Unveiling Spin Transition at Single Particle Level in Levitating Spin Crossover Nanoparticles

TL;DR

The paper tackles the challenge of understanding and controlling spin crossover (SCO) transitions in individual nanoparticles without substrate-induced artifacts. It introduces a substrate-free platform by integrating a quadrupole Paul trap with a multi-spectral polarization-resolved scattering microscope to confine, optically excite, and read out the spin state of a single SCO nanoparticle under controlled pressure and laser power. The authors demonstrate reversible LS↔HS switching with opto-volumetric changes up to ~6% and show that pressure tuning produces a similar volumetric expansion; dehydration under vacuum provides a mechanistic bias toward HS. This approach enables low-energy, real-time control of SCO at the single-particle level and offers a generalizable methodology for integrating SCO nanomaterials into ultralow-power photonic devices and nanoscale sensors, with potential extension to other SCO systems and on-chip photonic architectures.

Abstract

The ability to control and understand the phase transitions of individual nanoscale building blocks is key to advancing the next generation of low-power reconfigurable nanophotonic devices. To address this critical challenge, molecular nanoparticles (NPs) exhibiting a spin crossover (SCO) phenomenon are trapped by coupling a quadrupole Paul trap with a multi-spectral polarization-resolved scattering microscope. This contact-free platform simultaneously confines, optically excites, and monitors the spin transition in Fe(II)-triazole NPs in a pressure-tunable environment, eliminating substrate artifacts. Thus, we show light-driven manipulation of the spin transition in levitating NPs free from substrate-induced effects. Using the robust spin bistability near room temperature of our SCO system, we quantify reversible opto-volumetric changes of up to 6%, revealing precise switching thresholds at the single-particle level. Independent pressure modulation produces a comparable size increase, confirming mechanical control over the same bistable transition. These results constitute full real-time control and readout of spin states in levitating SCO NPs, charting a route toward their integration into ultralow-power optical switches, data-storage elements, and nanoscale sensors.

Figures (4)

• Figure 1: (a) Representative TEM image of Fe(NH$_2$trz)3(NO$_3$)$_2$ particles. (b) Schematic of the electron redistribution between the LS and HS configurations in an octahedral Fe(II) coordination compound upon different external stimuli (pressure, temperature and light irradiation). (c) Picture of a bistable SCO/PMMA solution at room temperature, with low (pink, LS) and high (white, HS) spin states. (d) Heating-and-cooling cycles of $\chi T$ as a function of temperature. (e) Absorption spectrum of NPs redispersed in ethanol. (f) Ellipsometry measurements with heating-and-cooling cycles of the real part of the refractive index for a bare PMMA layer (top graph) and SCO/PMMA film (bottom graph).
• Figure 2: Experimental setup for polarization-resolved scattering measurements on levitated SCO nanoparticles. (a) Schematic of the optical system: three continuous-wave (CW) probe lasers at 642, 785, and 852 nm (LDs) are combined via a fiber multiplexer, collimated, and polarization-controlled using a half-wave plate ($\lambda/2$) and linear polarizer (Pol) before entering the trap along the $x$-axis. A separate CW laser at 488 nm is used as a pump beam when required. The trap is placed inside a vacuum chamber with optical access through a microscope objective, and the scattered light is collected along the $y$-axis and imaged onto a camera. Inset: top-view 3D sketch image of trapped nanoparticles under simultaneous red probe and blue pump illumination. (b) Electrospray source used to inject SCO NPs into the trap. Dark-field and shadowgraphy microscopy images showing the emitter and Taylor cone formation. (c) Side-view schematic of the linear Paul trap indicating the positions of the DC endcap electrodes and AC rods, as well as the polarization direction of the probe laser and the scattered light collection geometry. (d) Photograph of the Paul trap during operation, showing the trapped SCO NPs illuminated by the probe beams.
• Figure 3: (a) Image of an array of isolated SCO NPs inside the trap, which are illuminated with a collimated red probe laser and excited with a focused blue pump laser. (b) Scattering micrograph and its (c) integrated profile of a trapped NP illuminated with a CW-laser at 642 nm. The dashed-red line illustrates a fit using a Lorentzian function. (d) Scattering intensity as a function of the laser linear polarization angle. The dashed-white line corresponds to a sinusoidal fit. (e) Visibility of SCO NPs as a function of their size at 642 nm, 785 nm and 852 nm laser wavelengths. The solid lines were calculated using Mie theory. The horizontal lines illustrate the experimental visibilities extracted from the data. The markers display the intersections with the Mie visibilities. Dashed vertical line corresponds to a size of 315 nm at zero laser excitation and ambient pressure. (f) Graph of the visibility at 642 nm versus NP size. These data were measured for three independent trapped SCO NPs irradiated using several laser intensities, which can be found using the color code on the right-hand side. (g) Relative laser-induced size change of trapped SCO NPs as a function of laser excitation intensity at $\lambda =$ 488 nm. The relative sizes were retrieved following the method illustrated in panel (e). Red circles (P↑) correspond to the up-sweep and blue triangles (P↓) to the down-sweep; dashed lines are guides to the eye. The vertical dashed line marks 6 mW, the power at which independent Raman measurements on non-trapped NPs indicate the HS state (Fig. S4).
• Figure 4: Pressure-dependent scattering of levitating SCO nanoparticles. (a) Experimental polarization-resolved scattering at $\lambda_{\mathrm{probe}} = 642$ nm for a single trapped $\mathrm{Fe(NH_2trz)_3(NO_3)_2}$ nanoparticle while the ambient pressure is reduced from 1013 to 0.1 mbar. Curves are vertically offset for clarity and colored according to pressure (log scale). (b) Experimental visibility $V$ at 642 nm for six isolated particles as a function of ambient pressure. Their initial average NP size at ambient pressure is 280$\pm$4 nm. (b1) Control particles (C1, C2) introduced in the high-spin state exhibit pressure-independent $V$. (b2) Particles initially in the low-spin state show a monotonic decrease in $V$ that saturates below $\sim 10$ mbar. The vertical dashed line marks the approximate crossover pressure, and the horizontal dashed line separates $I_{\mathrm{Lsc}} < I_{\mathrm{\parallel sc}}$ ($V > 0$) from $I_{\mathrm{Lsc}} > I_{\mathrm{\parallel sc}}$ ($V < 0$). (c) Relative particle size change, extracted using the multi-wavelength visibility method, as a function of ambient pressure.