1. INTRODUCTION

Recently, much attention has been focused on quantum cutting materials due to its application in Si solar cells. One major energy loss in Si solar cells is thermalization, which is expected to be considerably reduced if the absorbed UV/blue photon ($\lambda <500\mathrm{nm}$) is cut into two near-infrared (NIR) photons that can be absorbed by Si (${\lambda }_{\mathrm{abs}}<1100\mathrm{nm}$) [1]. Owing to the abundant energy levels and narrow emission spectra lines, rare earth (RE) ions play a great role in the quantum cutting (QC) downconversion (DC) process. The ${\mathrm{Yb}}^{3+}$ ion has been extensively used for its near-infrared emission around 980 nm (${}^2{\mathrm{F}}_{5/2}\to {}^2{\mathrm{F}}_{7/2}$), which is just above the bandgap of Si-based solar cells. Therefore, the QC phosphor based on Yb: rare-earth couples could be realized as a downconversion convertor in front of solar cell panels [2]. The main systems of NIR QC phosphors are ${\mathrm{Tb}}^{3+}\text{-}{\mathrm{Yb}}^{3+}$ [3–8], ${\mathrm{Pr}}^{3+}\text{-}{\mathrm{Yb}}^{3+}$ [5,9–13], ${\mathrm{Tm}}^{3+}\text{-}{\mathrm{Yb}}^{3+}$ [5,13–18] couples. For these couples, as the excited energy level of each donor ion is located at approximately twice the energy of that of ${\mathrm{Yb}}^{3+}$, the QC mechanism was supposed to be a second-order cooperative energy transfer (ET) process [5,8,12–17]. Therefore, when the couple is absorbed a blue light photon will release two NIR photons. That is to say, the ${\mathrm{RE}}^{3+}\text{-}{\mathrm{Yb}}^{3+}$ couple would theoretically contribute the highest NIR quantum efficiency (QE) of 200%. In the previous works, the calculated NIR QE on the basis of decline lifetime is close to 200% when the ${\mathrm{Yb}}^{3+}$ ion is in a high doping concentration. However, the highest value before the concentration quenching threshold of the Yb ion is always very low. In the other words, the Yb ion light-emitting intensity achieves maximum with low Yb ion concentration, which is most beneficial for Si-based solar cells. Therefore, there is still a strong visible light emission of spontaneous radiation transition in the donor ions when the Yb ion is in the maximum value doping concentration for ET. As a consequence, the actual energy transfer efficiency and quantum efficiency are not high. In order to better apply this in solar cells, it is necessary to further enhance the NIR light emission of ${\mathrm{Yb}}^{3+}$. If the spontaneous radiation of sensitizing ions could be suppressed, the energy is likely to further transfer to ${\mathrm{Yb}}^{3+}$. Thereby, the near-infrared emission of the ${\mathrm{Yb}}^{3+}$ can be enhanced. The existence of a photonic bandgap of photonic crystal can control spontaneous emission. Hence, to weaken spontaneous radiation of sensitizing ions and enhance the energy transfer efficiency, changing the structure of QC materials is perhaps a promising way.

Photonic crystals have attracted considerable interest since the concept was first independently proposed by Yablonovitch and John [19,20]. Such crystals are three-dimensional periodic dielectric composites in which the distribution of the refractive index varies on the visible wavelength scale. This periodicity in the refractive index may lead to the formation of a photonic bandgap. The existence of a photonic bandgap can manipulate the spontaneous emission from light-emitting materials embedded in photonic crystals [21–27].

Very recently, a few papers reported that spontaneous emission could be modified by photonic bandgap effects. The visible emission of ${\mathrm{Tm}}^{3+}$ will be suppressed if the photonic bandgap overlaps with the ${\mathrm{Tm}}^{3+}$ ions visible emission band. Since the energy will release in the another form, the infrared emission with the ${\mathrm{Yb}}^{3+}$ ions could be increased while the energy transfer between ${\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}$ is enhanced. In this work, the visible–infrared light-emitting photonic bandgap materials (${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb) with an inverse opal structure were prepared by a self-assembly technique in combination with a solgel method. The effect of the photonic bandgap on QC emission of rare earth ions was investigated in inverse opal photonic crystals. In the ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opal, we not only observed a strong modification of the visible emission spectra of ${\mathrm{Tm}}^{3+}$ but also obtained an energy transfer enhancement between ${\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}$ by suppression of the red emission of ${\mathrm{Tm}}^{3+}$.

