High-Index Semiconductor Nanoparticles as Low-Loss Alternatives to Gold for Refractive Index Sensing

Abstract

This study presents a comparative numerical analysis of Gold (Au) and high-index semiconductor nanoparticles for refractive index sensing in the visible range. While Au nanoparticles demonstrate high sensitivity ($\approx 150$ nm per refractive index unit), their performance is constrained by ohmic losses. In contrast, high-index dielectrics are shown to exhibit comparable extinction efficiencies driven exclusively by scattering, thereby minimizing thermal losses. Multipolar decomposition reveals that semiconductors support simultaneous electric and magnetic Mie resonances, the interference of which enables directional scattering phenomena unattainable in small metallic particles. These findings suggest that high-index nanostructures offer a robust, low-loss alternative to plasmonics for advanced sensing applications.

Figures (4)

• Figure 1: (a) Extinction efficiency $Q_\text{ext}$ within the visible range (350 to 750 nm) for Au nanoparticles with different values of the radius $r$ between 10 and 50 nm surrounded by a medium with $n=1.333$ (such as water). (b) Peak $Q_\text{ext}$ as a function of $r$, obtained from the cases shown in (a). (c) Spectral position of the peak $Q_\text{ext}$ as a function of $r$, obtained from the cases shown in (a). (d) $Q_\text{ext}$ within the visible range for an Au nanoparticle of 30 nm radius surrounded by media with different values of $n$ between $n=1$ and $n=2$. (e) Peak $Q_\text{ext}$ as a function of $n$, obtained from the cases shown in (d). (f) $\lambda_\text{max}$ as a function of $n$, obtained from the cases shown in (d).
• Figure 2: Extinction, scattering and absorption efficiencies within the visible range (300 to 900 nm) for high-index semiconductor nanoparticles of radius 100 nm made of (a) AlAs, (b) AlSb, (c) GaAs, (d) GaP, (e) Ge, (f) Si and (g) TiO$_2$, and for metal nanoparticles of radius 30 nm made of (h) Au and (i) Ag. In all cases, the refractive index of the surrounding medium is $=1.333$ (such as water). For high-index semiconductors, it is easy to see that the largest contribution to extinction near the resonances comes from scattering, whereas in metals absorption contributes significantly.
• Figure 3: Extinction efficiency and real part of the Mie coefficients $a_1$ (electric dipole), $a_2$ (electric quadrupole), $b_1$ (magnetic dipole) and $b_2$ (magnetic quadrupole) for a Si nanoparticle of 100 nm radius surrounded by a medium with $n=1.333$ (such as water) within the visible range (300 to 900 nm).
• Figure 4: (a) Real and imaginary parts of (top) and absolute difference $\mid a_1-b_1\mid$ between (bottom) the $a_1$ and $b_1$ Mie coefficients for a Si nanoparticle of 100 nm radius surrounded by a medium with $n=1.333$ (such as water) within the visible range (300 to 900 nm). $\mid a_1-b_1\mid$ is locally smallest at 564.56 nm and 839.79 nm, and thus these are the wavelengths where the Kerker condition $a_1=b_1$ is most closely fulfilled. The TM and TE components of the intensity scattered by the same Si nanoparticle for such incident wavelengths where $a_1 \approx b_1$ are shown in (b) and (c), respectively. In (c), a pure zero-backward scattering behavior is observed. On the other hand, in (b), only the TM component shows zero-backward scattering. (d) and (e) present a sensitivity analysis of panels (b) and (c), respectively, showing the TM and TE components for wavelengths that are $\pm$5% around the nominal value. These results confirm that the wavelengths closest to satisfying $a_1=b_1$ provide the optimal balance of strong forward scattering with minimal backward scattering.