Hi Qaiser,

We cannot tell what the real problem is without the experimental details. But for the M-S analysis, you should carry out an EIS analysis before the frequency-static M-S measurement. Also, you can adjust the voltage range (as I know, for titania, it could be from -0.6 to 0.3 V vs. SCE ), and also the scan step (a moderate rate like 20 or 50 mV/per ten steps). After getting the plot, extrapolate the linear region to get the flat band position, and compare it with your OCP (it should be approximately equal).

In addition, when it comes to a more complex system, some modifications should be made based on some mixing principles. The following references may be helpful.

Best regards.

http://pubs.acs.org/doi/abs/10.1021/jp068354l

http://pubs.acs.org/doi/abs/10.1021/ed084p685?journalCode=jceda8

Dear Qaiser,

for semiconducting films on metallic surfaces, single frequency MS measurement could lead to many artefacts. The capacitance at 1kHz could have contributions from the double layer (DL) and space charge (SC) layer capacitances. Did you try fitting your EIS spectra with equalent circuit models, what are the values for DL and SC? In principle, the frequency selection determines the success of single frequency MS analysis, do you have the lowest phase (highest capacitive response) at 1 kHz for all potentials?

As an alternative, you can collect full EIS spectra at different potentials and determine via fitting your space charge capacitance, which would be more reliable. Since your oxide will also undergo changes during polarisation -which again interferes with EIS-, especially in aqueous electrolytes, I would recommend you to use an organic electrolyte with addition of sodium perchlorate for conductivity. Propylene carbonate is a good candidate for this.

Best regards

Ozlem

Dear Qaiser,

first a word of caution:

It seems that there exists kind of confusion in the electrochemistry community what you can get out of  "Mott-Schottky analysis". What you can get is the majority carrier density, when you measure the dc voltage bias dependence of the depletion capacitance (the real part of the measured complex capacitance within a CERTAIN FREQUENCY RANGE ! ) as pointed out by Ozlem Ozcan and Chen Zeng. What you can not get out of these type of measurements is the type of the majority carriers. This is only possible if your contacts are trully Schottky-Ohmic contacts (almost never case in electrochemistry). In one polarity then the depletion capacitance decreases (depletion resistance increases ) - this is reverse bias. When you reverse the polarity of your electrodes, the depletion capacitance increases (depletion resistance decreases) - this is the forward bias. In the Mott-Schottky plot then the (1/Cdepl)^2 plot has a positive slope for reverse bias and negative plot for the positive bias. IT DOES NOT tell you the type of the majority carrier.

And now to your question :

1. I assume that your sample is a sandwich and one electrode is Ti. Then comes Titanium oxide. But what is the other electrode ? I can not really help you if I do not know excatly what is your sample geometry.

2. As has been already pointed out, when you have not determined the proper range of frequencies where you actually measure the depletion capacitance, the errors occur. So do this first.

3. Something in your experiment suggests that the reason for your observation of positive and negative slopes has been in fact the change in the polarity of your electrodes when changing your samples. In other words you did not insert your samples in the identical orientation relative to your electrodes.

4. I would always run the dc voltage bias from largest reverse to largest forward bias in order to clarify the situation. If you really get positive and negative slopes, then your system is indeed Schottky-Ohmic system (that would surprise me). If not, then you probably measure some unwanted artefacts.

My guess would be that Ti-Titanium oxide is a Schottky contact (high resistance rectifying contact). Then the other electrode should be Ohmic contact (low resistance non-rectifying Ohmic contact) if the positive/negative slopes you see have the physical meaning discussed above.

I hope this will help you on the way

With best regards

Petr

as Petr Viscor.  However, let me propose to apply : a 2-plus-electrical (ideally 4) contacts as an elegant technical implementation, for this "Mott-Schottky" measurement.  More analytically : form 2-electrical contacts on Ti-metal (1 for source, and another 1 for measurement ) and the same, e.g. 2-electrical contacts on the other side (the Ti-oxides' side), if possible. If not, then 2 on Ti-metal and 1 on the other side.

Ti // Titanium oxide layer (developed at open atmosphere)   is, already, a ... "sandwich".

