Upconversion Modulation through Pulsed Laser Excitation for Anti-counterfeiting

Lanthanide-doped upconversion nanomaterials are emerging as promising candidates in optoelectronics, volumetric display, anti-counterfeiting as well as biological imaging and therapy. Typical modulations of upconversion through chemical methods, such as controlling phase, composition, morphology and size enable us to rationally manipulate emission profiles and lifetimes of lanthanide ions by using continuous-wave laser excitation. Here we demonstrate that under pulsed laser excitation the emission color of NaYF4:Er/Tm (2/0.5%)@NaYF4 core-shell nanoparticles has an obvious transformation from green to red colors. Moreover, both pulse duration and repetition frequency are responsible for manipulating the upconversion emission color. The mechanism of the phenomena may be that the pulsed laser sequence triggers the emission levels to non-steady upconversion states first, and then cuts off the unfinished population process within the pulse duration. This pump source dependent and resultant tunable fluorescence emission enables NaYF4:Er/Tm (2/0.5%)@NaYF4 nanoparticles as a promising fluorophore in the transparent anti-fake printing.

(CW) laser with equivalent average power density, enabling deep tissue optical bioimaging 13,14 . However, new type of color-shifting UCNPs is still limited and comprehensive understanding of the upconversion modulation discipline by pulsed laser especially for its repetition frequency has not realized until now.
Herein we designed an active core-inert shell structured nanocrystal with the composition of NaYF 4 :Er/Tm (2/0.5%)@NaYF 4 to obtain an insight into the modulation mechanism on upconversion emission by pulse duration and repetition frequency, respectively. Selecting the Er/Tm ion pair was based on its pure red emission in hexagonal phase NaYF 4 nanoparticles under CW laser excitation. Actually, the GRR of nanoparticles with prominent green emission under CW laser excitation, such as β-NaYF 4 :Yb/Er (20/2%) nanocrystals, could also be changed under pulsed laser excitation. Nevertheless, the emission color is always located in the green region. In our design, under short pulse duration and low repetition frequency 980 nm near infrared (NIR) laser excitation, the as-synthesized NaYF 4 :Er/Tm (2/0.5%)@NaYF 4 nanoparticles exhibited green emission, which is different from its typical red emission under CW 980 nm laser excitation. We found that the GRR evidently changed with pulse duration and repetition frequency, respectively. This significantly improves the anti-fake quality when using NaYF 4 :2Er/0.5Tm@NaYF 4 as printing ink. Our proposed mechanism considered that the upconversion temporal population and depopulation proceeding which are influenced by pulse duration and repetition frequency give rise to combined impacts on the ultimate luminescence profile.
The typical upconversion spectra of the as-prepared three kinds of nanoparticles with equivalent mass concentration (1 mg/mL) were measured upon continuous 980 nm laser excitation (42 W/cm 2 ). As is shown in Fig. 2(a), the upconversion spectra of NaYF 4 :Er/Tm (2/0.5%) and NaYF 4 :Er/Tm (2/0.5%)@NaYF 4 upon CW 980 nm laser excitation gave rise to three dominant peaks at 520 nm, 541 nm and 654 nm, which are attributed to optical transitions of Er 3+ : 2 H 11/2 → 4I 15/2 , 4 S 3/2 → 4I 15/2 and 4 F 9/2 → 4I 15/2 , respectively. The doping Tm 3+ ions here interfere with Er 3+ ions' upconversion process by introducing a recycling energy transfer pathway, which contains transitions of 4 I 11/2 (Er 3+ ) + 3 F 4 (Tm 3+ ) → 4 I 13/2 (Er 3+ ) + 3 H 5 (Tm 3+ ) and 3 F 4 (Tm 3+ ) + 4 I 11/2 (Er 3+ ) → 4 F 9/2 (Er 3+ ) + 3 H 6 (Tm 3+ ), to facilitate red emission 15,16 (Fig. 2(d)). The overall