2. EXPERIMENTAL

Commercial colloidal suspension with a monodispersive polystyrene microsphere (PS) having an average diameter of 360 or 490 nm was used to fabricate opal templates by the vertical deposition process. The single size (360 or 490 nm) PS microsphere suspension or mixed microsphere suspension consisting of 360 and 490 nm PS microspheres were added into the glass containers filled with deionized water. Quartz substrates ($4\times 1\mathrm{cm}$) were vertically dipped into the glass containers filled with PS microsphere suspension, then placed in a 50°C oven. After the water was evaporated, the unitary order opal and binary disorder templates were formed on the quartz substrates.

The ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opal was prepared by the template-assisted method. The ${\mathrm{YPO}}_4$ codoped with 1 mol. % Tm and 10 mol. % Yb precursor sol was prepared by using ${\mathrm{Y}}_2{\mathrm{O}}_3$, ${\mathrm{Yb}}_2{\mathrm{O}}_3$, ${\mathrm{Tm}}_2{\mathrm{O}}_3$, and ${\mathrm{P}}_2{\mathrm{O}}_5$ as raw materials. Initially, ${\mathrm{Y}}_2{\mathrm{O}}_3$, ${\mathrm{Yb}}_2{\mathrm{O}}_3$, and ${\mathrm{Tm}}_2{\mathrm{O}}_3$ were dissolved in hot nitric acid to form a nitrate solution, which was then evaporated until drying out. The dry nitrates and ${\mathrm{P}}_2{\mathrm{O}}_5$ were dissolved in ethanol separately, then mixed together. The prepared ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb precursor solutions were used to infiltrate into the voids of the opal templates. ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opals were obtained by calcining at 950°C for 5 h in an air furnace. The ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opals prepared by unitary opal templates constructed with single size microspheres 360 or 490 nm in diameter were denoted as IPC-I and IPC-II, respectively. The ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opal prepared by the binary disorder template constructed with mixed microsphere 360 and 490 nm in diameter was denoted as the reference sample (RS). The downconversion emission measurements and luminescent decay curves of the inverse opals were carried out on an FLS920 under a 480 nm excitation. Transmittance spectra were measured by a Hitachi U-4100 spectrophotometer. The microstructures of the opal templates and inverse opals were observed by a scanning electron microscope (SEM; FEI QUANTA 200F). X-ray powder diffraction (XRD) data of the inverse opal photonic crystal were obtained on a Rigaku 2200 diffractometer.

3. RESULTS AND DISCUSSION

First, we observed morphologies of photonic crystal materials which we prepared by SEM. Figures 1(a) and 1(b) show the SEM images of a unitary opal template constructed with single microspheres 490 nm in diameter and a binary template constructed with mixed microspheres 360 nm and 490 nm in diameter, respectively. In a unitary opal, the packed microspheres form into a high ordered face-centered cubic structure with a (111) plane parallel to the surface of the glass substrate. The binary template showed that microspheres with two kinds of size were arranged in a completely disordered (random) pattern. Figures 1(c) and 1(d) show the SEM images of IPC-II and the reference sample, respectively. The SEM images of IPC-I are not shown due to its similarity to that of IPC-II. The lighter regions and the darker circles in the SEM images of IPC-II represent the walls of the inverse opals and the air spheres previously occupied by the PS microspheres, respectively. It can be clearly seen that the air spheres possessed a highly ordered hexagonal arrangement in IPC-II, which revealed that the ordered opal template was not destroyed during the sintering process. Inside each hollow region are dark holes corresponding to the air spheres in the sublayer. On the contrary, the air spheres were randomly arranged in the reference sample. The center-to-center distance between the air spheres in IPC-I and IPC-II is about 290 nm and 400 nm, respectively, both of which are about 20% smaller than the sizes of the corresponding PS microspheres used to form the template. Thus, it can be concluded that considerable shrinkage occurs during calcinations.

 

Fig. 1. (a) and (b) SEM images of unitary opal templates constructed with polystyrene microspheres 490 nm in diameter and binary templates constructed with polystyrene microspheres 490 nm and 360 nm in diameter, respectively. (c) and (d) SEM images of IPC-II and the reference sample, respectively.

Download Full Size | PDF

The structures of the inverse opal samples and the reference sample were also characterized. Figure 2 shows the XRD pattern of IPC-I, IPC-II, and the reference sample on a quartz substrate. The broadband range from 17° to 25° in the XRD pattern was considered as the diffraction of the quartz substrate. The tetragonal phase ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb can be obtained in the three sample types when sintered at 950°C for 5 h. It is indicated that the inverse opal structure did not changed the crystals type. No impurity peaks were observed, implying that inverse opal is ${\mathrm{YPO}}_4$ in a pure tetragonal phase. The ordered ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb inverse opals display iridescent regions, which can be easily observed with the naked eye. The observed bright iridescent color corresponds to the Bragg reflection from the ordered porous structure.

 

Fig. 2. XRD patterns of IPC-I, IPC-II, and the RS.

