Raman spectroscopy is a method for characterization and identification of materials, which is widely used in a broad range of scientific and industrial fields, like materials science, physics, chemistry, biology, medicine, security control, substance control and art examination. It is based on Raman scattering of light on molecules and crystals. As opposed to its elastic counterpart - Rayleigh scattering, Raman scattering is inelastic, which means that the scattered photon has a different energy than the incident one. Because of this, Raman scattering has low probability of occurrence and the scattered Raman intensity is very low. For this reason, many methods for the enhancement of Raman scattering have been developed through the years. One of the new methods of enhancement is based on photonic nanojet, which is a concentrated beam of light emerging from the shadow side of an illuminated microlens. This method of enhancement is characterized by a low cost and a simple principle of implementation. It is a non-invasive, reproducible and reliable way of enhancement. By careful choice of parameters, photonic nanojet can have very high intensity, very narrow width, or very long length. This makes it suitable not only for Raman scattering enhancement, but also for applications in nanolithography, super resolution, optical forces, data storage and similar fields. Although being a promising technique, the role and usage of photonic nanojet in Raman spectroscopy is currently underexplored. Moreover, the properties and conditions for emergence of the photonic nanojet generally are still not clear. This PhD dissertation is a result of four years of research on photonic nanojet and its usage for Raman spectroscopy. It is based on four published papers [1, 2, 3, 4], one still unpublished body of work, and a patent application [5, 6]. This research has resulted not only in the improvement of the method of Raman enhancement, but also with new findings in the general field of photonic nanojet. The research was performed from two angles, experimental and computational. The series of computer codes were written in order to model the photonic nanojet in various conditions. The codes, based on Generalized Lorenz-Mie theory, calculate the electric field intensity from scattering of a Gaussian beam on a dielectric microsphere, upon which a photonic nanoi Extended abstract jet emerges. A large amount of configurations was calculated which provided a systematic overview of photonic nanojet properties and its dependence on parameters. Also, the change of dependencies is detected and investigated by variation of other parameters. The parameters which are varied are the incident Gaussian beam wavelength, position and waist radius, and the microsphere radius. The microsphere refractive index was taken to correspond to SiO2 material. The investigated properties of a photonic nanojet are its maximum intensity, position, width and length. The incident Gaussian beam position is shown to be of critical importance for photonic nanojet properties. Two types of photonic nanojet are identified: Type 1 has lower intensity, its position is further away from the microsphere and has larger dimensions, while Type 2 has higher intensity, it is positioned close to the microsphere edge, and has smaller dimensions. The size matching between the incident beam waist radius and microsphere radius is shown to improve the intensity of the photonic nanojet, but it is not the main contribution. Proper positioning of the incident beam, small waist radius and short wavelength are shown to be important for high intensity. It is also shown that all parameters are important in their absolute value, and that size parameter from Lorenz-Mie theory cannot be applied. Furthermore, parameter combinations for the photonic nanojet of extremely high intensity, very narrow width, or extremely long length with long working distance are determined. In some regimes, intensity oscillations are also detected, and they are identified as whispering-gallery modes and Mie interferences. The occurrence of the photonic nanojet is also investigated when a high refractive index microsphere is used. The investigation followed three theoretical levels: geometrical optics, ray transfer matrix analysis, and Generalized Lorenz-Mie theory. Geometrical optics show that divergent incident light rays can be focused outside a high refractive index microsphere. Ray transfer matrix analysis show that divergent cone of a Gaussian beam produces output beam with a waist outside a high refractive index microsphere. The mathematical condition for that occurrence is derived. Finally, the Generalized Lorenz-Mie theory calculations show that a photonic nanojet can emerge outside the microsphere even when the refractive index of a microsphere is higher than two, which was up to now considered a limit in literature. The calculations also show the difference in focusing of the incident beam based on the refractive index of the microsphere, which is confirmed by the vertical Raman mapping. The Raman enhancement is optimized by variation of experimental parameters. Optimal position of the incident laser beam is determined by vertical Raman mapping, and explained with ray transfer matrix analysis. Laser beam profiles under the microscope objective are determined by a knife-edge method. Antenna effect of the microsphere for the enhancement is detected. The dependence of the enhancement on the collection fiber diameter, microscope objective and microsphere size is determined. Two microsphere materials were used: SiO2 and barium tiii Extended abstract tanate glass. The dependence on microsphere radius shows different behaviors depending on the objective used. The calculations of a photonic nanojet intensity are compared with experimental values of the Raman enhancement, which suggest that the photonic nanojet is not the only contribution to the enhancement. The enhancement strongly lowers by increasing the numerical aperture of the objective. The highest enhancement of the silicon substrate, of 19.29× is achieved in configuration of barium titanate glass microsphere of radius of 4.5 μm and 10× NA 0.25 microscope objective. The mechanism of the enhancement is discussed, which is separated into two contributions. The first contribution comes from the photonic nanojet, and the second contribution comes from the collection system. The model of the effective numerical aperture of the microsphere-objective system is presented, and compared with the experimental results. The usage of the microsphere for the enhancement was further improved by designing the new system for mechanical control of the microsphere. The system is called two-stemmed microsphere and allows positioning of the microsphere under the laser beam of the microscope objective independently of the substrate position. This way, Raman mapping can be performed in which each point is enhanced. The system is tested on a silicon substrate with domains separated by visible borders. Raman mappings are compared with atomic force microscope measurements. With two-stemmed microsphere, the intensity enhancement is 4× and the estimated resolution enhancement is 3×. Combined enhancement of SERS (surface-enhanced Raman scattering) and microsphere is achieved. The SERS substrates which were used were non-uniform and uniform silver nanoislands. The used analytes were 4-mercaptophenylboronic acid or 4-mercaptobenzoic acid. The non-uniform substrates combined with the microsphere show higher but less reproducible enhancement than the uniform substrates with the microspheres.