The question reminded me of a fact that even a large single crystalline TIO2 gets deep blue after annealing in vacuo. The bluish color in this case is usually ascribed to oxygen vacancies working as color centers.  The color of nanotubes may be due to a similar reason.

which solution you have used? organic medium or aqueous medium. normally the powder starts falling down when you used perchloric acid or similar electrolytes.

refer

 http://www.sciencedirect.com/science/article/pii/S0022459611000211

http://pubs.acs.org/doi/abs/10.1021/cm900410x

TiO2 nanotube formation is a tricky business (much more tricky than you might deduct from literature), and there are lots of parameters that will influence the formation process as well as the result. What exactly do you mean with "white"? I saw hundreds of anodized samples, but none of them appeared white. Blue or purple color is quite typical for thin oxide films.

I agree with Cluas Moseke. I too never observed bluish or purple color for the anodized samples. The color can be observed during the first 1 or 2 minutes after the initiation of anodization procedure 

Claus Moseke, I thought the color of TiO2 nanotube arrays should be white like TiO2 powder. Do you know why the color is not white. Is there any  scientific reason?  Does the thickness play any role in the color of the formed  thin film or the amorphous structure of TiO2 nanotube arrays or anything else?

I can help to explain the reason for the color of this TiO2 nanotube layer.

when white light (such as environmental light) illuminates a thin film, a series of constructive and destructive interferences arise from the phase difference of the light beams reflecting at the top and at the bottom of the thin film. The position of this is directly dependent on the refractive index of the thin film (n) and the thickness of the film L, so that it follows the Fabry-Pérot relationship:

ml=2nL

where m i s the order of the interference and l is the wavelength position of the interference.

If the film is thin enough, the multiple interferences may appear in the reflectance spectrum as an increased reflection at a specific wavelength which corresponds to the color you observe in the anodized film. 

The question reminded me of a fact that even a large single crystalline TIO2 gets deep blue after annealing in vacuo. The bluish color in this case is usually ascribed to oxygen vacancies working as color centers.  The color of nanotubes may be due to a similar reason.

As-grown nanotube-shaped titania films typically in the organic solutions with fluoride are amorphous and coloured in bronze to brownish tints depending on the thickness, and outer as well as inner diameter of tubes. The colour of these films, as other materials is a function of light absorption and reflection. Only in case of extreemly thin films the colour is determined by interference phenomena ascribed to Fabry-Peroth effect. In case of mkm-thick films, such as TiNT, bronze tones should be explaned by quite uniform scattering of visible light  spectrum inside the titania bulk without absorption of some spectrum part. As also well-known, TiNT films contain inside the tubes numerous debris particles also influencing in the film brightness decrese. Hydrothermal treatments of TiNT films result in the formation of hydrated anatase or anatase-brookite titania, colouring them whiter,  Annealing ot TiNT at 430 - 500 0C results in the foprmation of crystalline material coloured more in yellowish tints since the scattering of light inside the TiNT bulk changes due to the changes in optical properties of amorphous from crystalline anatase .

amorphous materials will display this property. it will take about 800 nm for an amorphous alumina film to obtain its "bulk" color, which is transparent for a glass. or white for a crystalline insulator. the whiter the more insulating.

each color cycle represents for alumina about 200 nm thickness and goes through the spectrum of color from a golden yellow, red, pink, purple, dark blue, light blue turquoise, green, then transparent, passing through various shades for each color. my favorite is the turquoise. Semi conductor or insulating films obtained by CVD in quartz tubes are a good example to visualise in operando with your eyeballs these changes as the growth takes place.

metallic films will not display this property and turns gray very rapidly. some insulating films will also act as full absorbers and turn black quickly due to the presence of multiple lengthscales absorbing the entire spectrum, for example fractal based structures incorporated in a film. hence, not only the chemistry plays a role but also the crystallinity and the morphology, which "dictate" the optical parameters, so to speak.

for other nanostructures, you can get the same effect, though more difficult to predict, for examples imagine etched pores or trenches made in a Si substrate will give different colors based on the pitch and diameter of the pores. and based on the path the light goes on the sample, that is the angle between the light, the source and your eye. go to a grazing angle and the color will change, c.f. G. Macias explanation.