Download Full Size | PDF

To investigate the three-dimensional ordered structural effect in photonic crystals, the photonic stop band of inverse opal arising from Bragg diffraction of the inverse opal order structure should be measured. Figure 3 shows the transmittance spectra collected from the reference sample and IPC-I and IPC-II at the normal direction of the (111) plane. The photonic bandgaps for IPC-I and IPC-II are at 480 nm and 650 nm, respectively, while there was no photonic bandgap in the reference sample. Figure 4 shows the visible emission and the NIR emission spectra of IPC-I, IPC-II, and the RS, respectively. All samples show typical ${\mathrm{Tm}}^{3+}$ emission bands, corresponding to the ${}^1{\mathrm{G}}_4\to {}^3{\mathrm{F}}_4$ transitions. In the NIR region, the emission band near 1000 nm is attributed to the ${\mathrm{Yb}}^{3+}\text{:}{}^2{\mathrm{F}}_{5/2}\to {}^2{\mathrm{F}}_{7/2}$ transition. Compared with the visible emission intensity of the ${\mathrm{Tm}}^{3+}$ emission band in the reference sample, the visible emission intensity of the ${\mathrm{Tm}}^{3+}$ was decreased in IPC-II due to the inhibition of the ${\mathrm{Tm}}^{3+}$ spontaneous emission in the inverse opal. On the contrary, the QC emission intensity of the ${\mathrm{Yb}}^{3+}$ in IPC-II were enhanced in comparison with this in the reference sample. These results were attributed to the enhancement of energy transfer from ${\mathrm{Tm}}^{3+}$ to ${\mathrm{Yb}}^{3+}$. When the bandgap is located at 480 nm, ${\mathrm{Yb}}^{3+}$ and ${\mathrm{Tm}}^{3+}$ light-emitting intensity weakened at the same time. The absorption of ${\mathrm{Tm}}^{3+}$ will diminish when the photonic bandgap overlaps the ${\mathrm{Tm}}^{3+}$ excitation light wavelength range. This phenomenon strongly suggests the existence of energy transfer from ${\mathrm{Tm}}^{3+}$ to ${\mathrm{Yb}}^{3+}$.

 

Fig. 3. Transmittance spectra of IPC-I, IPC-II, and the RS.

Download Full Size | PDF

 

Fig. 4. Visible emission spectra and quantum cutting emission spectra of IPC-I, IPC-II, and the RS under 480 nm excitation.

Download Full Size | PDF

The luminescence decay curves of ${\mathrm{Tm}}^{3+}\text{:}{}^1{\mathrm{G}}_4\to {}^3{\mathrm{F}}_4$ emission at 650 nm were also observed to explain the changing of the emission spectra, as shown in Fig. 5. We prepared singly doped ${\mathrm{Y}}_{0.99}{\mathrm{Tm}}_{0.01}{\mathrm{PO}}_4$ with no photonic bandgap to demonstrate the energy transfer efficiency. It can be clearly seen that the luminescence lifetimes at 650 nm for singly doped ${\mathrm{Y}}_{0.99}{\mathrm{Tm}}_{0.01}{\mathrm{PO}}_4$, the RS, and IPC-II are 0.275, 0.203, and 0.138 ms under the excitation of 480 nm, respectively. The luminescence lifetime at 650 nm for IPC-I is identical with the RS. The ${\mathrm{Yb}}^{3+}$ doping concentration in the RS, IPC-I, and IPC-II was 10 mol. %. The decline in the lifetime as a function of ${\mathrm{Yb}}^{3+}$ concentration can be explained by the introduction of extra decay pathways due to the ${\mathrm{Yb}}^{3+}$ doping: ET from ${\mathrm{Tm}}^{3+}\text{:}{}^1{\mathrm{G}}_4$ to ${\mathrm{Yb}}^{3+}$ enhances the ${\mathrm{Tm}}^{3+}\text{:}{}^1{\mathrm{G}}_4$ decay rate. Compared with the decay curves of IPC-II and the RS, the decline with IPC-II revealed that the energy transfer efficiency was further enhanced. In the photonic crystals, spontaneous emission will be inhibited because the optical electromagnetic modes do not exist within the photonic bandgap frequency range. While the 650 nm emission was suppressed, the photons released another way. We supposed that the energy transfer rate became much faster in IPC-II which caused the luminescence lifetime at 650 nm to decline. By dividing the effective decay time of ${\mathrm{Tm}}^{3+}$ in the RS and IPC-II by that in the ${\mathrm{Yb}}^{3+}$ free counterpart, the energy transfer efficiency is calculated to be 26.2% and 49.8% for the RS and IPC-II, respectively, by the equation ${\eta }_T=1-{\tau }_{\mathrm{RS}\mathrm{or}\mathrm{IPC}\text{-}\mathrm{II}}/{\tau }_0$ [28], which indicates a rather efficient ET in photonic bandgap samples. Here, ${\tau }_0$ is the decay lifetime of ${\mathrm{Tm}}^{3+}\text{:}{}^1{\mathrm{G}}_4$ in the singly doped samples. The quantum efficiency is calculated to be 126.2% and 149.8% for the RS and IPC-II, respectively. The relation between the ET efficiency and the total QE is defined as [3] ${\eta }_{\mathrm{QE}}={\eta }_{\mathrm{Tm}}(1-{\eta }_{\mathrm{RS}\mathrm{or}\mathrm{IPC}\text{-}\mathrm{II}})+2{\eta }_{\mathrm{RS}\mathrm{or}\mathrm{IPC}\text{-}\mathrm{II}}$, where QE for the ${\mathrm{Tm}}^{3+}$ ions, ${\eta }_{\mathrm{Tm}}$, is set to 1.

 

Fig. 5. Photoluminescence decays from ${}^1{\mathrm{G}}_4$ of ${\mathrm{Tm}}^{3+}$ in singly doped ${\mathrm{Y}}_{0.99}{\mathrm{Tm}}_{0.01}{\mathrm{PO}}_4$, the RS, and IPC-II. The excitation and emission wavelengths are 480 nm and 650 nm, respectively.

Download Full Size | PDF

In order to understand the luminescence modification mechanisms of Tm-Yb codoped ${\mathrm{YPO}}_4$ inverse opal photonic crystals, the schematic energy level diagram of ${\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}$ ions is shown in Fig. 6. The ${}^1{\mathrm{G}}_4$ excited state ${\mathrm{Tm}}^{3+}$ returns to the intermediate state ${}^1{\mathrm{G}}_4$ through two mechanisms: spontaneous fluorescence emission of ${\mathrm{Tm}}^{3+}$ and energy transfer. The energy transfer and the spontaneous emission of the ${\mathrm{Tm}}^{3+}$ compete with each other. In IPC-II, the photonic bandgap at 650 nm inhibits the 650 nm fluorescence emission of the donor ${\mathrm{Tm}}^{3+}$. Therefore, the energy transfer of the other mechanism is enhanced through the route $({\mathrm{Tm}}^{3+},{{}^1\mathrm{G}}_4\to {}^3{\mathrm{F}}_4)\to ({\mathrm{Yb}}^{3+},{{}^2\mathrm{F}}_{7/2}\to {}^2{\mathrm{F}}_{5/2})$. The ${}^1{\mathrm{G}}_4$ excited state of ${\mathrm{Tm}}^{3+}$ returns to the ground state by $({\mathrm{Tm}}^{3+},{{}^1\mathrm{G}}_4\to {}^3{\mathrm{F}}_4)\to ({\mathrm{Yb}}^{3+},{{}^2\mathrm{F}}_{7/2}\to {}^2{\mathrm{F}}_{5/2})$, making energy transfer much more readily. Therefore, the energy transfer between ${\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}$ is greatly enhanced by the suppression of the red spontaneous emission of ${\mathrm{Tm}}^{3+}$. Thus, the QC emission from ${\mathrm{Yb}}^{3+}$ is considerably improved in IPC-II.

 

Fig. 6. Simplified energy level scheme of the quantum cutting emission mechanism in photonic crystals. The photonic bandgap in the photonic crystals is at 650 nm.

Download Full Size | PDF

4. CONCLUSION

In this paper, the ${\mathrm{Tm}}^{3+}$ sensitized quantum cutting emission inverse opal of ${\mathrm{YPO}}_4\text{:}\mathrm{Tm}$, Yb was successfully fabricated. The effect of the photonic bandgap on the spontaneous emission of ${\mathrm{Tm}}^{3+}$ and the energy transfer efficiency of ${\mathrm{Tm}}^{3+}$ and ${\mathrm{Yb}}^{3+}$ was investigated. With the inhibition of ${\mathrm{Tm}}^{3+}$ emission by the photonic bandgap in the inverse photonic crystals, the ${\mathrm{Yb}}^{3+}$ NIR emission was enhanced at the same time and the fluorescence decay of ${\mathrm{Tm}}^{3+}$ was reduced. By monitoring the steady-state photoluminescence (PL) and the PL decay of the intermediate ${}^1{\mathrm{G}}_4$ level of ${\mathrm{Tm}}^{3+}$ as a function of the photonic bandgap, we demonstrated that the energy transfer efficiency and the quantum efficiency were increased to 49.8% and 149.8% by suppression of the red emission of ${\mathrm{Tm}}^{3+}$, respectively. The results demonstrate that the energy transfer efficiency and quantum efficiency is greatly improved. We believe that the present work will be valuable for not only the foundational study of quantum cutting emission modification but also for new optical devices in Si solar cells.