I am trying to simulate plasmonic resonance of gold nanoparticles in water medium in COMSOL Multiphysics 5.2. I have so far been unsuccessful in reproducing results reported in past works. When I tried it for a Silicon nanoparticle, the agreement was good with Mie approximation. However, with gold, there is always some difference. The dielectric functions for gold were obtained from Johnson and Christy (1972) which is commonly cited as the source of this information. I have also checked whether my results are independent of the grid and the perfectly matched layer. In the literature, people have reported that there is exact agreement between Mie approximation and FEM. Can anybody help me by pointing out where I might be wrong? Thank You!