1. Introduction

Super-resolution microscopy is of great importance in investigating the structures and dynamics of nanoscale architectures in recent years [1]. Superior to other super-resolution techniques, stimulated emission depletion (STED) microscopy could achieve video-rate imaging in real time without additional data processing [2]. Furthermore, STED microscopy combined with the two-photon excited (TPE) technique, namely two-photon excited stimulated emission depletion (TPE-STED) microscopy, has been developed as an important and promising approach in the studies of deep biological tissues [3,4 ]. However, there are some restrictions related to conventional STED microscopy. The imaging depth of TPE-STED is still limited by the short wavelength of the STED laser beam because of strong scattering, although longer excitation wavelengths have been proposed to achieve greater tissue penetration [5]. Furthermore, conventional fluorescent probes suffer from the problem of photo-bleaching induced intermittent detection. Meanwhile, further development of TPE-STED microscopy has been restricted by the high cost of high power femtosecond (fs) excitation and STED lasers. To date, STED probes with solutions to all the above problems have not yet been reported.

Lanthanide-doped upconverting nanoparticles (UCNPs) have been recently developed as a promising class of luminescent biomarkers [6,7 ]. The unique properties of UCNPs, including their non-photobleaching, non-photoblinking, spectrally narrow multi-color emissions and autofluorescence-free properties, make them a powerful and attractive class of contrast agents in biological applications [6]. The luminescence process of UCNPs mainly relies on several energy transfer processes, such as energy transfer upconversion (ETU), ground state absorption (GSA), excited state absorption (ESA), cross relaxation (CR) and photon avalanche (PA) [8]. The contribution of these processes could be modulated by either physical or chemical methods, which provides the possibility to optically deplete the luminescence by manipulating the energy transfer processes [9,10 ]. To the best of our knowledge, the extensively-used Er³⁺, Ho³⁺ or Tm³⁺ co-doped NaYF₄-UCNPs have never been studied for multi-photon super-resolution microscopy, although overly complicated and time-consuming subdiffraction resolution microscopy has been studied using low efficient Pr³⁺:YAG nanoparticles [11]. Many efforts have been devoted to the luminescence generation mechanisms of Er³⁺-doped UCNPs under 795-nm, 980-nm or 1490-nm near infrared (NIR) continuous wave (CW) irradiation, which inspired to optically regulate the energy transfer processes among the rich ladder-like energy levels of Er³⁺ [12–15 ].

In this work, for the first time, an optical luminescence depletion mechanism employing NaYF₄:Yb³⁺/Er³⁺ UCNPs is proposed, and its potential for multi-photon STED-like microscopy was carefully investigated. The ESA process induced by an 1140-nm laser from ²H_11/2 (Er³⁺) to ²K_15/2 (Er³⁺) transfers the relaxed energy to Yb³⁺ in NaYF₄:Yb³⁺/Er³⁺ UCNPs and then the two-photon green emission is depleted. A theoretical model analysis using steady-state rate equations was performed to investigate this proposed mechanism, coincided with the drastically rising blue emission centred at 478 nm. The green emission depletion efficiency is approximately 30% and could be optimized by a CW 1140-nm laser. The non-photobleaching and non-photoblinking properties of these nanomaterials under two lasers co-irradiation were verified. The advantages and potential of applying NaYF₄:Yb³⁺/Er³⁺ UCNPs to super-resolution microscopy were also discussed in detail.

2. Experimental setup

The optical experimental setup was built to perform the experiments, as shown in Fig. 1 . The 1140-nm fs laser was obtained from an optical parameter oscillator (OPO), pumped by a Ti: sapphire fs laser (Mira HP, Coherent), while a 795-nm CW light beam was generated from another laser. An 850-nm long-pass filter and an 800-nm band-pass filter were used to purify the laser spectra. These two lasers were spatially overlapped into the same pathway using a polarizing beam splitter (PBS), and directed into a multiphoton laser scanning microscope system (MPE FV10-s, Olympus). The spatially overlapped lasers were reflected by a 690-nm dichroic mirror (DM), and focused on the sample using an objective (NA = 1.35, Olympus). The emission luminescence was backforward collected by the same objective, and captured by a compact spectrometer (QE65000, Ocean Optics). Two half-wave plates (690 nm-1200 nm) were placed on the optical pathway to each adjust the power of one of the two lasers. The excitation power of the 795-nm laser was kept at 1.4 mW in the NaYF₄:Yb³⁺/Er³⁺ UCNP experiments. A blocker was placed in front of the OPO while single laser irradiation was performed. The FWHM of 1140-nm laser was recorded to be 18 nm using a NIR spectrometer (NIR Quest, Ocean Optics).