Best

Hi Abdo,

I'm not working with TiNT, but with interference systems like TiOx/Ti. 

In my opinion you obtained thin titanium oxide layer on Ti (plate?) during anodization. 

The formation process of interference colors corectly was explained by Gerard Macias.

Below I added links for some publications, where you can find more information about interference colors of ditanium oxided formed by laser, anodizaton and PVD process:

1. https://www.researchgate.net/publication/267868470_Characterisation_of_coloured_TiOxTiglass_systems

2.

https://www.researchgate.net/publication/266142815_Optical_properties_of_laser_induced_oxynitride_films_on_titanium

3.

https://www.researchgate.net/publication/268219117_The_influence_of_process_parameters_on_the_laser-induced_coloring_of_titanium

4.

https://www.researchgate.net/publication/230238473_Interference_colors_of_thin_oxide_layers_on_titanium

In position 3 you have sequence of interference colors calculated for dielectric layers (with thickness

One more possible explanation, which is feature for TiO2 nanoparticles: Ti3+ ions. Such colores were observed after temperature-programmed reduction in hydrogen at temperatures >350C

The TiO2 layer in the as anodized stage is amorphous in nature. also the XPS analysis reveals a different oxidation states (Ti3+ and Ti2+) for the amorphous film. the amount of Ti3+ is more at the metal oxide /metal interface. So the presence of nonstoichiometry and thin oxide film contributes to the blue or purple color in the initial stage of the anodization and as time progresses the color disappears due to the thickening of the oxide film.  

I think that the color comes from the fact that your film thickness is lower than or of the order of the wavelength of visibile light..  If you have an inhomogeneous film with thicknesses greater than 0.5 microns, the film will scatter light and will appear white or yellow. 

Correct me if I'm wrong, but the thin film interference color depends on the thickness and the ANGLE OF INCIDENCE (and the other material properties). If color is the same using different angles, then in my opinion it is not thickness dependant.

For nano tube arrays it should depend on the pitch. I saw long ago poster for anodic etching of aluminium and thus creating highly ordered "honeycomb" arrays with controllable properties. (as far as I remember one of the possible applications was to create thermally induced surface light waves)

White is if light is reflected evenly in respect to wave length and in random directions (assuming no absorption). If there are coloring impurities, the color will intensify with decreasing the size.

It is quite simple people. Thinking in a single TiO2 crystal with large band gap (~ 3 eV), it would be transparent, since visible light (1.6 - red to 3.1 - violet) is below this range.

So ... why large band gap materials, like TiO2, is not transparent in the powder as is the single crystals? The answer: because the existence in this form of material of a relative large amount of punctual defects, most of them in the surface of the particles (powder form).

Punctual defects introduces localized states within the band gap that allows electron transitions associated with interactions with light in the visible range. As was pointed by our college Tetsuya Aruga, probably oxygen vacancies are associated with states close to the band edges that allows transitions in the blue (2.6 to 2.9 eV).  

That is all folks!

I vote localized states.

Hello ! It is excellent; (восклицание) excellent!, very good ... perfect question ! I have and know device with unique possibilities.! This device have possibilities not only give for You answer on Your question. This new device have possibilities for give inforamtion about (!!!) kinetic of Your process!!! Kinetic ! Wit the help of this device OPEN new possibilities for such ROAD for research. I am Dr. Oleg V Anokhin. My e-mail addreee - [email protected] - I ready for full cooperation with Your possibilities.

I am depositing TiO2 films and it is showing dark brown color. I am using Ar:O2= 1:2.

Have you seen your material through SEM? I have had the same experience and the explanation is quite simple and frustrating: your tubes are detaching and the color you see after annealing is the footprints of your nanotubes.