Abstract：Star formation is the fundamental process in the galaxy formation and evolution. In this talk, I fill focus on three key questions about the star formation in galaxies. One is the relationship between the product (such as the stellar mass and metallicity) and the star formation process. Another is the relation/interaction between the fuel (such as molecular gas) and star formation. The last one is the effect of active galactic nuclei (AGN) on star formation in nuclear region.

Metal--poor galaxies (MPGs) play an essential role in understanding the galaxy evolution, especially for the galaxies at the early stage of evolution, or for those evolve slowly. we search for these MPGs from the LAMOST ExtraGAlactic sur-vey in nearby universe and the eBOSS survey at the 0.6 < z < 0.9 universe. We find more than 10 extremely metal--poor galaxies, then study the relationship between star formation rate (SFR), stellar mass (M*) and metallicity (Z) at the early period of galaxy. The M* - Z relation (MZR) of the MPGs at intermediate redshift is systemically lower than the relation in nearby MPGs and normal star forming galaxies (SFGs) at intermediate redshift. This result indicates the evolution of MZR. When fix the stellar mass, the SFRs of MPGs at intermediate redshift are much higher than MPGs in nearby universe and SFGs at intermediate red- shift. After taking the SFR into account, the scatter in M* – Z – SFR relation (FMR) is slightly reduced, which means that the metallicity is weakly dependent on the star formation rate in MPGs. Furthermore, the FMR at intermediate redshift is lower than in nearby universe, indicating the evolution of FMR.

The research about the relation between local metallicity distribution, stellar mass surface density ($\Sigma_*$) and the global galactic parameters is popular in these years. Using the IFU data of MaNGA survey, we select the star--forming galax-ies, then study the relationship between local metallicity, stellar mass surface density and total stellar mass. We find the metallicity increases steeply with increasing $\Sigma_*$ at a fixed M？. Similarly, at a fixed $\Sigma_*$, the metallicity increases strongly with increasing M* at low mass end, while this trend becomes less obvious at high mass end. We establish a new M* – $\Sigma_*$ – Z relation, which reduces the scatter in $\Sigma_*$ – Z relation about 30%. These results suggest that the M* – $\Sigma_*$ – Z relation is a more universal and funda-mental relation than MZR and $\Sigma_*$ – Z relations.

The star formation activities in nuclear region will be affected by the central AGN and gas properties. Using the ALMA band 3 and VLT/MUSE data, we perform a spatially resolved analysis of molecular gas and ion-ized gas in the central region of NGC 1365, and explore the star formation activities and kinematics in dust lanes, circumnuclear ring, and outflow biconical regions. We find the resolved main sequence relation and K--S relation is super-linear at a resolution of 180 pc, with steeper slopes. The K--S relation is broken down at a low gas surface density regime, which might be caused by the efficient depletion of molecular gas or the decorrelation between star formation and molecular gas in arms. We detect a few of regions in the inner star-forming ring, showing the highest star for-mation efficiency (SFE). These regions are regarded to harbor massive star clusters and might be caused by the cloud- - cloud collisions in the denser molecular gas regime. The star formation in the southwestern dust lane is much weaker than the northeastern one. This result supports the scenario of negative feedback of outflows, because the radiation from the central AGN and starburst can prevent the molecular gas from cooling even in the denser clouds. We also find two obvious non-circular motion components of molecular and ionized hydrogen gas, reaching velocity up to 100 km/s. These motions probably indicate the scenario that the outflows from AGN and starburst could sweep out the low--density molecular gas and diffuse ionized gas on the surface of the disk.