 

Fig. 1 The scheme of experimental setup. P1, P2: half-wave plate. F1: 800-nm band-pass filter. F2: 850-nm long-pass filter. B: blocker. M: silver reflector mirror. PBS: polarizing beam splitter. DM: 690-nm dichroic mirror. OL: objective lens. S: sample. FL: focus lens. Inset: FWHM of 1140-nm laser.

Download Full Size | PDF

3. Results and discussion

3.1 1140-nm laser assisted ESA in Er³⁺-singly doped UCNPs

The NaYF₄:5%Er³⁺ and NaYF₄:18%Yb³⁺/5%Er³⁺ UCNPs used in the experiments are synthesized following our previous protocols [16, 17 ]. Under 795-nm excitation, Er³⁺-singly doped UCNPs can generate two-photon green emissions. According to previous research, the energy transfer process E0 between Er³⁺, that is, ⁴I_11/2(Er³⁺) + ⁴I_11/2(Er³⁺)→⁴I_15/2(Er³⁺) + ⁴F_7/2(Er³⁺), is dominant for green luminescence generation of Er³⁺, as depicted in Fig. 2(a) . The systematic luminescent process could be described as follows. To begin with, the energy level ⁴I_9/2 is populated as an Er³⁺ ion absorbs a single 795-nm photon through GSA. After that, the energy levels ⁴I_11/2 and ⁴I_13/2 are populated due to the occurrence of multi-phonon relaxation from ⁴I_9/2, and then generate downconverting luminescence, including 980-nm and 1500-nm emissions [18]. Next, upper levels including ²H_9/2, ⁴S_3/2 and ⁴F_9/2 are populated by rapid relaxation from ⁴F_7/2, pumped by E0. Green emission is mainly a result of E0 because E0 is dominant in populating ²H_9/2 and ⁴S_3/2, although other energy transfer processes also take part in luminescence generation [15]. So, in order to deplete the green emission, the population of ²H_11/2 and ⁴S_3/2 needs to be cut down since they are mainly responsible for green luminescence generation. Calculation of the gaps between the energy levels reveals that a photon energy of 1140-nm light precisely well matches the energy gap from ²H_11/2 to ²K_15/2 and from ⁴I_13/2 to ⁴F_9/2 [Fig. 2(a)] [19]. The energy gap between ²H_11/2 to ²K_15/2 allows that the excitation energy of ²H_11/2 could be pumped to higher energy levels, which might be depopulated by CR or non-radiative multi-phonon relaxation, and as a result contribute less to the green emission. This assumption suggests that an 1140-nm laser can work as a depletion laser beam for two-photon green emission of NaYF₄:Er³⁺ UCNPs.

 

Fig. 2 (a) Proposed mechanism of luminescence generation of 795-nm laser excited Er³⁺-singly doped NaYF₄ UCNPs with/without 1140-nm irradiation. (b) Luminescence intensity of Er³⁺-singly doped NaYF₄ UCNPs under 795-nm CW excitation with/without 1140-nm irradiation. Inset figure: amplified luminescence spectrum from 350 nm to 500 nm.

Download Full Size | PDF

Fig. 2(b) shows the luminescence spectra of the prepared NaYF₄:Er³⁺ UCNPs under 795-nm excitation with/without 1140-nm irradiation. After employing 1140-nm depletion laser beam, the green emission unexpectedly increased instead of being depleted. An overall enhancement of luminescence is observed, including 410 nm and green emission. Considerable 410-nm luminescence, originated from other CR processes, was detected from the nanoparticles under 795-nm laser excitation [12]. These results could be attributed to the following analysis. On one hand, while the excitation and the depletion lasers are working simultaneously, the population of ²H_11/2 and ⁴S_3/2 are partially compensated by the continuous relaxation from ⁴F_7/2, which counteracts the depletion process. Meanwhile, the population of ²K_15/2 excited from ²H_11/2 would quickly return to lower energy levels (e.g., ²H_11/2), through a multi-phonon relaxation process, and again contribute to the green emission. During this process, some excitation energy would decay to ²H_9/2 and contribute to the arising of 410-nm luminescence. This unsatisfactory performance could be improved by doping other elements to introduce additional energy transfer pathways. In short, we have demonstrated that an 1140-nm laser could induce ESA in Er³⁺ from ²H_11/2 to ²K_15/2, although green emission depletion was not achieved in NaYF₄:Er³⁺ UCNPs.

