Various gas phase synthesis routes have been employed to produce TiO2 based nanoparticles for photocatalytic applications. Synthesis routes have been selected depending up on the feasibility/constraints of the particular process to produce a given kind of material. The first attempt at producing the photocatalysts involved the synthesis of pure TiO2 nanoparticles by a flame aerosol process. The growth behaviour of TiO2 particles which has significant influence on their crystallinity and surface area, has been controlled by varying the process parameters and selecting suitable flame configurations. An existing method to determine the crystallinity of the powder has been modified which facilitates comparison of the degree of crystallinity of the nanoparticles synthesized by various methods. Resultant TiO2 nanoparticles were characterized by X-ray diffraction (XRD) and nitrogen physisorption to determine the degree of crystallinity quantitatively and surface area, respectively, and the decomposition rate of an industrial dye, methylene blue, has been taken as the measure of the photocatalytic activity (PCA). The importance of an optimum combination of crystallinity and surface area for improved PCA has been shown by comparing the PCA of the as-synthesized nanoparticles with commercial TiO2 (Degussa P25). By depositing another semiconductor material on TiO2, PCA of the resultant composite photocatalyst can be improved due to the effective separation of the charge carriers. The deposited semiconductor has been chosen by virtue of its conduction band potential which should be higher than that of the TiO2. Accordingly, SnO2/TiO2 composite nanoparticles have been synthesized in a single-step by feeding evaporated precursor mixtures into an atmospheric pressure diffusion flame. For the lowest concentration (3.4 mol %) of SnO2 employed in this study anatase phase of TiO2 is stabilized, while segregation of SnO2 is seen at medium (6.9 to 12.4 mol %) and high concentrations (20.3 mol %). Though the equilibrium phase diagram predicts complete solubility of one oxide in another at all compositions, segregation of SnO2 phase is observed which is explained by the usage of diffusion flame in the present study. A particle formation mechanism of SnO2/TiO2 composites is proposed based on the single component aerosol formation. Photocatalytic activity of the composite particles is tested for the degradation of methylene blue and is compared with pure TiO2 synthesized under similar conditions. Improved photocatalytic activity of the composite particles is attributed to the stabilized anatase phase and better charge separation due to the coupling of TiO2 and SnO2 within the composite nanoparticles. The adsorption properties of the SnO2/TiO2 particles are similar to the pure TiO2 particles and the PCA can be further improved coupling the advantages of adsorption properties and charge carrier separation in a single system. WO3 is the kind of semiconductor material that has suitable conduction band potential and surface acidity to perform the dual function mentioned above. Due to the non-availability of the suitable volatile precursor for WO3, WO3/TiO2 composite nanoparticles have been synthesized by using flame spray synthesis. W and Ti precursors were dissolved in a suitable solvent and sprayed into the high temperature acetylene flame using an atomizing gas. Particles with controlled W:Ti ratios were produced at various flow rates of precursor solution and the resulting powders were characterized by BET (Brunauer-Emmett-Teller) surface area analysis, XRD, TEM (Transmission electron microscopy), Raman and ATR-IR (Attenuated total reflection infrared) spectroscopy. Two-dimensional coordinatively unsaturated wolframyl species were well dispersed on the TiO2 surface for the samples equal to or less than 3.6 mol% WO3 and contributed to increase the surface acidity. Crystalline WO3 was formed for the samples above 3.6 mol% WO3. The formation of crystalline WO3 is attributed to the enhanced rate of condensation of W species with increasing loading of tungsten. The variation of lambda (defined as the ratio of the actual oxygen-to-fuel ratio of the reactants to the stoichiometric fuel-to-oxygen ratio) affects the type of surface species dispersed on TiO2 particles and thereby the resultant acidity. The improved photocatalytic activity of the composite particles is attributed to the increased surface acidity and better charge separation due to the coupling of WOx species and TiO2 within the composite nanoparticles.  The development of a practical photocatalytic system focuses on the cost effectiveness of the process. The usage of the expensive solar concentrators and artificial ultraviolet (UV) irradiation for photocatalytic reactions has negative influences on the cost effectiveness. In addition to the cost effectiveness resulting from using solar energy, abundance of the visible light (55 % compared to 6 % of UV) motivates to synthesize TiO2 nanoparticles that absorb visible light. Nitrogen doped TiO2 nanoparticles that absorbs visible light have been synthesized by hot-wall reactor synthesis. The extent of doping of nitrogen is controlled by varying the hot-wall reactor temperature and the flow rate of the gaseous reactants. Increasing the nitrogen concentration above a certain value reduces the PCA and is attributed to the formation of Ti3+ that act as recombination centre for the charge carriers. The concentration of nitrogen at which significant Ti3+ formation takes place has been systematically investigated by XPS (X-ray photoelectron spectroscopy) measurements.