Figure 1

2D finite element simulations of the distribution of the absolute value (denoted by the color or brightness) of electric field components of an incident TM wave with $\lambda =758\mathrm{nm}$, in a ferrofluid system of thickness $1\mu \mathrm{m}$ containing chains of ${\mathrm{Fe}}_3{\mathrm{O}}_4$ nanoparticles coated by an Ag shell. The ferrofluid system is replaced by an effective medium described by the parameters calculated according to Eq. (1). The direction and size of the blue or dark arrows indicate, respectively, the direction and magnitude of the local power flow or Poynting vector.Reuse & Permissions

Figure 2

Cross-sectional view of 3D finite element simulations of the distribution of the absolute value of electric field components of an incident TM wave with $\lambda =758\mathrm{nm}$, in a ferrofluid system of thickness $1\mu \mathrm{m}$ containing columns aggregated by chains of ${\mathrm{Fe}}_3{\mathrm{O}}_4$ nanoparticles coated by an Ag shell. For convenience we replace each column with a solid cylinder by assuming the solid cylinder to possess the same electromagnetic responses as the column. This assumption is reasonable, at least to some extent, since the incident wavelength is much larger than the size of coated nanoparticles. We set the cross sections of the cylinders to exist in a hexagonal lattice since equal spacing between columns was experimentally observed [11]. The radius of the cylinder is 30 nm, and the center-to-center separation between two adjacent cylinders is 122 nm.Reuse & Permissions

Figure 3

Real part of the effective refractive index $\mathrm{Re}(\eta )$ calculated according to Eq. (7) as a function of $n$, $p$, and $\varphi$ (degree) at $\lambda =758\mathrm{nm}$. Three cross sections correspond to $n=10$, $p=0.25$, and $\varphi =70°$, respectively.Reuse & Permissions

Figure 4

Same as Fig. 3, but for imaginary part of the effective refractive index $\mathrm{Im}(\eta )$.Reuse & Permissions