3.2 Green emission depletion by 1140-nm laser in NaYF₄:Yb³⁺/Er³⁺ UCNPs

In order to achieve depletion of green emission of Er³⁺-doped UCNPs, Yb³⁺ co-doping was proposed as an effective way to change the energy flowing pathway and regulate the steady state of excitation energy distribution. According to previous research, the absorption band of Yb³⁺ is near 980 nm and does not interact with the 795-nm or 1140-nm laser directly [20]. The energy transfer between Yb³⁺ and Er³⁺ would be dominant and surpass the energy transfer processes between Er³⁺ under 800-nm band excitation, which has been investigated in previous research [15, 20 ]. As depicted in Fig. 3(a) , with Yb³⁺ co-doping, two energy transfer pathways, E1 and E2, mainly contribute to the luminescence of NaYF₄:Yb³⁺/Er³⁺ UCNPs. The E1 process indicates the energy transfer process between Yb³⁺ and Er³⁺, that is, ⁴G_11/2(Er³⁺) + ²F_7/2(Yb³⁺)→⁴F_9/2(Er³⁺) + ²F_5/2(Yb³⁺). This process was recently proposed by Anderson et al, and has successfully been applied to explain the blue and red emission generation processes in the popular β-NaYF₄ host [21]. In addition, the E0 process is dominant for green emission generation in NaYF₄:Er³⁺ UCNPs and NaYF₄:Yb³⁺/Er³⁺ UCNPs [15, 20 ]. In NaYF₄:Yb³⁺/Er³⁺ UCNPs, the luminescence generation process could be explained as follows [Fig. 3(a)]. The excitation energy of the 795-nm laser is absorbed by ⁴F_9/2 of Er³⁺ and shortly relaxed to ⁴I_11/2, while the E2 process leads to a two-photon process and results in the energy accumulation of ⁴F_7/2. The green emission is generated from ²H_11/2 and ⁴S_3/2, which is populated through the multi-phonon relaxation of energy in ⁴F_7/2. The red emission not only comes from the sequential relaxation from ⁴S_3/2, but also originates from the E1 process. In principle, Yb³⁺ provides an additional energy transfer pathway to direct the energy from ⁴G_11/2 to ⁴F_9/2. In this case, the ESA process from ²H_11/2 and ²K_15/2, induced by 1140-nm laser irradiation, could transfer the energy to neighboring Yb³⁺ through the E1 process instead of contributing to the green emission by sequential multi-phonon relaxation.

 

Fig. 3 (a) Proposed mechanism of luminescence generation of 795-nm laser excited NaYF₄:Yb³⁺/Er³⁺ UCNPs with/without 1140-nm irradiation. (b) Luminescence intensity of NaYF₄:Yb³⁺/Er³⁺ UCNPs under 795-nm CW excitation with/without 1140-nm irradiation. Inset figure: amplified luminescence spectrum from 350 nm to 500 nm.

Download Full Size | PDF

Fig. 3(b) shows the experimental results of NaYF₄:Yb³⁺/Er³⁺ UCNPs luminescence depletion using an 1140-nm laser as the depletion beam. After adding 1140-nm laser, the green emission was effectively depleted as expected. We also observed obvious a distinct rise in 478-nm luminescence [Fig. 3(b), inset], which originates from ²K_15/2/²G_9/2 to ⁴I_13/2 transition. Compared to Er³⁺ singly-doped UCNPs [Fig. 2(b), inset], in Yb³⁺/Er³⁺ co-doped NaYF₄ UCNPs, the enhancement of 478-nm luminescence was much greater than that of 410-nm enhancement. These experimental results substantially confirmed our hypothesis that the excitation energy of an 1140-nm laser could be transferred to Yb³⁺ instead of inducing 410 nm or green emission. The experiment using 795-nm fs and 1140-nm fs co-irradiation was also performed to further investigate our mechanism. The experimental results show that the excitation of 795-nm fs (part of seed light pumping OPO) made no difference in terms of depletion efficiency as well as emission spectroscopy, mainly due to the fact that the decay time of UCNPs (us) is several orders longer than the short interval time (13 ns) of high repetition fs pulses. In this part, a green emission depletion mechanism of NaYF₄:Yb³⁺/Er³⁺ UCNPs by using an 1140-nm depletion laser was experimentally demonstrated.