emission intensity of NaYF 4 :Er/Tm nanocrystals was remarkably enhanced after NaYF 4 shell coating owing to elimination of quenchers on the surface of NaYF 4 :Er/ Tm nanoparticles 10,17 . Note that the enhancement factor for red emission is more prominent than that of the green emission, further elevating the red to green ratio. This effect implies that more activators participate in the recycling energy transfer procedure rather than are captured by surface quenchers after constructing the core-shell structure. Moreover, except for adapting inert shell (NaYF 4 shell layer) to improve photoluminescence efficiency, we also tried to coat active shell (NaYF 4 :Yb shell layer) onto the NaYF 4 :Er/Tm core. However, the Yb 3+ ions, which possess the largest cross-section at 980 nm among all lanthanide ions, absorbed pump photons first, and then transferred the energy to Er 3+ and Tm 3+ ions individually. As a result, the upconversion route and the emission spectra ( Fig. 2(b)) changed, thus violating our pre-demand of red emission color under CW laser excitation. Herein NaYF 4 :Er/Tm@NaYF 4 is the most suitable sample in our design. The slopes of ln-ln plots of emission intensity at 541 nm and 654 nm versus 980 nm excitation power density clearly indicate that the upconversion populations of 4 S 3/2 and 4 F 9/2 energy levels are both realized through two-photon upconversion process ( Fig. 2(c)).
As for the upconversion spectra measurement of NaYF 4 :Er/Tm@NaYF 4 nanoparticles under pulsed laser excitation, firstly the pulse duration time was set within the range from 100 µs to 6 ms, while the repetition frequency was fixed at 100 Hz. Figure 3(c) is the schematic diagram illustrating the experimental setup for upconversion spectra measurement. The pulse generator was used to shape 980 nm NIR laser with controllable pulse duration time and repetition frequency and afterwards send the modulated laser to upconversion nanocrystals, which generate spectrum information captured by the optoelectronic detector. It is clearly depicted in Fig. 3(a) that by gradually prolonging the pulse width the emission color was adjusted from green, yellow to red. We further checked the emitting color dependence on frequency under permanent pulse duration. It was found that the emission color was also tuned from green to red with the frequency increasing meanwhile the pulse duration was set as 100 µs (Fig. 3(b)). When the pulse duration was longer, the frequency adjusting effect on emission color displayed the same tendency but in a narrow scope. These results indicated that both pulse width and frequency were able to manipulate the emission color of NaYF 4 :Er/Tm@NaYF 4 nanoparticles and under excitation of short pulse duration and low frequency laser the green emission was 'extracted' from upconversion process while the red emission was suppressed. It was noted that when pulse duration was longer or the repetition frequency was higher, the emission intensity became stronger as a result of enhanced pump power density. Due to the special multicolor emission features in homogeneous materials, the core-shell nanoparticles were designed as transparent fluorescence ink for improving anti-counterfeiting level. A series of character patterns were firstly embed into a PMMA plate through laser ablation. The cyclohexane solution with NaYF 4 :Er/Tm@NaYF 4 dispersion was injected into a part of character patterns. In ambient lighting, all the character patterns seem identical. However, under 980 nm laser excitation (CW or long pulse), only the labelled patterns ('IAM') exhibited red emission ( Fig. 4). Significantly, as gradually adjusting the pulsed laser conditions, the 'IAM' patterns changed to yellow and green color. This unique property greatly enhanced the anti-counterfeiting level.
