Different colours are due to difference in coating thickness. As such these are interference colours. Please refer my following paper for details

A. K SHARMA, Anodizing titanium for space applications, Thin SOlid Films, Vol. 208 (1992) 48-54.

Dear Claudia Rosa Morales Vargas,

Two common possibilities are:

1) small variation in composition of the film.

2) variation in thickness of the film: if your film is transparent, then you can have optical interference. Variation in deposition time will shift the position of the colours.

Dear Claudia,

a cilindrical cathode implies different distance/Ohmic loss among the edges and the midlle of the plate and the cathode. This produces non-homogeneous current distribution over the plate surface giving different oxide thickness/colour.

Try to use a flat cathode parallel to the plate with similar size and shape; mask  the edges of the plate to avoid current concentration.

Shortly spoken: everything what might have influence on your oxide thickness. Do you get a continuous change it is perhaps related to the distance as written before already. If it is not continuous you should think about a discontinuos reaction caused by the sample surface prep (reactivity changes), a incontinuos reaction by a varying field density, or no respectively too slow stirring up of your solution. 

Thanks you all for your answers. I was thinking about the current distribution as the problem. But as I've been working with that cathode before, I wasn't so sure about it. I was not stirring the solution either.

Dear Claudia,

stirring is useful, but it is not the core of your problem. In my experience your layer, and the relevant colour, should form just during the first seconds of polarization (potentiostatic, I suppose), after which the current goes down due to the hindrance of the formed oxide.

The colour you obtain is mainly due to the current peak in the first seconds, more than to evolving gas or limiting diffusion layer. In other words, you are obtaining a sort of coloured fingerprint of  the peak-current density distribution along the surface.

So, in my opinion, the core of your problem lies in the cell geometry: a uniform current distribution produces a uniform colouring.

Different colours are due to difference in coating thickness. As such these are interference colours. Please refer my following paper for details

A. K SHARMA, Anodizing titanium for space applications, Thin SOlid Films, Vol. 208 (1992) 48-54.

Thanks you all for your answers. So now I know everything is about current distribution. I will update you with my results.

Claudia. in my experience the color of the titanium oxide layer depends on each process factors (voltage, current, cell geometry , concentration of the solution among others). For your case, I think the problem is in the cathode size, to grow a layer in a larger area you need a cathode sufficient large to generate a homogeneous distribution of charge. particularly soil using a cathode with minimum area 3 times the area of the anode.

I agree with AK Sharma and Anderson as well. While color difference is generally due to film thickness  ( interference patterns), prior history of the substrate before anodization will have an impact too.  If uneven thickness in the oxide layer exists prior to anodization, it will contribute to changed color patterns.  Also how the crystalline patterns are seen on the surface will make a difference.  In Al anodization, a spangled finish is obtained by annealing and quenching the sample  under different conditions. If such an  anodized Aluminium is dyed, it will show various colors. Thus a lot of factors may be involved.

I think the color pattern is due to the orientation of the electrodes.  You have a flat plate working electrode and a cylindrical counter electrode.  Current densities on the working electrode will vary with the distance between the counter and working electrode.  Larger the distance between the working and the counter lower will be the current density ( increased electrolyte resistance).  The oxide layer grown will therefore vary in thickness based on the current density. When thickness of the oxide layer varies you will get color variation.  You can add a neutral electrolyte  ( such as sodium phosphate etc) to increase the overall conductivity of the electrolyte.  The small variation in the resistance due to distances will be minimized by this method . Try it and find out.

The color pattern is due to the coatigs thickness of the oxide. The thickness of the layer grown is due to the applied voltage between the counter electrode and the substrate. The electrolyte composition used is not very important, can use H2SO4 (5 g/L), H3PO4 (10 g/L), etc. for view the different colours.

Dear Claudia,

I agree the other researchers that different interference colours of thin transparent oxide films can be ascribed to different oxide thickness. It is always a challenge in process upscaling to remain a uniform current distribution. You will find more information about improving the current distrbution in literature about industrial anodizing or industrial electrodeposition. Size and arrangement of the electrodes might be a parameter as well as additional electrodes or apertures.

Hi, I agree about thin film interference, on top of that the deposited phase may not be stechiometric, so you might have colors stemming from color centers defects. Check if there is any TiO that generally possesses a blue color