3.3 Modeling the mechanism of green emission depletion

To better understand the proposed depletion mechanism, we performed model analysis for NaYF₄:Yb³⁺/Er³⁺ UCNPs using steady state rate equations. To begin with, the following rate equations were used to express the upconversion process of NaYF₄: Yb³⁺/Er³⁺ under 795-nm and 1140-nm lasers co-irradiation:

In these equations, $N_i(i=1-10)$refers to the population densities of different energy levels, as indicated in Fig. 2(a) and Fig. 3(a). $F_{795}$ and $F_{1140}$ represent the excited power density of the 795-nm and 1140-nm laser, respectively. ${\tau }_i(i=1-10)$ refers to the lifetime of corresponding states in Er³⁺ ions.${\sigma }_1$ is the absorption cross sections of the 795-nm laser, while ${\sigma }_2$ and ${\sigma }_3$ are the 1140-nm light absorption cross sections of ⁴I_13/2 and ²H_11/2, respectively. ${\beta }_i$ donates the phonon-assisted non-radiative relaxation rate of the corresponding energy levels. $W$is the cross relaxation coefficient of E2, while $K_{ET9-5}$ is the rate constant for energy transfer from ⁴G_11/2 to ⁴F_9/2. $N_{Yb0}$ represents the population density of the Yb³⁺ ions in the ground state. In this model, we assumed that $N_1=N_0=C$ for the low excitation laser power, while $N_0$ is the population density of the Yb³⁺ ions in the ground state.

Based on Eqs. (1) and (2) , as $2W{N}_3^2$ could be neglected under low concentration Er³⁺ doping and low excitation laser power, we could obtain:

By rearranging Eq. (5), we could also get:

According to Eqs. (7)-(9) , we could solve $N_8$, $N_9$ and $N_{10}$ as described in the following equations:

where

and

In ²G_7/2/²K_15/2/²G_9/2, ⁴G_11/2 and ²H_9/2, the non-radiative relaxation is a dominant process for the low luminescence generation of these energy levels [Fig. 2(b) and Fig. 3(b), insets]. Thus, it is reasonable to assume that ${\beta }_8\gg 1/{\tau }_8$, ${\beta }_9\gg 1/{\tau }_9$ and ${\beta }_{10}\gg 1/{\tau }_{10}$. Under this approximation, $N_8$ could be described as:

With the combination of Eqs. (6), (10), (14) and (15) , the correlation of $N_6$ and $F_{1140}$ could be derived as:

Equation (16) has expressed the correlation of $N_6$ and $F_{1140}$. Under the 795-nm CW laser excitation with fixed power, the right part of this equation remains immutable. The correlation of $N_6$ and $F_{1140}$ is a rational function$y=a/(b+cx)$, where $x$represents the power of the 1140-nm laser, and $y$ is the intensity of green emission. To deplete the green emission, that is, to decrease the value of$N_6$, increasing the excitation power of the 1140-nm laser would be the most feasible approach.

In addition, if $K_{ET9-5}=0$ (corresponding to the case in NaYF₄:Er³⁺-singly doped UCNPs), we could also obtain an analogous expression of $N_6$ in Er³⁺-singly doped UCNPs:

Equation (17) suggests that increasing the power of the 1140-nm laser would not induce green emission depletion in Er³⁺-singly doped UCNPs, while the experimental results in Fig. 2(b) show that an 1140-nm fs light beam induced enhancement of green luminescence. This is probably ascribed to the fact that the excitation energy of 1140-nm photons would partially contribute to the luminescence generation of Er³⁺.

Regarding the competition between 410-nm and 478-nm luminescence under 795-nm and 1140-nm laser co-irradiation, we could derive the correlation of $N_8$ (410-nm population) and $N_{10}$ (478-nm population) by using Eqs. (12) and (13) :

In NaYF₄:Yb³⁺/Er³⁺ UCNPs, $R_{410/478}$ is relatively small because the value of $K_{ET9-5}{N}_{Yb0}$ is typically large due to high concentration of Yb³⁺-doping [21]. In contrast,$N_8$ in NaYF₄:Er³⁺-singly doped UCNPs is somehow larger than $N_{10}$in our experiment [Fig. 2(b), inset]. This result further proofs the accuracy of our theoretical model, which has qualitatively explained the proposed depletion mechanism and the correlation between 410-nm and 478-nm enhancement in different materials, although ${\beta }_8$,${\beta }_9$ and ${\beta }_{10}$ might vary slightly with Yb³⁺ doping.

