Revealing Light-Driven Dynamics at Nanostructured Solid-Liquid Interfaces with In-Situ SHG

TL;DR

The work tackles the problem of noninvasively probing light-induced surface-potential dynamics at solid–liquid interfaces by implementing a nanophotonic SHG platform that amplifies interfacial signals beyond conventional limits. A periodic array of $\text{a}$-Si:H nanodisks yields SHG enhancement >200×, with $I_{SH} \propto E(\omega)^4$, enabling real-time, in-situ monitoring of interfacial susceptibility and surface potential at a silicon–oxide–electrolyte boundary. The authors identify two competing light-driven pathways—photocharging at low illumination and photothermal effects at high illumination—whose balance can be tuned by nanostructure geometry, revealing a non-monotonic dependence of $\Phi_0$ on electrolyte concentration and geometry, and exposing ion–electronic coupling via changes in silicon polarizability. The study provides a quantitative framework for actively controlling interfacial charge distributions to improve solid–liquid energy conversion devices, while offering strategies to separate thermo-optical artifacts from genuine photothermal contributions in nanostructured interfacial systems.

Abstract

Light and heat are key drivers of interfacial chemistry at solid-liquid boundaries, governing fundamental processes in sustainable energy conversion systems such as photoelectrochemical and hydrovoltaic devices. However, non-invasive probing of light-induced surface potential dynamics at these interfaces remains challenging due to limited surface sensitivity. Here, we introduce a nanophotonic approach that amplifies second harmonic generation (SHG) from nanostructured solid-liquid interfaces by over two orders of magnitude, providing real-time, all-optical access to light-driven interfacial phenomena. Using in-situ SHG at silicon-oxide-electrolyte interfaces, we uncover two concurrent pathways for light-mediated modulation: (i) low-intensity illumination induces photocharging via carrier generation and trapping, while (ii) high-intensity excitation leads to photothermal heating that modifies surface group dissociation through temperature-dependent reaction equilibria. We further show that nanostructured semiconductor interfaces deviate markedly from the monotonic electrolyte-concentration dependence predicted by Gouy-Chapman theory. Instead, the surface potential exhibits a pronounced non-monotonic behavior governed by interfacial geometry, consistent with prior device-level observations. Importantly, SHG measurements reveal that this concentration-dependent modulation of surface potential directly alters the electronic polarizability of silicon, exposing the underlying ion-electronic coupling at the solid-liquid boundary. By combining nanophotonic design, in-situ SHG probing, and quantitative modeling, this work establishes an experimentally validated framework for actively manipulating interfacial charge distributions to advance the performance of solid-liquid energy conversion technologies.

Figures (14)

