I have been depositing FeGa (magnetic material with magnetostriction close to 400 ppm (bulk value)) and trying to make thin films on top of Silicon wafer. The composition of my target is Fe 81% and Ga 19% (atomic percentage). Mostly I got the Fe-Si compound which matches the xrd data (3 peaks, 110, 200, 211)
But I want to have the FeGa compound and since Ga volume is low in the target, eventually the sputtered composition has low Ga percentage. So it seems to me that Fe tends to form Fe-Si than FeGa in the film. I am using dc bias of 30, 45, 60, 100, 200 watt power with a rf bias at the substrate (5 to 10 watt) with 3.3 militorr.
1. Is there any technique I can apply to increase the adhesion of Ga with Fe?
2. In principle, other than power, what can I do (like deposit something that will improve adhesion). Ga is liquid in room temperature.
3. Is any volatile nature of Ga causing anything not to form bonds with Fe (my target is solid by the way) . If I increase the deposition pressure in the sputtering chamber, will it help?
4. Does Cu and Ti layer increase adhesion for this type of material?
Can anyone help with any ideas to sort this problem out , at least I can try different idea to see any improvement?
5. Will annealing improve the result or very low melting point of Ga will case the adhesion weaker?
I hope if you make the alloy of Fe-Ga with respective atomic percentages (81%Fe and 19% Ga) in Vacuum arc melting furnace as button sample(Try to arc melt 4 to 5 times to get homogeneous alloy sample). Then use this Fe-Ga alloy sample as a sputtering target for thin film deposition. I hope this technique may help you.
Use multi target chamber to deposit alloy film; As you pointed out your problem that the Ga has lower percentage in the single alloy (FeGa) target. I suggest you that arrange two different target of Fe and Ga if possible. And optimize the sputtering rate or deposition rate for both the target at different powers for different deposition time. In this way you adjust the power and time of deposition for both target. So you may able to deposit the alloy film.
Hasnain, Your first problem seems to be the chemical reaction and interdiffusion of the depositing Fe with the silicon substrate. What temperature is the substrate? I was not thinking Fe would diffuse into silicon so much at room temperature.
Maybe you should try and prepare a silicon substrate with a pre-sputtered layer that will be a chemical reaction and diffusion barrier? Gold? SiO2? Platinum? I do not have a great idea, just now. But if silicon oxide is stable against Fe vapor "attack", then simply oxidizing the wafer in very hot steam or oxygen plasma may be easy for some labs. Or you could just buy silicon wafers with so-called "thermal oxide" layer already grown on them, usually 500 nm thick.
If you like the diffusion barrier idea, you can probably do a literature search and find what other people have used successfully.
Thanks a lot. I can buy the wafer with oxide. Let's try that. I did have FeSi phase due to diffusion since the native oxide layer was too thin to prevent it. Let's hope it makes FeGa phase for thin film
Hatem, you and Rughopatruni may have a point, but I for one agree with Sajid!!
If you have the equipment to do co-sputtering, then co-sputtering is far-and-away the best way to make these FeGa films. One can control the composition at the turn of a knob. A FeGa target is troublesome in its phase equilibria (or metastable states) and microstructure.
But as Hasnain well knows, Ga is liquid at 31°C, and a sputter target will surely get hotter than that. So a pure Ga sputter target may seem like a problem. But sputtering from liquid gallium is easy, except that the target must be facing gravitationally upward to contain the liquid metal. I have done it. Others have too. The vapor pressure of liquid gallium at < 100°C is low (like < 1E10 Torr), so this is not a problem.
What is the mechanism behind liquid metal sputtering. I have never done it. So is it the same principle by which solid metals get sputtered? Is the liquid metal sputtered more quickly than solid solution?
Hasnain, The mechanism of sputtering at a liquid-vacuum interface is the same as at a solid-vacuum interface: Momentum transfer during atomic collisions. I did not perform experiments comparing sputter (or deposition) rate off of solid versus liquid Ga target. I would intuitively assume that the liquid sputters somewhat faster. The reason would be, if it were true, that the binding energy of the Ga atoms to each other would be less in the liquid than in the solid. (Else, where would the heat of solidification arise?) This might be a small difference in Ga compared with other metals. The sputter rate of liquid gallium is not extremely high. There are solid metals that sputter faster.
One more way to make alloy is to depositing bilayer and anneal at variour range of the temperatures. So that may reach at particular phase you want to get. It may be useful if you have the single target sputtering set-up.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction