Effect of Pr³⁺ concentration in luminescence properties & upconversion mechanism of triple doped NaYF₄: Yb³⁺, Er³⁺, Pr³⁺

The absorption spectra were utilized to confirm the presence of pure Yb³⁺, Er³⁺, and Pr³⁺ dopants and to understand the energy levels and transitions in Yb³⁺, Er³⁺, Pr³⁺ triple-doped NaYF₄ samples, thereby providing valuable information about the electronic structure of the material. A Perklin-Elmer UV–vis-NIR Lambda 35 spectrophotometer was used to collect the optical absorption spectra. The powder samples were placed and examined in a sample holder with a quartz window. The range of wavelengths that were measured fell within the spectrum of 190 nm to 1100 nm. The absorption spectroscopy for the Yb³⁺, Er³⁺, Pr³⁺ triple doped NaYF₄ samples were investigated with and without Er³⁺ dopant. The absorption spectra of samples 20Yb1Pr and 20Yb2 Er1Pr are shown in figure 3.

Zoom In Zoom Out Reset image size

The intense peak at 976 nm is common for both spectra and originates from ²F_5/2→²F_7/2 absorption of Yb³⁺. For the blue lined spectrum (20Yb1Pr), the absorption bands located at 445, 469, 483, 587 nm correspond to the transitions from the ground state ³H₄ to specific excited states ³P₂, ¹I₆ + ³P₁, ³P₀ and ¹D₂ of Pr³⁺, respectively. For the red lined spectrum (20Yb2Er1Pr), in addition to the aforementioned transitions belonging to the Pr³⁺ the absorption bands located at 520, 540, and 654 have been observed, which can be assigned to ²H_11/2→⁴ I_15/2, ⁴S_3/2→⁴I_15/2, and ⁴F_9/2→⁴I_15/2 transitions of Er³⁺, respectively [30].

Figure 4 exhibits the visible and near-infrared emission spectra of NaYF₄: Yb³⁺, Er³⁺, Pr³⁺ nanocrystals excited by a 980 nm laser diode. Emission spectra in the visible region were observed at 377, 407, 465, 488, 520, 540, and 654 nm, which can be attributed to specific transitions of Er³⁺energy levels such as ⁴G_11/2→⁴I_15/2, ²H_9/2→⁴I_15/2, ⁴F_5/2→⁴I_15/2, ⁴F_7/2→⁴I_15/2, ²H_11/2→⁴I_15/2, ⁴S_3/2→⁴I_15/2, and ⁴F_9/2→⁴I_15/2, respectively (inset figure 4). When Pr³⁺ ions were doped into the 20Yb2Er sample, all peak intensities were quenched with the increase in the Pr³⁺ dopant concentration, shown in figure 4. However, no emission peaks belonging to Pr³⁺ ions were observed at the visible region originating from the ³P_0,1→³H_4–6, ³F_2–4,¹D₂→³H₄ transitions [31].

Zoom In Zoom Out Reset image size

In the NIR region, depicted in figure 5, the luminescence intensity at 1530 nm (Er³⁺:⁴I_13/2→⁴I_15/2) was significantly quenched and weak emission intensity at 1290 nm (Pr³⁺:¹G₄→³H₅) changes irregularly upon doping with Pr³⁺ions. It is evident that in the population of ⁴I_13/2 and ¹G₄ levels, cross relaxations seem to be the main process affecting the emission from these levels.

Zoom In Zoom Out Reset image size

In order to comprehend the upconversion mechanism of our samples, we analyzed the excitation power dependence of UC emission intensities for NaYF₄: Yb³⁺, Er³⁺, Pr³⁺ samples excited at various pump powers using the 980 nm diode laser and the pump power dependent upconversion luminescence spectra are recorded, either for emissions in visible or NIR regions, as depicted in figures 6 and 7, respectively. The relation between upconversion emission intensity and excitation power was established using equation (2),

$$\begin{eqnarray}&&I\,\alpha \,{P}^{n}\end{eqnarray} \tag{ 2 }$$

Where I is the emission intensity, P is the power of excitation and n is the number of photons involved in the upconversion process. The value of n can be ascertained by analyzing the slope value of a linear fitting in a logarithmic-logarithmic plot of equation (2).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In the visible region, nanoparticles under a 980 nm laser diode indicated that the number of photons involved in the upconversion process was found to be closely related to the concentration of dopant ion, as the increment in concentration can either promote or impede the upconversion process. This relationship is illustrated in figures 6(a) and (b), which shows the impact of increasing the Pr³⁺ concentration from 0.1% to 1%, respectively.