• Figure 1: Second harmonic generation modulation and in-situ monitoring at nanostructured electrochemical interface. A) Overview of the device structure for the SHG experiment, highlighting the periodic disk array positioned in the electrolyte medium, excited by both fundamental (FW) and pump laser beams. The FW facilitates probing of interfacial dynamics via SHG, while the pump beam modulates interfacial charges/potential and temperature. The inset shows that the nonlinear optical process (SHG) occurs when a high-intensity fundamental (FW) beam at frequency $\omega$ is excited. B) The compression cell is mounted on an inverted microscope, showing the excitation path featuring the FW (centered at 1030 nm with a bandwidth of 6 nm), and a pump laser (centered at 633 nm with a bandwidth of 10 nm). BS is the beam splitter; BP is the band pass filter. The bottom graph shows an output SH intensity (centered at 515 nm with a bandwidth of 4.3 nm) detected by the collection path for a representative case. C) This illustration depicts the interfaces between semiconductors, oxide, and electrolyte, highlighting that the susceptibility is made up of four distinct contributions. D) SH enhancement in the disk array (red line) is approximately 200 times greater than in the planar film (black line) under identical conditions, with the dashed blue line indicating SH intensity for the disk array without electrolyte, underscoring the significant change due to the solid-liquid interface. As $I_{SH} \propto E(\omega)^4$, the inset shows the FW electric field enhancements in a periodic nanodisk array. The maximum electric field enhancement on the surface is $\approx5$, and the scanning electron microscopy (SEM) image of the nanostructured array. E) In-situ SHG recorded at a 10 seconds acquisition time, illustrating the change in SH intensity due to the transition of the electrolyte from DI water (gray shaded region) to 0.01 mM KCl (blue shaded region) and back, demonstrating the concentration dependence of SH intensity. The inset shows the SH image obtained in DI water, which shows pronounced enhancements in nonlinear polarization around the disk.
• Figure 2: Interfacial potential and susceptibility changes as a function of electrolyte concentration and SiND geometry. A) SH spectra at 0.01 mM electrolyte concentration across different FW powers, demonstrating that while the spectrum position remains constant, the peak height increases with incident power. B) SH intensity plotted against FW light intensity across different electrolyte concentrations, with data fitted to a straight line exhibiting a slope of 2. C) Normalized SH spectra for various electrolyte concentrations, highlighting a noticeable red shift with increasing concentration. The inset shows the non-normalized spectra, highlighting that both peak intensity and spectra positions vary with electrolyte concentration. D) Interfacial potential estimated from SH measurements for varying disk diameters and electrolyte concentrations. E) Spectra positions obtained as a function of SiND diameters for various electrolyte concentrations.
• Figure 3: Photocharging and photothermal modulation of the interfacial susceptibility probed by second harmonic generation. A) Schematic illustration of the light-induced (right) heating and (left) changes in the capacitive interface, affecting surface charge and potential. The inset on the top highlights the initial surface charge conditions at ambient temperature T$_0$. B) Time traces of the light-induced fractional change in effective susceptibility ($\Delta\chi$), normalized to the initial value in the dark ($\chi_0$), for SiNDs with D = 490 nm and D = 510 nm in DI water under 633 nm pump irradiation. Two distinct regimes, depending on the pump laser intensity, are observed. At low intensity, the photocharging effect is observed; at high power, the photothermal effect dominates due to light-induced heating of the structures. C)The reflectance spectra for the SiND arrays in DI water with diameters of 490, 510, and 530 nm. D). Fractional change is total susceptibility as a function of light intensity for the two SiNDs, showing photocharging and photothermal regimes, with transition points at distinct light intensities. E)SH spectra recorded using deionized (DI) water for disk arrays with diameters D = 510 nm (red lines) and D = 530 nm (blue lines) under 633 nm pump laser irradiation (dashed lines) and subsequent pump turn-off (solid lines). F) The time-trace of non-linear polarization $P_{SH}$ during 633 nm excitation for a sample with disk diameter D = 530 nm, showcasing variations under different electrolyte conditions, with dashed lines indicating the decay of the non-linear polarization baseline. G) Corresponding to the measurements in panels F, the normalized susceptibility values obtained by decoupling the TO effect induced change in $E(\omega)^2 \propto R$.
• Figure S1: The operando spectro-photo(electro)-chemical setup. A-B) Sample configuration depicting the (A) bottom and (B) top illumination configuration in the reflection mode. C) An external pump circulates the electrolyte through the designated inlet and outlet ports, and an additional port for a reference electrode is available for precise electrochemical measurements. The nanostructured sample can be arranged either on the top or bottom, based on whether front or back illumination is employed. A hot plate is used to heat the electrolyte for temperature-dependent measurements. D)In the (bottom-right), we have the excitation path featuring the FW (centered at 1030 nm with a bandwidth of 6 nm), and a tunable wavelength pump laser (wavelength, power, and bandwidth can be varied). At the (top-left), an inverted microscope is mounted with the flow cell as shown in panel C. On the (bottom left), the 4F collection path facilitates the signal, which then travels through the spectrometer and is monitored via a CCD camera. The various critical elements in this setup are as follows: M1-M5 signify mirrors; L1, L2 are collimating lenses; TL denotes the tube lens; DM LP900 is the long pass dichroic mirror; FS consists of 0.17 mm thin fused silica substrates for reflecting a small percentage for real-time power monitoring; NDF is a neutral density filter for controlling the power level; BS is the beam splitter; NF is a notch filter designed to eliminate the pump excitation wavelength from the collection path; SP is the short pass filter; and BP is the band pass filter. The bottom graph shows an SH intensity (centered at 515 nm with a bandwidth of 4.3 nm) detected for a representative case.
• Figure S2: Influence of band bending at the silicon–oxide side on interfacial charge and electric field regulation.Calculated surface charge at the solid–liquid interface for n-type silicon under different dopant concentrations (1/cm$^3$). B) The potential profile across the semiconductor-oxide electrolyte system for various bulk potentials of silicon and dopant concentrations. The inset illustrates how changes in band bending modify the interfacial potential profile, thereby shifting the chemical equilibrium on the electrolyte side. C) Surface charge variation as a function of bulk potential of silicon for various electrolyte concentrations. D) Corresponding potential profile for various bulk potentials of silicon and electrolyte concentrations.