Population Control of Upconversion Energy Transfer for Stimulation Emission Depletion Nanoscopy

Abstract Upconverting stimulated emission depletion microscopy (U‐STED) is emerging as an effective approach for super‐resolution imaging due to its significantly low depletion power and its ability to surpass the limitations of the square‐root law and achieve higher resolution. Though the compelling performance, a trade‐off between the spatial resolution and imaging quality in U‐STED has been recognized in restricting the usability due to the low excitation power drove high depletion efficiency. Moreover, it is a burden to search for the right power relying on trial and error as the underpinning mechanism is unknown. Here, a method is proposed that can easily predict the ideal excitation power for high depletion efficiency with the assistance of the non‐saturate excitation based on the dynamic cross‐relaxation (CR) energy transfer of upconversion nanoparticles. This allows the authors to employ the rate equation model to simulate the populations of each relevant energy state of lanthanides and predict the ideal excitation power for high depletion efficiency. The authors demonstrate that the resolution of STED with the assistance of nonsaturated confocal super‐resolution results can easily achieve the highest resolution of sub‐40 nm, 1/24th of the excitation wavelengths. The finding on the CR effect provides opportunities for population control in realizing low‐power high‐resolution nanoscopy.


Supplement information note2: Optical light path of STED
We achieved the stimulated emission depletion(STED) microscopy based on a conventional scanning confocal system. Figure S2 provides a schematic illustration of an UCNPs STED nanoscopy system.
In the present implementation of UCNPs STED nanoscopy, a single doughnut beam of 808nm laser that generated by a Vector phase plane in the focal region is co-aligned to the Gaussian focal excitation spot of 980nm. The optical system was built on a confocal scanning facility that employed a 3-axis closed-loop piezo nanopositioner( Mad city Lab). A single-mode fibre-coupled 980nm diode laser (BL976-PAGn900, controller CLD1015, Thorlabs, maximum output power 900 mW) was used as the excitation source. One objective lens was used as the collimation lens(Thorlabs RMS10X), after collimation, the excitation beam was transmitted through the first half-wave plate(HWP, WPH05M-980, Thorlabs). And then the 4f optical system is applied with two 300mm lenses. The excitation beam was transmitted through a long-pass dichroic mirror(937 long pass filter), then reflected by a second short-pass dichroic mirror (T750spxrxt, Chroma), and the excitation beam was transmitted through the first quarter-wave plate (QWP, WPH05M-980, Thorlabs). and focused through an oil-immersion objective(Olympus; 100×, NA = 1.4)onto the sample slide. A similar setup for 808nm depletion laser light path. RMS10X lens was used as the collimation lens, and then halfwave plate and PBS were used two maintain the polarization purity. A vortex phase plate(VPP-1a, RPC Photonics) was inserted in the 808nm beam path before the 4f optical system. Then, the longpass filter is allowed 808nm laser to go through and merge with a 980nm beam. The luminescence signal from the sample was collected by the oil-immersion objective lens. The excitation and emission were split by the 750nm short-pass filter. Before coupling the emission into a multi-mode fibre, there has a flip mirror that can transfer the emission into the camera. The end of the multi-mode fibre was connected to a single-photon avalanche diode(APD, SPCM-AQRH-14-FC, Excelitas). To select the emission from upconversion nanoparticles, the bandpass filter was inserted into the emission light path. By using the vortex phase plate, a doughnut-shaped PSF was generated at the focal plane. In the dual-laser STED system, two beams of 980nm and 808nm were carefully aligned to ensure the precise

Supplement information note3: PSF under different excitation conditions
To investigate the relationship between excitation power and resolution of STED, a series of confocal and STED imaging was carried out using different levels of excitation power. As shown in Figure S2, we keep depletion power at 3.4 MW/cm 2, and then excitation power in STED scanning gradually falls from 5 mW to 1.1 mW. Prior to STED scanning, one confocal image was recorded as depicted in Figure S3, and then the depletion beam was supplied to deplete the emission from emitters. As can be seen, the full-width half-maximum (FWHM) of STED PSF is limited shrunk by the depletion beam when the excitation power is 5 mW. It confirms that at the excitation power closed to the saturation power of excitation, and the depletion efficiency is not high enough to deplete the emission pattern to achieve the resolution under the diffractive limit. When the lower excitation power was applied to the confocal and STED, it is obvious that the FWHM of the confocal and STED had a significant decrease. The non-saturation excitation of the confocal will induce higher resolution imaging. When the smaller PSF of the confocal occurs, a higher resolution of STED will be achieved. With 1.1mW excitation power, the FWHM of STED can reach 40.6 nm. Figure S3. The confocal imaging of UCNP and its cross-section shrinking into the diffractive limit processing under the ultralow non-saturate excitation power. Based on this non-saturated excitation, the upconversion nanoparticles can achieve the highest resolution with the assistance of fixed depletion power. The cross-section of UCNPs got from confocal and STED image.

Supplement information note4: Emitter core/core-shell structure modulation at the depletion process
To discuss the core-shell structure for the enhancement of depletion efficiency, core and core-shell structure UCNPs were synthesized with the same Tm 3+ doping concentration of 12 mol%. Under dual-laser illumination, the optical depletion ratio of 455 nm emission was measured as a function of the 808nm intensity (the results were normalized by the emission intensity). As shown in Figure S4, with the same excitation power of 980 nm laser, the core structure of UCNPs has a lower depletion power and higher depletion efficiency, which demonstrates our assumption. We further investigate the different excitation power effects with core and core-shell structures. 1 mW excitation power has a much lower depletion threshold and higher depletion efficiency than 4 mW and 8 mW, which further exhibits in our previous experiment. The lower excitation power, the lower depletion threshold, and the higher depletion efficiency from the STED system. Figure S4. The depletion efficiency of core and core-shell structure nanoparticles. The results got from 12% Tm 3+ doped core and core-shell nanoparticles.
In a typical experiment, 1 mmol LnCl 3 ·6H 2 O (Ln=Y, Yb, Tm) with the molar ratio of 72:20:8 were added to a flask containing 6 mL OA and 15 mL ODE. The mixture was heated to 160 °C under argon flow for 30 min to obtain a clear solution and then cooled down to about 50 °C, followed by the addition of 5 mL methanol solution of NH 4 F (4 mmol) and NaOH (2.5 mmol). After stirring for 30 min, the solution was heated to 80 °C under argon flow for 20 min to expel methanol, and then the solution was further heated to 300 °C for another 90 min. Finally, the reaction solution was cooled down to room temperature. The products were precipitated by ethanol and centrifuged (9000 rpm for 5 min), then washed 3 times with cyclohexane, ethanol and methanol to get the nanoparticles.
To get the nanoparticles with bigger size, layer-by-layer epitaxial growth has been employed. The shell precursors preparation is similar to that for the core nanoparticles synthesis until the step where the reaction solution was slowly heated to 150°C and kept for 20 min. Instead of further heating to 300 °C to trigger nanocrystal growth, the solution was cooled down to room temperature to yield the shell precursors. For epitaxial growth, 20 mg as-prepared core nanocrystals were added to a containing 6 ml OA and 6 ml ODE. The mixture was heated to 170 °C under argon for 30 min, and then further heated to 300 °C. Next, 0.15 ml as prepared shell precursors were injected into the reaction mixture and ripened at 300 °C for 2.5 min, followed by the same injection and ripening cycles for several times to get the nanocrystals with the desired size. Finally, the slurry was cooled down to room temperature and the formed nanocrystals were purified according to the same procedure used for the core nanocrystals.