Figure 6(a) shows that for the 20Yb2Er0.1Pr sample, the slope values for UC emission peaks at 407, 488, 520, 540, and 654 nm were 3.11, 2.53, 2.54, 2.53, and 2.66, respectively. Since the slope values are approximately three, the upconversion mechanism should be a three-photon process. Subsequently, upon adding 1 mol% of Pr³⁺ ions, the slope values decreased to 2.82, 2.81, 2.25, 2.23, and 2.23 for the 407, 488, 520, 540, and 654 nm emissions, respectively, as demonstrated in figure 6(b). The slope value for 407 and 488 nm emission is close to 3, indicating that three-photon processes are involved in populating the ²H_9/2 and ⁴F_7/2 levels. For the 520, 540, and 654 nm emissions, the slope values are close to 2, indicating two-photon processes are involved in populating the ²H_11/2, ⁴S_3/2, and ⁴F_9/2, respectively. The decrease in slope values for 520, 540, and 654 nm emissions from three to two photons can be attributed to the transitions ⁴F_7/2→³P₀ and the following multiphonon relaxations (MPRs) and cross-relaxation (CR) processes. Additionally, competition between the energy transfer from ⁴I_13/2 (Er³⁺) to ³F_3,4 (Pr³⁺) level and upconversion processes pumping to ⁴F_7/2 where the depletion of intermediate excited states of ⁴I_13/2 level might be effective [32].

In the NIR region, 0.1% and 1% Pr³⁺ doped NaYF₄: Yb, Er samples were examined for 1290 nm and 1530 nm emissions at different pump powers of the 980 nm diode laser and the pump power-dependent upconversion luminescence spectra is recorded as shown in figures 7(a), (c). The log–log plots for 1290 nm and 1530 nm emissions, logarithmic upconversion emission intensity versus logarithmic excitation power density graphs, are illustrated in figures 7(b), (d). For the 1290 nm emissions, the slope value of the 20Yb2Er0.1Pr sample was found to be 0.58, as shown in figure 7(a). As Pr³⁺ ion concentrations increased in the system, the slope value of the 20Yb2Er1Pr sample decreased to ∼0.08, as depicted in figure 7(d). These results suggest that both slope values correspond to the single photon process for populating the ¹G₄ level. The evidence for 1530 nm emission from the log–log plot exhibits the slope value of 1.30 and 0.24 for the 20Yb2Er0.1Pr and 20Yb2Er1Pr samples, respectively. The decrease in slope values observed as Pr³⁺ ion concentration increases, shown in figures 7(b), (d). The variation in slope value for 1290 nm and 1530 nm depending on Pr³⁺ concentration is given in figure 7(e). The rate of decreasing slope value at 1530 nm is greater than 1290 nm, indicating a faster depletion for the ⁴I_13/2 level compared to the ⁴I_11/2 level. Consequently, the energy transfer rate of ⁴I_13/2→³F_3,4 is faster than ⁴I_11/2→¹G₄.

Decay time analysis was conducted on the samples to achieve a better understanding of the energy transfer and upconversion mechanism. Samples with increased concentration of Pr³⁺ dopant were examined under 980 nm pulsed laser excitation. To demonstrate the concentration-dependent photoluminescence dynamics of the Pr³⁺ dopant, decay time analysis was monitored at 520 nm (figure 8(a)), 540 nm (figure 8(b)), 1290 nm (figure 8(c)), and 1530 nm (figure 8(d)). The emission lifetimes (τ) were calculated using a single exponential decay function (equation (3)) and results presented in table 3.

$$\begin{eqnarray}&&I\left(t\right)={I}_{1}{\exp }^{(-t/{\tau })}\end{eqnarray} \tag{ 3 }$$

where I(t) represent the photoluminescence intensity at a specific time point corresponding to the on and off of the pulsed laser. τ is the decay lifetime of the upconversion luminescence.