Dear Ioannis Samaras:

Thank you very much for adding nice information to my knowledge.

What I think, is may be at  negative external potential or very high positive potential , The accumulation and inversion space charge layer I get as compare to depletion region and semiconducting layer TiO2 behave likes a metal.

Dear Qaiser,

this paper could be helpful.

http://dx.doi.org/10.1016/S0022-0728(02)00814-8

I don't know which potentiostat you are using, but application notes from Gamry usually give good practical tips, even if details of software will be different.

https://www.gamry.com/application-notes/EIS/basics-of-electrochemical-impedance-spectroscopy/

Regards

Ozlem

Dear Qaiser,

now I am getting really confused. Let me ask you my original question in another way.

One electrode is Ti on which you grow your oxide. Where is the other electrode ? And of what material is it made ?

Many greetings

Petr

Dear Petr:

I am using three electrode system, Counter , reference and working electrode.

Thanks

Which are the values of : "(Vn =?)negative external potential or very high (Vp=?) positive potential"

(vs. what, you mean vs. Ref. electrode) ?

Dear Qaiser,

I thoughtn that your system was solid state system, but now it seems that it is a liquid electrolyte (what kind ?), sandwiched between Ti+TiO2 and "pure" Ti electrode. You mention ,in your answer to Ioannis, that you use also "a reference electrode2, something that seems to be a standard way to configure electrochemical system.

Am I right in my guessing or do I still not understand your basic experimental set-up ?

With best regards

Petr

Dear Qaiser,

now I understand the system. The only think left is the geometry. Is your electrode configuration co-centric ? The graphite rod in the middle and the Ti working electrode  around it at radius of (?). The aqueous NaCl is at what nolar concetration ?

I can see that at the end we will get there and I might be even able to come up with some useful comments-

With best regrads

Petr

Dear Qaiser,

excellent !! That was precisely what I wanted to know. I will think about the problem and will come back to you.

With best regards

Petr

Dear Qaiser,

the refs[1,2]  might be helpful.

1.  see page 100: ELECTROCHEMICAL STUDIES ON TiO2 NANOTUBES MODIFIED Ti ELECTRODES  http://www.scientificbulletin.upb.ro/rev_docs_arhiva/full599_630749.pdf

2. Dye-sensitized anodic TiO2 nanotubes http://ac.els-cdn.com/S1388248105002298/1-s2.0-S1388248105002298-main.pdf?_tid=0b282bee-df0e-11e6-a41a-00000aab0f01&acdnat=1484916389_a3da51358a064b95929de7c3f17068e9

Dear Qaiser Ali Khan (which is your first name ?),

I have given your problem a further thought and because it seems to appear on the Research Gate quite often, I have picked up a couple of paragraphs from one of my texts for illustration (attached).

Now, what is discussed in the attachment is a single interface (the example is for Cr(Au)-Si-Cr(Au) interface , with very well defined geometry (both the thickness and the area).

However, your system has three such interfaces (Ti-TiO2 interface, TiO2-NaOH electrolyte interface and finally NaOH-Carbon electrode interface) and therefore, the analysis will be more complicated. It can even be that your "change of sign" might be caused by some interfaces being in reverse bias, while the other are in the forward bias.

My suggestion therefore would be to start with one interface (Ti-TiO2) and progressively add more.

And one more thing, it is not correct to assume that when the density of mobile charges in a semiconductor increase, the interface resistance decrease. This resistance is almost entirely determined by the difference in the chemical potential across the interface and since the chemical potential moves towards the band edges with increasing doping, irrespective the type, it is likely that this resistance will even increase.

I hope this will help you on the way.

With best regards

Petr

Dear Qaiser,

welcome,

A late answer: If the semicondcutor is n type and the electrolyte is acting as metal forming rectifying contact, then you will observe a normal MOTT-SCHOTTKY plot behavior with positive slope where the the positive voltage is applied on the n-semiconductor electrode. please see similar behavior in the link: https://en.wikipedia.org/wiki/Mott%E2%80%93Schottky_plot,

Such plots are sensitive to the presence of interface sates and interface charges. Such states and charges can distort the normal behavior. If you display your results and describe your device one may be able to interpret the results.

Best wishes