To understand color modulation mechanism through pulsed diode laser excitation, the time-dependent evolutions of emission intensity at 541 nm and 654 nm of as-prepared NaYF 4 :Er/Tm and NaYF 4 :Er/Tm@NaYF 4 nanoparticles were investigated under 980 nm laser excitation with 8 ms duration time and 20 Hz repetition rate. As is indicated in Fig. 3(d), in the pulse duration time the 541 nm and 654 nm emission of the as-synthesized nanoparticles rise to their steady states within several milliseconds, as the populations of the emission energy levels are involved with sequential optical transitions. We found that without Tm ions the upconversion sample would give rise to green emission color ( Figure S3). Moreover, after Tm ions doping the rising evolutions of green and red emission were more evidently separated ( Figure S4), which was probably ascribed to the fact that two energy transfer processes participate in 4 F 9/2 population process. The separation of rising section makes it possible to timely activate the emission levels using modulated pulsed laser. As a consequence, under short pulse width excitation and a relatively low frequency the green emission intensity is stronger than the red intensity. When the pulse duration is prolonged while fixing the repetition frequency at 100 Hz, the red emission follows a gradual approach to its steady equilibrium state, yielding enhanced emission intensity. Therefore, the emission color changed from green to yellow and finally to red with pulse duration increasing. The lifetime of upconversion nanocrystals can be calculated based on the decay curve. Typically, the lifetime curve is fitted with the equation: where I 0 is emission intensity at the initial decay time t 0 and τ is the calculated lifetime of the emission energy level. Therefore, the fitted lifetimes of red emission and green emission in NaYF 4 :Er/Tm@NaYF 4 are 323 µs and 198 us, respectively. It is worth noting that, as displayed in Fig. 3, lifetimes of 541 nm and 654 nm emissions are prolonged after shell coating, verifying the enhanced luminescence intensity in NaYF 4 :Er/Tm@NaYF 4 compared to NaYF 4 :Er/Tm. We next investigated how the repetition frequency influences GRR under fixed pulse duration. It is convinced that the pulse duration time is not the only reason for tuning the GRR, otherwise the GRR will be permanent with frequency changing. As is indicated by the luminescence decay curves in Fig. 3(d), the red emission and green emission take 4 ms and 2 ms for their complete deactivation to initial states after the excitation source is shut down. The depopulation time is independent on pulse width because it is the nature character of the energy level. For illustrating, our frequency-dependent emission experiments, shown in Fig. 5(a), indicate that the red emission intensity is more sensitive to frequency than the green under pulsed 980 nm NIR laser excitation. The frequency effect on red or green emission intensity can be divided into two sections, which are separated by the time points that the intervals between two pulses are just equal to their complete deactivation time. When the pulse width is 500 µs, at which point the green emission level is close to its maximum population whereas the red emission is at its initial population state (Fig. 3(d)), the critical frequencies for red and green emission are therefore 220 Hz (~1/(4 ms + 500 µs)) and 400 Hz (1/(2 ms + 500 µs)), respectively. It is found that separated by their critical frequencies the red emission and green emission intensities grow with different rising gradients. The regular pattern of intensity dependence on frequency is further verified by red to green ratio changing with frequency (shown in Fig. 5(b)). It can be concluded from Fig. 5(b) that the slope of red to green ratio are 4.27 × 10 −3 , 6.07 × 10 −3 and 4.81 × 10 −3 when the frequency located in the scale of <220 Hz, 220 Hz < frequency < 400 Hz and frequency > 400 Hz, respectively. Obviously, the slope of red to green ratio presented a slight decrease after a slight increase. This phenomenon can be interpreted by the fact that with frequency increasing the subsequent pulse first reaches the complete deactivation point of red emission level and afterwards the red emission intensity begins to increase faster whereas the green emission intensity evolution is still at its slow increment scale (<400 Hz). When the repetition frequency exceeds 400 Hz, the green emission intensity starts to rise faster leading to depression of red to green ratio slope.

Conclusion
Traditional photon upconversion in lanthanide ions doped materials are usually realized through CW laser excitation, thus the emission color of one fixed type nanoparticle is typically monotonous merely with minor change. We here demonstrated that an identical sample, NaYF 4 :Er/Tm (2/0.5%)@NaYF 4 core-shell structured nanocrystal, could display green, yellow and red emission colors, which depends on elaborately manipulating the pulse duration or repetition frequency of the pumping laser. The microsecond-scale short pulse duration of pumping laser triggered population separation of green and red emission energy levels before reaching steady states due to their differentiable rising time. And the pulse repetition frequency affects emission evolution with different tendency depending on whether the time between two pulses is longer than its absolute decay time or not. Therefore, we attributed the upconversion emission modulation to a synthetical effect derived from rising and decay nature of emission levels, which are influenced by the pulsed laser. Our study showed that the NaYF 4 :Er/Tm@NaYF 4 core-shell upconversion nanocrystal could be a promising fluorescent substance for invisible and color-tunable anti-counterfeiting ink.

