decide to make new samples from scratch.

Jephias, I mean to use silane self-assembly to serve as adhesion layer between a-Si:H film and Ag. silane monolayer should be 20 A thick

I get you now. I will;l explore that option and see how feasible it is. can you handle the process in your lab? Is the SiH4 adhesive in liquid state ? I know SiH4 is a gas at room temperature and highly explosive. I do not want to deal with that problem.

Silanes in organic chemistry are mostly liquid compound, e.g.: 3-metoxy-amino-silane (Aldrich, Gelest etc) or any others. You could prepare 3-mM solution of this compound in solvent and left it for overnight self-assembly in air-free condition (e.g.: Shlenk line). Once you came in the morning, wash it 3 times in the same solvent which you use to prepare self-assembled monolayer and rush to deposit Ag layer on top ASAP

Sounds like a more practical solution than I initially thought. I will definitely explore that option. Thanks Vladimir.

Jephias, you are wellcome, please let me know if you will require more details.

I will definitely do that. thanks once again.

You can also try depositing a thin layer (0.5-0.8 nm) of silicon oxide on Si or glass substrates for better adhesion of metal on the same. Si oxides can be easily deposited by e beam technique.

@Chaitanya. Generally there is always a native oxide layer on Si and in most cases this layer is 1 nm or thicker. Are you suggesting that I need to do the HF cleaning process first to remove the native oxide? This will add a lot of process steps and hence costs which I am trying to minimize. Please advise me if I am wrong.

Dear Jephias, it might make a difference whether minimizing cost is a requirement all along your activity or rather a goal for the "finalized version" of the procedure you're trying to establish. It could make sense to first establish a valid - maybe not so cost effective - solution and then to work on reducing cost by understanding the requirements.

@ Kai. That makes sense. However, I have been depositing Ag on native oxide and the results were not really excellent. I was now thinking of trying silane.

There may be differences between native oxide and the one that is prepared on cleaned

substrates through controlled deposition. But it is upto your requirement which options to choose from.