As the global market grows rapidly, forgery of products or intellectual property becomes a big problem. Companies are trying to protect their products from counterfeiting. As a result, it is very important for a company to develop its own anticounterfeiting technology. Upconversion (UC) phosphors, which convert near-infrared (NIR) to visible light, have attracted considerable attention as potential security materials because they can easily distinguish counterfeit products from original products using self-emission under NIR illumination [1-5]. The state of the art on UC phosphor technology is well introduced in recent literature [6, 7]. Most UC materials contain lanthanide ions (Ln³⁺) as activators and sensitizers. These lanthanide ions are incorporated into inorganic hosts and play a key role in determining the luminescence color and efficiency. The composition of host materials also directly affects the UC emission color or efficiency even if the activator and sensitizer ions are same. Thus, many efforts have been dedicated to find novel hosts with the intention of achieving an improved UC performance [8-11].

Lanthanide ions such as Er³⁺, Ho³⁺, Tm³⁺ and Yb³⁺ have been used as the dopants for UC materials. Yb³⁺ is an excellent sensitizer because it has a large absorption crosssection at 980 nm compared with other Ln³⁺ ions (Ln³⁺ = Er³⁺, Ho³⁺ and Tm³⁺) [12]. Therefore, ion pairs such as Er³⁺/Yb³⁺, Ho³⁺/Yb³⁺ and Tm³⁺/Yb³⁺ are used to improve the UC luminescence. Host materials for UC phosphors are required to have high transparency, good optical or chemical stability and low phonon energy [6]. Given this, CeO₂ is a good host because it has high transparency in the visible region, excellent chemical stability, and low photon energy (~470 cm^-1) [13]. Thus, ceria has been used in a number of research fields including catalysts, phosphors, and sensors [14-16]. Also, lanthanide ion radii are close to the ionic radius of Ce⁴⁺ so that the dopants can be easily substituted into the lattice of CeO₂. There are a number of studies that have focused on Er³⁺ or Er³⁺/Yb³⁺-doped CeO₂ characteristics [13, 17-19]. CeO₂:Er³⁺/Yb³⁺ has both green and red UC emissions, and the emission ratio of green to red is controllable by changing the concentration of Yb³⁺. Similarly to the Er³⁺/Yb³⁺ couple, Ho³⁺/Yb³⁺ couples can also generate green or red depending on the type of host material. Referring to the previous studies including Y₂O₃ [20-22] and Gd₂O₃ [23, 24], using a Ho³⁺/Yb³⁺ pair rather than an Er³⁺/Yb³⁺ ion pair can be more beneficial for implementing high intense green UC phosphors. Nevertheless, there are not much studies on the synthesis and UC characteristics of CeO₂:Ho³⁺/Yb³⁺ compared with CeO₂:Er³⁺/Yb³⁺. In 2010, for the first time, Babu et al. reported the green emission of CeO₂:Ho³⁺/Yb³⁺ and applied it to bio-imaging and therapeutics [25]. Avram et al. investigated the dependence of the UC emission color of CeO₂:Ho³⁺/Yb³⁺ on the excitation wavelength, and they observed the tunable ratio of red to green while increasing the powder density of an ac laser diode at 980 nm [26].

Although host and activator are the same, the luminescence properties of phosphor particles depend on their size and distribution, morphology, activator concentration, and activator distribution within hosts. To achieve good performance in various application fields, phosphor particles are needed to have fine size (about 1 μm) and spherical morphology. Also, high phase purity or uniform distribution of activators is critical to achieve high luminescence. Therefore, it is important to develop a synthesis method that can produce phosphor with excellent optical and morphological properties. Spray pyrolysis has been applied to prepare various functional materials [27-30]. One particle comes from one droplet in the spray pyrolysis. All elements consisting of phosphor can exist a homogeneously mixed state in the produced particle in a molecular level as long as no phase separation occurs during the drying step. Consequently, the spray pyrolysis can easily produce finesized spherical particles having homogeneous distribution of activators within host particles. Therefore, spray pyrolysis is a good synthesis method that can prepare multi-component phosphors such as BaMgAl₁₀O₁₇:Eu, Y₃Al₅O₁₂:Ce and (Y, Ln)VO₄:Eu into spherical shape and fine size [31-33]. Nevertheless, to our best knowledge, there is no report on the synthesis of CeO₂:Ho³⁺/Yb³⁺ using the spray pyrolysis process. In this work, the spray pyrolysis was applied to prepare CeO₂:Ho³⁺/Yb³⁺ having high UC emission to be identified by the naked eye. To do this, the UC properties were optimized by controlling the Ho³⁺ and Yb³⁺ concentration. The crystallinity of CeO₂:Ho³⁺/Yb³⁺ was controlled by changing the calcination temperature from 900°C to 1200°C in order to find the relationship with the UC intensity. Finally, the UC mechanism of CeO₂:Ho³⁺/Yb³⁺ prepared by spray pyrolysis was investigated by monitoring the dependency of the UC intensity on the IR pumping power while changing the activator concentration.

The CeO₂:Ho³⁺/Yb³⁺ particles were synthesized by a spray pyrolysis process consisting of an ultrasonic nebulizer (1.7 MHz), a quartz tube (ID = 55 mm and length = 1200 mm), an electrical furnace and a Teflon bag filter. Cerium nitrate hexahydrate (Aldrich, 99.99%), holmium oxide (Aldrich, 99.9%) and ytterbium oxide (Aldrich, 99.99%) were used as the starting materials. Spray solution was prepared by the following procedure. The activator (Ho³⁺) and sensitizer (Yb³⁺) precursors were dissolved by using nitric acid as a nitrate form and followed by adding the purified water until the total solution becomes 250 mL. Cerium nitrate was dissolved in 250 mL of purified water and mixed with the activator and sensitizer solution. The total salt concentration was kept at 0.2 M. The Ho (x) and Yb (y) contents in Ce_1-x-yO₂:Ho_x/Yb_y were changed from x = 0.001 to x = 0.007 and from y = 0.005 to y = 0.060. The prepared precursor solution was turned into droplets using an ultrasonic nebulizer and carried into the quartz reactor (900°C) by air (30 L/min). The produced particles were collected by a Teflon bag filter and calcined at the temperature range from 900 to 1200°C for 3 h in an air environment.

The UC emission of all samples was measured using a spectrophotometer (PerkinElmer, LS 50) under the excitation of 980 nm IR laser (Optoenergy, PL980P330J). The dependency of UC emission on the pumping powder was monitored by varying the pumping power from 100 to 1000 mW. The crystal phase of CeO₂:Ho³⁺/Yb³⁺ powder prepared was identified by X-ray diffraction (XRD, Rigaku, MiniFlex600) measurement. The morphology of the prepared particles were identified by using high-resolution scanning electron microscopy (HR-SEM, Hitachi S4800) at the Korea Basic Science Institute (KBSI).

Figure 1 shows the UC emission spectrum and energy level diagram of CeO₂:Ho³⁺ (x = 0.005) and CeO₂:Ho³⁺/Yb³⁺ (x = 0.005, y = 0.005) particles prepared by spray pyrolysis and calcined at 1000°C. The observed three peaks are due to the ⁵F₄ /⁵S₂→⁵I₈ (550 nm, green), ⁵F₅ →⁵I₈ (670 nm, red) and ⁵F₄ /⁵S₂→⁵I₇ (760 nm, NIR) transition of Ho³⁺ ions [26]. Figure 1(b) shows the energy level diagram of Ho³⁺ and Yb³⁺. The absorption cross section of Yb³⁺ ions is much larger than that of Ho³⁺ ions. Thus, the Yb³⁺ ions absorbs much more photons than the Ho³⁺ ions. The ²F_5/2 energy level of Yb³⁺ is located slightly higher than the ⁵I₆ level of Ho³⁺. Thus, the energy transfer (ET) from Yb³⁺ to Ho³⁺ is possible and can take part in both ground state adsorption (GSA) and excited state adsorption (ESA) processes. The photons populated in the ⁵I₆ state of Ho³⁺ through the GSA process can be further excited to the ⁵F₄ /⁵S₂ level via the ESA or ET processes. The photons excited to the ⁵F₄ /⁵S₂ level decay to the ⁵I₈ and ⁵I₇ states of Ho³⁺, emitting green and NIR light, respectively. A part of photons in the ⁵F₄ /⁵S₂ level can also relax nonradiatively to the ⁵F₅ level and radiatively to the ground state (⁵I₈), which corresponds to the red emission at about 670 nm. The red emission can be also possible through the consecutive path of GSA (⁵I₈→⁵I₆), non-radiative relaxation (⁵I₆→⁵I₇), ESA (⁵I₇→⁵F₅ ) and radiative decay (⁵F₅ →⁵I₈). For CeO₂:Ho³⁺, as shown in Fig. 1(a), the green emission is strong, but other two (red and NIR) peaks are weak. That is, the UC emission of CeO₂:Ho³⁺ is achieved mainly through the ⁵F₄ /⁵S₂→⁵I₈ transition after the ground-state absorption (GSA) (⁵I₈→⁵I₆) and excitation-state absorption (ESA) (⁵I₆→⁵F₄). The UC emission of CeO₂:Ho³⁺/Yb³⁺ is much higher than that of CeO₂:Ho³⁺. Given this, the Yb³⁺ co-doping leads to the enhanced UC intensity due to an effective energy transfer from Yb³⁺ to Ho³⁺.

The UC properties of CeO₂:Ho³⁺/Yb³⁺ prepared by spray pyrolysis were investigated by changing the calcination temperature. Figure 2(a) shows the UC spectra of CeO₂:Ho³⁺/Yb³⁺ particles and the green emission intensity at 550 nm is shown in Fig. 2(b) as a function of the calcination temperature. There is no difference in the emission spectrum except the intensity with changing the calcination temperature, indicating that the main UC path of CeO₂:Ho³⁺/Yb³⁺ is not affected. The UC emission of the CeO₂:Ho³⁺/Yb³⁺ sample calcined at 1200°C was photographed under irradiation of 980 nm IR. As shown in the inset of Fig. 1(a). The CeO₂:Ho³⁺/Yb³⁺ particles have excellent green emission. The emission intensity increases gradually as the calcination temperature increases, which is mainly due to the increase of crystallinity. To confirm this, XRD measurements were carried out. Figure 2(c) shows XRD patterns of the prepared CeO₂:Ho³⁺/Yb³⁺ samples. All peaks are in good agreement with the cubic phase of CeO₂ and no impurity phase is observed regardless of the calcination temperature. For solid oxides, the crystallinity can be evaluated from the crystallite size. That is, the larger the crystallite size, the higher the crystallinity. The crystallite size was calculated by Scherrer’s equation, and the resulting values are shown within Fig. 2(c). The crystallite size increases as the calcination temperature increases, indicating that the crystallinity is gradually improved. Figure 2(d) shows the dependence of the emission intensity on the crystallite size. The emission intensity has a linear relation with the crystallite size. This result supports that the enhancement in the crystallinity is critical to improve the emission intensity.

Figure 3 is SEM photos of CeO₂:Ho³⁺/Yb³⁺ particles prepared by changing the calcination temperature from 900 to 1200°C. The formation mechanism of CeO₂:Ho/Yb hollow particles in spray pyrolysis is also shown in Fig. 3(f). The as-prepared particles have spherical shape that is maintained even after the calcination at 1200°C. The prepared particles show a hollow structure that is frequently encountered in a conventional spray pyrolysis process. No significant change in the particle morphology was observed by increasing the calcination temperature. As shown in Fig. 3(f), the spray pyrolysis produces particles through drying, precipitation and pyrolysis (thermal decomposition). The fast drying of droplets increases the surface concentration of droplets. As a result, the salt concentration on the surface of the water droplet reaches the supersaturation point, so surface precipitation occurs and forms a shell layer. Thereafter, as the water drying progresses further, the precipitation proceeds on the inner surface of the initially formed solid shell. In this step, evaporated water molecules escape solid layers, making the precipitated layer porous. Next, thermal decomposition (pyrolysis) takes place, producing many gas molecules. Because the solid layer is porous, these gas molecules can be ejected outside the solid layer without the deformation of the particle. Finally, the shell layer is turned into nano-sized crystals by the high-temperature calcination, which can be identified in Fig. 3(e). The obtained CeO₂:Ho³⁺/Yb³⁺ particles are about 1 μm in size and have a hollow structure. Based on the SEM result, it was confirmed that spherical and fine-sized CeO₂:Ho³⁺/Yb³⁺ particles with high green UC emission could be synthesized by spray pyrolysis.

Figure 4 shows the effect of the Ho³⁺ and Yb³⁺ content on the UC properties of CeO₂:Ho³⁺/Yb³⁺ particles prepared by spray pyrolysis. First, at the fixed Ho³⁺ content (x = 0.005) the UC emission is monitored by changing the Yb³⁺ content (y) from 0.005 to 0.06 and the resulting spectra are shown in Fig. 4(a). The UC intensity was significantly improved by the introduction of Yb³⁺, and the highest emission intensity was observed at y = 0.02. There was no significant improvement in the red or NIR emission. Next, the UC emission is examined by changing the Ho³⁺ content (x) at y = 0.02 and the resulting spectra are shown in Fig. 4(b). The highest UC emission was obtained at x = 0.003. Figure 4(c) shows the effect of both Ho³⁺ and Yb³⁺ contents on the UC emission intensity (λ_em = 553 nm). From these results, the optimum contents of Ho³⁺ (x) and Yb³⁺ (y) to achieve the highest UC intensity were determined as x = 0.003 and y = 0.02.

The luminescence intensity of CeO₂:Ho³⁺/Yb³⁺ is reduced when the Ho³⁺ content is larger than 0.003 and the Yb³⁺ content is larger than 0.02. The concentration quenching of the emission intensity in phosphor materials is basically connected with the non-radiative energy transfer between dopants. The critical distance (R_c) between dopants for the non-radiative energy transfer can be calculated by the following equation [34].

where z_c is the critical concentration, V is the unit cell volume (158.5 Å) and N is the number of cations in the unit cell (N = 4). Then, the calculated critical distances are about 15.6 Å and 29.3 Å for Yb³⁺ (z_c = 0.02) and Ho³⁺ (z_c = 0.003), respectively. Also, the average critical distance for the total dopants (z_c = 0.023) is about 14.9 Å. The exchange interaction is known to be possible when the critical distance is less than about 5 Å. The calculated critical distance is much larger than 5 Å. Therefore, the non-radiative energy transfer between Ho³⁺ or Yb³⁺ ions mainly occurs through the multipolar interaction processes. To elucidate the type of multipolar interaction, the following relation between the emission intensity (I) and the activator concentration (z) can be used [35].

where K and b are interaction constants. The multipolar character (Q) is 6, 8, and 10 for dipole-dipole, dipolequadruple and quadruple-quadruple interactions, respectively. The Q value can be estimated from the slope of ln(I/z) vs. ln(z) assuming that βz^Q/3≫1. Figure 5(a) shows the fitting of ln(I/z) vs. ln(z) (z = x for Ho³⁺ and z = y for Yb³⁺). The resulting Q values are 4.74 and 5.73 for Ho³⁺ and Yb³⁺, respectively. These values are close to 6. In addition, as shown in the inset of Fig. 5(b), the Q value is 5.62 when calculated based on the total dopant concentration, indicating the energy transfer occurs via the dipole-dipole interaction process.

The emission was monitored while changing the IR pumping power in order to investigate the UC mechanism of CeO₂:Ho³⁺/Yb³⁺. Figure 6(a) shows the emission spectra measured with a current variation of the IR laser for the sample of CeO₂:Ho_0.003/Yb_0.02. The emission progressively increases with the increase of laser current. In UC phosphors, the emission intensity (I) is known to be proportional to Pⁿ, (I∝Pⁿ), where P is the IR pumping power and n presents the photon number for achieving one UC emission. Then, the n value is equal to the slope of ln(I) vs. ln(P) plot. As shown in Fig. 6(b), the n value estimated from the fitting resulting is 1.95, which is close to 2. Therefore, the green emission of CeO₂:Ho_0.003/Yb_0.02 is mainly achieved by the two-photon process. For the CeO₂:Ho_x/Yb_y samples prepared by changing the Ho³⁺ and Yb³⁺ content, the n values were estimated and shown in Figs. 6(c) and 6(d) as a function of the dopant content. For CeO₂:Ho_0.005, the n value is 1.31. As shown Fig. 6(c), the n value progressively increases up to 2 while increasing the content of Yb³⁺. At a fixed Yb³⁺ content (y = 0.02), the n values increase monotonically from 1.69 to 2.0 while increasing the Ho³⁺ content from 0.01% to 0.5% (Fig. 6(d)). According to the previous report [36, 37], the UC mechanism can be explained by considering the competition between linear decay and upconversion for the depletion of photons in intermediate states. The intermediate and emission energy levels for the green emission are ⁵I₆ and ⁵F₄ /⁵S₂ of Ho³⁺, which are denoted as N₂ and N₄ in Fig. 1(b), respectively. When the dominant depletion of photons in the N₂ level is achieved by the linear decay, the emission from the N₄ level is proportional to P² (n = 2). On the contrary, the UC emission is proportional to P (n = 1) when the upconversion is dominant for the photon depletion of the N₂ level. For the sample prepared without Yb³⁺, the n value is 1.31. Given this, the upconversion dominates the depletion of photons in the intermediate N₂ level when no Yb³⁺ is doped. In the case of codoping Yb³⁺, the n values progressively increase up to 2.0 with increasing the Yb³⁺ content. The n values remain closely to 2.0 even when the Yb³⁺ concentration is 2.0 mol% or larger. Figure 6(e) shows the n value as a function of the Yb³⁺/Ho³⁺ mole ratio. The n value rapidly increases until the Yb³⁺/Ho³⁺ mole ratio becomes about 5 and steadily decreases when the Yb³⁺/Ho³⁺ mole ratio is larger than about 10. This result indicates that the upconversion mechanism strongly depends on the Yb³⁺/Ho³⁺ ratio. The energy transfer (ET) from Yb³⁺ to Ho³⁺ can occur in the transition of ⁵I₈→⁵I₆ (N₀→N₂) (GSA) or ⁵I₆→⁵F₄ /⁵S₂ (N₂→N₄) (ESA). If the energy transfer is mainly involved in the ⁵I₆→⁵F₄ /⁵S₂ transition (ESA process), the photons at the intermediate level (⁵I₆, N₂) should be depleted by upconversion as the Yb³⁺ content or the Yb³⁺/Ho³⁺ ratio increases. As a result, the green emission should be proportional to P¹ as the Yb³⁺ content increases. This situation is not in agreement with the experimental data. According to the results shown in Fig. 6(c), even if the Yb³⁺ content increases to larger than 2.0 at%, the n value does not decrease to 2 or less. Therefore, to meet this experimental result, the energy transfer from Yb³⁺ to Ho³⁺ should take place dominantly through the GSA process and the photon depletion at the intermediate level should be mainly caused by the linear decay. However, if the Yb³⁺ content is too high compared with the Ho³⁺ concentration, the energy transfer from Yb³⁺ to Ho³⁺ through the ESA process becomes large and not negligible. That is, the photons at the intermediate energy level (N₂) are competitively consumed by upconversion and linear decay. This is in agreement with the experimental results in which the n value decreases below 2.0 when the Yb³⁺/Ho³⁺ ratio is greater than 10.

CeO₂:Ho³⁺/Yb³⁺ was prepared synthesized by spray pyrolysis. The resulting CeO₂:Ho³⁺/Yb³⁺ particles with spherical shape and hollow structure showed intense green emission due to the ⁵F₄ /⁵S₂→⁵I₈ transition of Ho³⁺ ion and minor peaks in red and NIR under the 980 nm IR irradiation. The Yb³⁺ co-doping clearly made it possible to largely improve the upconversion green emission due to the energy transfer from Yb³⁺ to Ho³⁺. In terms of achieving the highest emission, the optimal content of Ho³⁺ (x) and Yb³⁺ (y) ions were found to be x = 0.003 and y = 0.02 in CeO₂:Ho_x³⁺/Yb_y³⁺. The upconversion emission intensity showed a linear relationship with the crystallite size of CeO₂ and the concentration quenching between the same doping ions was proved to occur mainly via a dipole-dipole interaction process. The green upconversion of CeO₂:Ho³⁺/Yb³⁺ optimized in terms of the emission intensity was achieved by a typical two-photon process. The energy transfer from Yb³⁺ to Ho³⁺ was concluded to mainly occur in the ground-state adsorption step (⁵I₈→⁵I₆) of Ho³⁺ ions unless the Yb³⁺/Ho³⁺ ratio is greater than 10.