Synthesis method.
The synthesis procedure is similar to that utilized in previous work 18 only with minor modifications.
Synthesis of NaYF 4 : Er/Tm (2/0.5 mol%) Nanoparticles. In a typical experiment, 2 mL of water solution containing Y(CH 3 CO 2 ) 3 (0.39 mmol), Er(CH 3 CO 2 ) 3 (0.008 mmol) and Tm(CH 3 CO 2 ) 3 (0.002 mmol) were pipetted into a 100-mL flask, then 3 mL oleic acid and 7 mL 1-octadecene were added into the flask. The resulting mixed liquid was heated at 150 °C for 1 h with stirring and then cooled down to room temperature. Thereafter, 2 mL methanol solution containing NaOH (0.5 mmol/mL) and 2 mL methanol solution with NH 4 F (0.4 mmol/mL) were added into the flask and stirred at 40 °C for 30 min, after which time the mixture was heated to 90 °C to remove the methanol. After eliminating bubbles in the liquid, the solution was heated to 290 °C and kept at this temperature for 1.5 h in an argon air atmosphere. Finally, the mixture was cooled down slowly to room temperature. The resulting liquid was washed with ethanol and cyclohexane for 3 times, collected by centrifugation, and re-dispersed in 5 mL of cyclohexane.
Synthesis of NaYF 4 :Er/Tm (2/0.5 mol%)@NaYF 4 :Yb (x mol%) (x = 0, 20) Core-Shell Nanoparticles. Firstly, 1 mL of water solution containing Y(CH 3 CO 2 ) 3 and Yb(CH 3 CO 2 ) 3 with a total lanthanide amount of 0.2 mmol, were pipette into a 100-mL flask. The precise amounts of Y(CH 3 CO 2 ) 3 and Yb(CH 3 CO 2 ) 3 were calculated depending on their mole ratio in the shell layer. And then we added 3 mL oleic acid and 7 mL 1-octadecene into the flask. The resulting mixture was heated at 150 °C for 1 h with stirring and then cooled down to room temperature.  Subsequently, we added as-synthesized 2.5 mL cyclohexane solution containing NaYF 4 :Er/Tm nanoparticles and methanol solution of NaOH (1 mL, 0.5 mmol/L) and NH 4 F (2 mL, 0.4 mmol/L) into the flask. The following treatment to the mixture is the same as the method for synthesizing NaYF 4 :Er/Tm nanoparticles except for dispersing the product in 2.5 mL cyclohexane finally.
Synthesis of NaYF 4 :Er (2 mol%) and NaYF 4 :Er (2 mol%)@NaYF 4 Nanoparticles. To understand the upconversion process, these two types of nanoparticles were prepared using the same method displayed above.
Data availability statement. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Characterization. XRD patterns were obtained by a Rigaku D/max 2550 X-ray diffractometer using CuKa radiation (λ = 0.154 nm). TEM images were performed on a Hitachi 7700 transmission electron microscope operating at 200 kV. UC luminescence spectra and decay curves were recorded on a Fluorolog ® -3 Spectrofluorometer by Horiba using an external (continuous or pulsed) 980 nm diode laser as the excitation source. All measurements were taken under room temperature.

Anti-counterfeiting Preparation.
To show the anti-counterfeiting ability of NaYF 4 :Er/Tm@NaYF 4 nanocrystals, the cyclohexane solution containing NaYF 4 :Er/Tm@NaYF 4 was used as ink to be injected into the character marks in the polymethyl methacrylate (PMMA) plate. The security information was read out by a camera when using pulsed 980 nm laser to excite the labeled characters.