Zoom In Zoom Out Reset image size

The decay time at 520 nm for the sample without Pr³⁺ ions was calculated to be 111.98 μs, which originated from the ²H_11/2 energy level of the Er³⁺ ion. As the Pr³⁺ concentration was increased to 0.1% in the second sample, decay time decreased to 92.61 μs. With further increases in the Pr³⁺ concentration to 0.5% and 1%, decay times decreased to 72.42 μs and 62.55 μs, respectively. The decay time at 540 nm presents similar behavior with increasing Pr³⁺concentration, as shown in table 3. It is expected that the decrease in the lifetime of ²H_11/2 and ⁴S_3/2 levels might be caused by an energy transfer of Er³⁺:⁴F_7/2→ Pr³⁺: ³P₀.

In the NIR region, decay time was studied as well. The decay time for the samples without Pr³⁺ and with 0.1% Pr³⁺ was not detected at 1290 nm. By increasing the Pr³⁺ concentration to 0.5% Pr³⁺ and 1% Pr³⁺, the decay time is determined to be 32.63 μs and 40.52 μs, respectively.

This increment in decay times at 1290 nm for the samples with varying Pr³⁺ concentrations demonstrates the existence of energy transfer (⁴I_11/2→¹G₄). On the other hand, the decay time at 1530 nm (⁴I_13/2→⁴I_15/2) for all samples demonstrated an irregular behavior as Pr³⁺ concentration increased (table 3). Hence, there should be another mechanism explaining this irregularity in lifetime at 1530 nm.

Based on these findings, the energy transfer from Er³⁺ to Pr³⁺ can be estimated. As deduced from decay time and power-dependent analysis, it is possible to figure out the upconversion mechanism, as shown in Scheme 1.

Zoom In Zoom Out Reset image size

The energy level diagram, Scheme 1, presents possible energy transfer mechanisms of triple doped NaYF₄: Yb³⁺, Er³⁺, Pr³⁺ systems. Upon 980 nm excitation, emissions observed in visible regions including blue (407, 465 and 488 nm), green (520 and 540 nm), red (654 nm), and near-infrared (845, 1290, and 1530 nm) region can be attributed to the population of ⁴G_11/2 and ⁴F_7/2 states of Er³⁺ either through the excited state absorption (ESA) or energy transfer upconversion (ETU) or both. First, Yb³⁺ absorbs near-infrared (980 nm) photons and is excited from the ground state to the ²F_5/2 level. Upon NIR excitation, the intermediate state of Er³⁺ (⁴I_11/2 ) can be populated either through energy transfer (ET1) or ground state absorption (GSA). Then a second 980 nm photon, or energy transfer from a Yb³⁺ ion (ET2), can populate the ⁴F_7/2 level of the Er³⁺ ion. Subsequently, the Er³⁺ ion can relax non-radiatively to the ²H_11/2, ⁴S_3/2 and ⁴F_9/2 levels leading to green ²H_11/2→⁴H_15/2 (520 nm), ⁴S_3/2 → ⁴I_15/2 (540 nm), red ⁴F_9/2 →⁴I_15/2 (654 nm) and infrared ⁴I_9/2→⁴I_15/2 (845 nm) emissions. Alternatively, the Er³⁺ ion in the ⁴I_11/2 state may nonradiatively decay to the ⁴I_13/2 state and then populate the ⁴G_11/2 state via ET3 or following ESAs. Simultaneously, this state relaxes to ²H_9/2,²H_11/2, ⁴S_3/2, ⁴F_9/2, and ⁴I_9/2 levels by MPRs, generating blue, green, red, and infrared emissions as well. Meanwhile, the ⁴F_7/2 state can be de-excited via ET4 and transfer its energy to the ³P₀ state. It is expected that, if an energy transfer ET4 exists there should be emissions stemming from transitions due to the ³P_0,1→³H_4–6, ³F_2–4, and ¹D₂→³H₄. However, no visible emissions are observed related to those transitions. Even if ET4 exists, visible emissions could be hampered through numerous cross-relaxations (CR1, CR2, and CR3) and MPR processes [31, 33].

Later on, both cross relaxations and ET5 (Er³⁺:⁴I_11/2 →Pr³⁺:¹G₄) populate the ¹G₄ level, and weak emission is generated at 1290 nm as a result of ¹G₄→³H₅ transitions. In the CR2 process, Pr³⁺ ions relax from the ¹D₂ to ¹G₄ level and promote the ⁴I_13/2 (Er³⁺) level resulting in 1530 nm emission. The near-resonant ET6 from Er³⁺: ⁴I_13/2 →Pr³⁺:³F₃, causes the quenching of 1530 nm emission through the MPR process [34].