3.4 Exploring depletion efficiency, decay time and photostability

We also explored the power dependent depletion efficiency of green emission. As depicted in Fig. 4(a) , the optical depletion efficiency of green luminescence increases with the power of 1140-nm depletion laser beam, and the intensity depletion efficiency achieved approximately 30%. The data could be well fitted with a rational function $y=a/(b+cx)+d$, which is analogous to the form of Eq. (16). The term d originates from the discontinuous depletion power function of the laser beam, which is a high-repetition fs laser instead of a CW laser. This result indicates that the depletion efficiency limit of this material could be broken by an 1140-nm CW laser. Besides, the laser density was calculated to be ~100 KW/cm². Compared to typical STED laser density (1-100 MW/cm²), this laser density is three orders smaller and would be more friendly to biological tissues [22].

 

Fig. 4 (a) The green emission depletion efficiency of the 1140-nm laser. The experimental data was well fitted with a rational function, where the independent variable is the depletion power density and the dependent variable is the intensity of the green emission. (b) The decay time of the green emission from Yb³⁺/Er³⁺ UCNPs under 795-nm excitation with/without 1140-nm depletion.

Download Full Size | PDF

For imaging application, the decay time of UCNPs is an important parameter. Here, to confirm the possibility of using these UCNPs for microscopy, the lifetime of green emission from the Yb³⁺/Er³⁺ UCNPs are measured following our previous protocol [Fig. 4(b)] [23]. The calculated lifetime is 55 µs and remains unchanged under 795-nm excitation with/without 1140-nm depletion. This is accorded to Eq. (16), which indicates that the applying 1140-nm laser would not change the decay time of green emission. Compared to previous work, this 55 µs decay time is acceptable for scanning imaging applications. The potential streak effect in the high speed scanning imaging can be eliminated by reduced lifetime material design (e. g. Controlling the nanoparticle size or the doping concentrations of lanthanide elements) or the addition of confocal pinhole [23–25 ].

In a previous work, the non-photobleaching and non-photoblinking advantages of UCNPs under single wavelength excitation was demonstrated theoretically and experimentally [26]. However, it still remains unclear whether these unique properties would change under double wavelength irradiation. The experimental method was illustrated in our previous work, and the results were shown in Fig. 5 [23]. The time traces of a single NaYF₄:Yb³⁺/Er³⁺ UCNP during 30-min under single [Fig. 5(a)] wavelength irradiation suggests that these UCNPs have very high photostability. In the case of two wavelength irradiation, the green luminescence also appears stable [Fig. 5(c)], while the slight unsteadiness could be ascribed to the power fluctuation of the OPO system, as the intensity occurrences still fitted the Gaussian function well [Fig. 5(d)], just as they did for single wavelength excitation [Fig. 5(b)]. To further confirm the non-blinking property, temporal resolution of 2 µs emission was recorded under single [Fig. 5(e)] and double [Fig. 5(g)] wavelength irradiation. No dark state of the green emission could be observed during 0.5 second, and the intensity occurrences still followed the Gaussian function distribution [Fig. 5(f), Fig. 5(h)]. From the experiments, the non-photobleaching and non-photoblinking properties of the Er³⁺-doped UCNPs remained unchanged under two wavelength irradiation. In fact, according to previous research, it was the steady state emission of multi-ions doped in a single UCNP that led to the non-photobleaching and non-photoblinking advantages [27]. This coincides with our experimental results. These unique properties enable UCNPs to be ideal candidates for STED-like microscopic imaging.

 

Fig. 5 Intensity fluctuations of green luminescence of a single NaYF₄:Yb³⁺/Er³⁺ UCNP under 30 mins or 0.5 second. Time trace recording of 30-min irradiation of a single UCNP under 795-nm excitation only (a) or 795-nm and 1140-nm co-irradiation (c). Experimental data and theoretical fitting of 30-min intensity occurrences of a single UCNP under 795-nm excitation (b) or 795-nm and 1140-nm co-irradiation (d). Time trace recording of 0.5-second irradiation of a single UCNP under 795-nm excitation only (e) or 795-nm and 1140-nm co-irradiation (g). Experimental data and theoretical fitting of 0.5-second intensity occurrences from a single UCNP under 795-nm excitation (f) or 795-nm and 1140-nm co-irradiation (h).

