In Raman spectra,  shifting of peaks towards lower or higher wavenumber is related to chemical bond length of molecules. The shorter bond length causes to shift higher wavenumber or vice versa. If chemical bond length of molecules changes due to any internal or external effects , then it may cause to shift wavenumber. 

Actually, I don't know doping process, but in Raman spectra, observing wavenumber and intensity provide us information about chemical structure of a sample and amount of specific molecules on it, respectively. 

I hope it helps to you.

The electron–phonon coupling in organic materials intimately correlates electron, or excitation dynamics, with structural skeletal changes that happen in the molecule along particular vibrational modes and these vibrations are particularly enhanced in the vibrational Raman spectrum.

By using the Effective Conjugation Coordinate model (ECC) that establishes the existence of a collective C=C/C–C mode (the C=C and C–C bonds simultaneously enlarge and shrink, respectively), you can observe in your Raman spectra:

(i) the ECC mode is the strongest band of the spectrum

(ii) the frequency of the ECC mode downshifts with the increase of conjugation that happens, among other circumstances, with the enlargement of the chain size in aromatic oligomers and with quinoidization of the backbone.

(iii) the frequency of the ECC mode upshifts whit the increasing of the aromatic character.

The are plenty of publications you can find for a deep review of this topic.

shifting of peaks towards lower or higher wavenumber is specific to the chemical bonds and symmetry of molecules. so if chemical bond length of molecules changes due to any internal or external effects , then it may cause to shift wavenumber. 

A structural change in the molecule is normally the cause of spectral shifts, but changes in temperature and stress within the sample may also affect the band position and shape (see D.J. Gardiner's work on indenting silicon).  This is not my field but some reading around your subject area should give clues as the the observations you make for your samples.

Thanks for your nice discussion .

I think statement 1 is true, What about statement 2 

1. Decreasing peak intensity of metal doped TiO2 than Pure TiO2 shows increment of metal into TiO2 lattice.

2. Increasing peak intensity of metal doped TiO2 than Pure TiO2.

Hi the shift peak of raman  si related to chemical or several paramter such asn deficience oxygene,  grain size, strain, non -stochiometer, different annealing temperatur.

 good luck

Mustafa Unal can you please give reference for your statement

" In Raman spectra,  shifting of peaks towards lower or higher wavenumber is related to chemical bond length of molecules. The shorter bond length causes to shift higher wavenumber or vice versa. If chemical bond length of molecules changes due to any internal or external effects , then it may cause to shift wavenumber. "

Only small shifting is observed due to substitutional doping. It depends on bond lengths as well as type of dopants. In crystalline compounds reduction in height and assymetry in peaks shows the less crystallinity and improper arrangement of atoms in lattice.

in my case for 1 % doping it shifted to lower wavenumber and when dopant concentration increases it shifted to higher wavenumber. what may be the reason for this? there were some changes in morphology. is there any link between morphology and this changes?

It happens due to the increase/decrease of particles size and the change in bond length of the molecules.

Many microstructural factors can cause this Raman shift. I have reported a shift to lower wavenumbers for crystalline samples annealed at a higher temperature which also meant larger crystal sizes. Also, I have also inferred cationic disorder from the full-width-at-half-maximum of the main peaks.

Cyril Ehi-Eromosele did you publish this Raman shift due to crystalline order? If yes, could you refer to your publication, please.

Charlotte Fritsch, the manuscript is presently under consideration for publication. I hope your work is going fine.

Here is one paper that explains the shift.

"If a crystal is subjected to a tensile stress, we can envision the atoms being pulled apart, or chemical bonds lengthened, relative to their normal positions and lengths in an unstressed crystal. As the chemical bond length increases, and the force constant remains the same, we should expect the vibrational frequency to decrease. A shift of the Raman peak position to lower frequencies is exactly what is observed from materials that have been subjected to a tensile stress. Conversely, if a compressive stress is applied, we would expect the atoms in a crystal to move closer together, or the chemical bond lengths to be shortened, relative to their normal positions and lengths in an unstressed crystal. The resulting compressive strain in the crystal results in Raman peak positions shifted to higher frequencies."

https://www.spectroscopyonline.com/view/stress-strain-and-raman-spectroscopy

You can see the explanation of the Raman spectra section.

https://pubs.acs.org/doi/10.1021/acsomega.0c04837#

I recommend Dr. Mustafa Unal answer. It is interesting