In the center you probably have redeposition of another composition and density, giving another color.

Best

Paul

Carbon contamination

Of course it would be due to the redeposition of some components. It depends if you sputtered in Ar, N2 or O2 atmosphere. There could be some oxide, probably Ti, or some nitride, on the terget!

Good Luck!

P

It is happening for variety of materials due to surface phase transformations, recystallization and roughness changes.

The different yielding coefficient can't explain the dark and gold area, otherwise you should see color changes depending on the etching of the target. Are you sure of the quality of the vacuum, it may be due to an oxidation or nitriding the surface of the target. Probably you have a magnetron target.

The rings shape comes from the sputtering area delimited by the magnetic field. The iridescence from different thicknesses of the redeposited materials. I dont know how different the sputtering yields of the individual components are, but I assume if you are depositing PbZrTiO3 the composition of the target already considers it. As far as you are probably depositing with Ar/O2 atmosphere you are probably redepositing the PbZrTiO3 itself that I believe is white/yellowish or grayish... About the composition of your films, it would probably change until you reach an equilibrium in the target surface compostion as I guess you already have. If you do a long presputtering with Ar only, you probably recover in some extent the color and the pristine target composition but then you will have to condition your system again in order to have constant composition in the films... Hope it is useful.

In my opinion, while a redeposition over the target (in case of r.f. sputtering) is possible, inhomogeneous sputtering from the target is also possible (I am talking about the inhomogeneous color) due to different sputtering yields of the individual components, especially if the plasma is not uniform.

As mentioned above this effect is due to the lines of the magnetic field affecting the plasma you won't get a smooth target after few cycles of deposition. Second for the colors one problem I suspect is thermal when ventillating rapidly after deposition. I wonder if it had some effects on your calibrated deposition parameters. Wish you luck

Since you are depositing metal oxides, I assume you are doing reactive sputtering in an oxygen environment. You probably have what is called target poisoning. There is lots of information available on the web for dealing with that. You will have to vary the conditions to control the target condition. you may need a more advanced power supply.

As mentioned above three phenomenons are present : flow of Ar+/O2 ; magnetic field lines and theating process all are included to reach what observeed on your samples.

You must change the three parameters one by one to dilimit the effect of each of them.It will be a good paper for you to be published.

I like to think that this is target material quality related and sputtering rate uniformity. This could cause segregation of some of the metals that make up the target. It would be interesting to do Auger or XPS study of the surface composition changes. The vacuum plays major role as well. You could be reacting some of the metal with the residual gas in the system.

For several months during my graduate school days I was making some interesting plasma created thin polymer films. They were very well correlated to the efficiency of my liquid nitrogen trap.

I assume that you are using an Argon atmosphere for sputtering. Does the color of the rings depend on the thickness of the sputtered gold? Also, at what thickness of gold do the rings start to appear? The reason I mention gold thickness is because I wonder if it's possible to deposit just enough gold for the electrode conduction needed but before the ring structure begins to appear.

I very much doubt that the rings are produced by composition variations as I can't think of any reason for a ringed composition variation, caused by crystallites with differing orientations that would produce such a large "pink" area in your photo. I'm assuming this is a polycrystalline sample of PZT.

The dark and gold ring are formed by redeposition of sputtered materials and it's amorphous phase in case of pzt. If you maintain same deposition condition, it's not critical.for the film quallity. However, if you want to change working pressure and gas ratio, you should carry out long presputtering becaue it changes the sputtered area of pzt target.

As mentioned several times before The colour difference seems to be related most probably to The higher abrasion speed induced by The magnetron. Considering problems with thin Film composition i would be much More concerned about process parameters as e.g. Temperature, which habe a very large effect due to Pb reevaporation...

It's so grateful to receive all your ideas and advise.

I think I should supplement more information. My sputtering is RF magnenron sputtering, in pure Ar environment. I usually make vacuum under 5e-6Torr (6.65e-4Pa). However, Mr. Khalifa Aguir mentioned about vacuum quality, I have to say I am not 100% confident about it as it is an old machine.

I can make a summary for all the reasonable speculation from all of you

The dark and gold color.

1. the ring related with material elimination which constrained by magnetic field.

2. Both dark and gold area / the gold and gray color area / the center gold area caused by redeposition.

3. surface oxidation or nitridation.

4. Carbon contamination.

5. surface phase transformations, recystallization.

6. It is due to target material quality, and segregation of some composition.

7. target poisoning phenomenon. target condition and power should be improved.

8. thermal effect.

advises on relation between surface change and film quality are in accordance:

for long time sputtering, the surface change does not affect the film quality, but if the sputtering condition changed, long time pre-sputtering is required to make it reach a new equivalent state.

I believe this tip to be very convincing!

I am thinking about a good way to prove the speculation. From now on, I should carefully gather the information along my experiment.

my own opinion about it is;

1. redeposition(I think most possible reason for gold area in center, but for the gold ring, it's weird to keep such a perfect geometry)

2. Different yielding coefficient of composition of target. But few people think so--!. I think since the magnetic field constrains the plasma, above the dark and gold ring, the magentic filed may formed 2 strong annular stripe, so Ar+ impact on target surface with 2 different energy, then on the surface different particles are emitted. Even so, eventually, the sputtered out composition will reach a equilibrium to be the same with target.

3. Oxidation and nitridation seems possible, but I still trust my machine to be free of vacuum problem.^ ^

I had obtained the same phenomenon using others targets and the Ar+ ions ,

one paper is published on IOP Journal conf.serie in 2012.anothetr paper is now prepared in the same way :

Ion beam irradiation is an efficiency technique for phase formation and material modification as a non-equilibrium state. At low energy (300 eV), the Ar+ ions bombardment lead to the formation of small nanodots on the InX (X=Sb,P,N) surface compounds. It is interesting in understanding the effect of the argon ion bombardment on the surface topography at normal incidence.Furthermore, reactive sputtering of InX films in oxygen indicated no evidence of direct formation of In2O3. It was suggested therefore that oxygen accelerates the formation of an intermediate amorphous indium oxidation process phase inside the InX matrix, which eventually oxidizes InX completely to form In2O3.

@Alina Manole, may I ask you about pre-sputtering? Usually, how long is your preference pre-sputtering time?

I know pre-sputtering times from 0,2 to 10 hours.

for getting a feeling of the situation I look at the power- time- plot at constant voltage. If the power increased some percent, there could be removed some percent oxide/nitride.. on the target. (rising temperature often also increases the power). For keeping oxide-layers thin one should avoid venting a hot target to air.

oxidation effect due to sputtering is evident in all material surfaces as it was experimentaly verified.The duree of sputtering and the heating proceess are the two parameters to be controlled to obtain such effect or to avoid it.

I only came across this chat thread now. But I am experiencing the same problem with In2O3 target. After sputtering there is a dark ring on the target where ion bombardment happens and I did sputtering in Ar.

I did XPS on the deposited film and found 30% carbon. So it's carbon contamination yes. And it happened with other target as well ( Au and Cr).

And it is also true that when sputtered in Ar/O2 true colour of the target reappear. We are still struggling to find the source of C contamination. Is vacuum of 1x10E-6 mbar not enough?

Thuy