Download Full Size | PDF

4.5. Discussions and perspective

In single-photon STED microscopy, the excitation and depletion wavelengths are always in the visible region [28]. These wavelengths not only induce disturbing autofluorescence, but also suffer from strong scattering in biological tissues. This problem has to some extent been remitted by TPE-STED microscopy. These NIR excitation wavelengths, typically around 800 nm, have been applied in STED microscopy and increase the imaging depth from 120 µm to ~300 µm [5]. Previous investigations have shown that the fundamental depth limit for TPE microscopy is approximately five attenuation lengths, which means an ideal 750-µm-depth could be achieved using 800-nm excitation light [29]. However, it is still challenging to perform TPE-STED microscopy at such a large depth since the depletion wavelength is located in the visible region and suffers from strong scattering. To solve this problem, NaYF₄:Yb³⁺/Er³⁺ UCNPs is proposed as a novel potential probes of STED-like microscopy with depletion wavelength at 1140 nm. Compared with previously reported wavelengths of STED lasers, 1140 nm has a much longer attenuation length, as depicted in Fig. 6 [30, 31 ]. To the best of our knowledge, this is the first report where the wavelength of the depletion laser is longer than that of excitation in multi-photon super-resolution microscopy. Meanwhile, in previous work, an Electric field Monte Carlo (EMC) method was proposed to investigate the focal spot of STED beam [32]. The result indicates that a NIR STED beam would have much larger penetration depth in a turbid tissue due to lower scattering, similar to the case of Gaussian beam. It should be noted that the spherical aberration is also an influential factor for obtaining super-resolution images in deep tissue [28]. However, this problem can be solved by adding adaptive optical devices [33]. Since UCNPs have been successfully applied in deep tissue multi-photon microscopy, these nanoparticles, with proper hydrophilic and surface functionalize processing, would have much potential to obtain high quality super-resolution images in large depth regions of biological tissues [6, 17, 34, 35 ]. With non-photobleaching and non-photoblinking advantages, UCNPs are powerful candidates to perform longstanding three-dimensional nanoscopic imaging.

 

Fig. 6 Wavelength-dependent attenuation length spectrum of a brain tissue model based on Mie scattering and water absorption. This model was proposed by Horton, et al. [31] The absorption length of water was measured by Hale, et al. [36] Green solid and dashed arrows: excitation and depletion lasers wavelength of single photon excited STED microscopy of brain slice [28]. Yellow solid and dashed arrows: excitation and depletion lasers wavelength of TPE-STED microscopy of brain slice [37]. Pink solid and dashed arrows: potential excitation and depletion lasers wavelength of TPE-STED microscopy in this work.

Download Full Size | PDF

Note that an 1140-nm CW laser can be a more efficient device to optically deplete the upconverting luminescence. On one hand, as illustrated in Fig. 1, the FWHM of the 1140-nm laser was 18 nm, much larger than the gap between two confirmed real states. It is better to use a 1140-nm CW laser since more precise energy bandgap match would bring about less waste of the depleting laser energy. Previous research has provided an ideal laser ready for application, which would probably be commercially available in the future [38]. On the other hand, an 1140-nm CW laser could break the depletion efficiency limit resulting from the low-duty-cycle and high peak power of the fs laser. In addition, another advantage called non-wavelengthmixing, that is, the excitation, emission and depletion lasers are spectrally separated, indicats that sophisticated optics such as mirrors and filters would be simplified. What’s more, the ultrasmall upconverting nanodots have already been successfully synthesized and applied in protein-targeted imaging, which provides preliminary preparation for super-resolution microscopy in subcellular structures [23, 39 ]. Owing to the application of these materials, an all-CW-laser STED-like imaging system would help to extend the research and application of multi-photon super-resolution microscopy with low complexity and high cost-effectiveness performance.

4. Conclusion

In this work, an optical depletion mechanism of upconverting luminescence in NaYF₄:Yb³⁺/Er³⁺ UCNPs was proposed and investigated. The depletion mechanism, the combination of ESA and energy transfer between Yb³⁺ and Er³⁺, was theoretically designed and experimentally achieved under 795-nm and 1140-nm co-irradiation. The theoretical analysis based on steady-state rate equations also supported the mechanism and was further confirmed by the surprising emerging luminescence at 478 nm. The depletion efficiency of NaYF₄:Yb³⁺/Er³⁺ UCNPs was as high as 30%, which could be effectively optimized by employing an 1140-nm CW laser. A number of advantages of using NaYF₄:Yb³⁺/Er³⁺ UCNPs in potential multi-photon super-resolution microscopy, including suitable wavelengths with less scattering and absorption, the avoidance of photoblinking and photoblinking, and a simple and cost-effective optical system, are discussed in details. Future use of these nanoparticles in deep tissue multi-photon STED-like microscopy would be highly promising and attractive.