I think you can try for AFM  in non contact mode to get the morphology of  your titanium piece before and after modification.

You can use Grazing incidence diffraction with ref:

Vineyard, G.H. (1982), “Grazing Incidence Diffraction and the Distorted-Wave Approximation for the Study of Surfaces”, Physical Review (B), 26, p4146-4158.

Barner, P (1992), “Dynamic Diffraction Studies on the Synthesis of Ceramics Using Synchrotron Radiation”, Trans.J. Br. Ceram.Soc., 91 (1),26-29.

If you have synthesized TiO2 nanotubes on Ti plate by anodization and wanting to know the lengths of the tubes, then you put a small piece of your sample inside FESEM, tilt the sample and measure the lengths near the edges.

XRR - Best and cheaper way. RBS is also a good way.

I still would throw in that a reliable value requires reproducibility of your results and you should not try to save all samples from destructive investigations at the price of good pictures and proof of your data. Produce more samples, make cross-sections and proper SEM/TEM. Compare the thickness you obtained in this way with results of whatever non-destructive method you'll pick and only if they fairly aggree go on with the non-destructive characterisation

On the sample cover small area (1 square mm) with tape or paint (depends on temperature of produce process). After finish the process remove tape. This will uncover Ti surface, so you will have the step between layer and base. Use AFM or even light profilometer to take 3D diagram, which let you to see differences in the height.

I suggest XPS to measure exact thickness of coatings

Thank you all!! There are few issues.

1. The surface can be masked in a point location (1 sq mm) with tape or paint, however, in AFM or profilometer it is not turning up well because, the oxide grows from surface down. I guess for profilometer, we need a coating which is deposited from the surface up.

2. AFM in non-contact mode might not help because the depths we are talking about in anodization is in tens to hundreds of nanometers. One can't get a clear idea about the oxide thickness.

3. Cutting polishing and viewing the edge is very tricky because the angle of cutting and polishing might not be precise and might lead to parallax errors when observed in SEM since the cut has to be 100% normal (perpendicular) to the surface.

4. Observing the edge is ruled out because, when anodization is done electrochemically, the entire 3 dimensional surface is oxidized. 

5. I have tried grazing incidence X-ray diffraction, and ended up with Titanium peaks, especially in thin oxides. How to avoid X-ray penetration? and what minimum incidence angle is preferred?

6. That leaves us with XPS. I think depth profiling in XPS seems to be the only option. Please do correct me if I am wrong.

There is a relative simple way to accurately determine thickness of oxide layers on metals (or any Kind of layers on a substrate) and that is using Electron Probe Micro Analysis EPMA (=WDX) thin film Analysis. It is non-desctructive and it simply measures the mass-thickness of oxigen on the surface. One uses a sufficent high primary electron energy that the electrons passes through the oxide layer and you measure the O-Ka Xray Line in comparison with any kind of reference material with a known oxigen content. e.g. we use a garnet cristal (Yttrium Iron Garnet). But any kind of well defined oxide will work. Then with an evaluation software like "STRATAgem" the mass-thickness is detemined from the the so called K-ratio, which is the ratio of O-Ka intensities on the sample and the standard. There are some assuptions which enter the results i.e. 1. you have to assume the stochiometry of the oxide e.g. TiO2, and 2. you have to assume the density of the oxide which is needed to calculate the gemoetrical thickness [nm] from the mass-thickness [µg/cm2].

The method exist already since 1990 but is not wide spreaded. The principles can be read in "Surface Film X-Ray Analysis" from J.L. Pouchou and F. Pichoir in SCANNING Vol 12, 212-224 (1990). The Software is available at http://www.samx.com/microanalysis/products/stratagem_us.html . The EDX Software from Oxford Instruments has also an optional "thin film analysis" tool, but to my opinion the STRATA Software is better and more flexible.

Best regards

Kirsten  

may I suggest some classical books

1. oxides and anodic films by ASHOK VIJH

2. ANODIC OXIDE FILMS by L.YOUNG for oxide films

All CSIR ( INDIA) has these books

I suggest you use EDX to estimate the thickness, if you do not want to alter structural properties. (do not know the thickness, so I assume the electron goes beyond your oxidized layer into the Ti under 25kV or so)

1. Check with XPS what oxide do you have, TiO2? TiO1.5?

1: Use the same setting for EDX, do not let others change in between;

2: measure on the Ti;

3: measure on your oxidized sample;

4: measure on some TiO2 ot TiO1.5 sample.

5: estimate the thickness via simplifying the shape of electron beam in the sample.

reference of  the method see:  

In our school, we measure the thickness of TiO2 layer as expressed in Fig.

try ellipsometry by the different refractive index

and XPS is the other way

ELLIPSOMETRY is known since 1980    --try

Hi,

By a Scanning Tunnelling Microscopy(STM) or by Atomic force Microscopy(AFM), refer to the herewith attached article;

-Habib.K.,.Eling.V.,and Wu.C.," Measuring Surface Roughness of an Optical Thin Film With Scanning Tunneling Microscopes",Journal of Materials  Science Letters,Vol.9,PP.1194 (1990). 

Explore using Microtome machine.

Thank you for all your valuable suggestions. I managed to manually scratch the sample surface and observed it in FE-SEM hoping to find an upright cross section, and fortunately I could see the cross section. However, from this observation, I could also conclude that ellipsometry will be tedious since anodization at different conditions yields different porosities and complex structures. This complexity changes everything.

Electron microscopy is the answer to your question. You can easily measure the coating thickness using a FEGSEM . Also Surface Probe techniques help to determine the relief of a surface.