Herein, the quenching in blue (407 nm), green (520 and 540 nm), red (654 nm) and near-infrared (845 nm) emissions can be attributed to the decrease in the ⁴I_13/2 population. The decrease in the ⁴I_13/2 population leads to the decrease in the ⁴F_9/2 population through ET2 and subsequently followed by the decrease in the ⁴G_11/2 population through ET3. Two mechanisms affect the population of the ⁴I_13/2 level. One of the mechanisms is the CR2 process that is highly affected by the particle size and Pr³⁺ concentration. The other mechanism is the ET6 energy transfer process affected by only Pr³⁺ concentration. The probability of the CR2 relaxations in smaller particles is quite high due to the random distribution of rare earth ions. While the probability of CR2 relaxation is quite less in the larger particles due to the strict distribution of rare earth ions [35]. To better understand the CR2 mechanism, populating ⁴I_13/2, we have analyzed the intensity ratio of R/G, as shown in table 4.

Comparing 20Yb2Er (23.66 nm) and 20Yb2Er0.1Pr (31.08 nm), as particle size increases R/G ratio decreases from 0.407 to 0.102, indicating the decrease in CR2 relaxations depopulating the ⁴I_13/2 level. Pr³⁺ concentration is another effect that populates ⁴I_13/2 level through the CR2 process. However, the CR2 processes that are populating ⁴I_13/2 level are negligible for small concentrations (0.1 mol % Pr³⁺ doped sample) [36]. Notably, the cross relaxations stemming from Er³⁺ ions are ignored as the Er³⁺ concentration is constant for all samples. The ET6 process mitigating emission intensity due to non-radiative decays is negligible for the 0.1 mol % Pr doped sample as well. Overall, the less likely CR2 relaxation process results in a depopulation of the ⁴I_13/2 level, giving rise to a shorter decay time from 34.89 μs to 19.83 μs.

By increasing the Pr³⁺ concentration to 0.5 mol% (sample 20Yb2Er0.5Pr), both decrease in particle size to 21.18 nm and increase in Pr³⁺ concentration are effective on the increase in CR2 relaxations populating ⁴I_13/2 level. Although the rise in ET6 process in parallel with the Pr³⁺ concentration depopulates ⁴I_13/2 level. Yet, the CR2 processes populating ⁴I_13/2 levels are dominant leading to an increase in intensity ratio of R/G to 0.144. Hence, a longer decay time was observed from 19.83 μs to 31.93 μs.

Further increasing Pr³⁺ concentration to 1 mol% (sample 20Yb2Er1Pr), particle size increased to 22.48 nm leading to a decrease in CR2 relaxations but their contribution is negligible as particle sizes are almost equal. The increase in Pr³⁺ concentration causes an increase in CR2 relaxations, which populate the ⁴I_13/2 level. However, due to the decrease in the distance between the lanthanide ions, ET6 processes increase and, as a result, ⁴I_13/2 is depopulated. The population of ⁴I_13/2 with CR2 process is estimated to be dominant over the depopulation of ⁴I_13/2 with ET6 process, causing a further increase in the intensity ratio of R/G to 0.18 and, Hence, a longer decay time observed from 31.93 μs to 38.04 μs.

Meanwhile, an energy transfer ET5 may occur from Er³⁺ (⁴I_11/2) to Pr³⁺ (¹G₄) level due to their well-alignment. This transfer leads to a weak emission at 1290 nm, corresponding to the transitions of ¹G₄→³H₅. The observed emission at 1290 nm seems to be affected by CR2 relaxations enhancing the population of ¹G₄ level as well. As stated before, particle size and Pr³⁺ concentrations are the factore that highly affect the CR2 relaxation process [35, 36]. Herein, it is estimated that within the CR2 process, particle size is more effective than Pr concentration. Even if Pr³⁺ concentration is a significant factor in the population of ¹G₄ level through CR2, the gradual increment in 1290 nm would be anticipated in parallel with Pr³⁺ concentration. Nevertheless, the results show that the enhancement of ¹G₄→³H₅ transitions is more associated with particle size rather than Pr³⁺ concentrations.