1. Introduction

Lanthanide (Ln³⁺)-doped luminescent nanomaterials are promising candidates with significant potential in imaging and display devices [1, 2], lighting [3–7], anti-counterfeiting [8], as well as in biological applications for nanomedicine [6, 9–12]. The unique optical properties arise from the weak interaction of the 4f electrons with the ligand field due to the shielding effect of the outer 5s and 5p orbitals resulting in sharp intra-4f transitions [13]. Ln³⁺ emission has been observed to cover the ultraviolet to the infrared regions of the electromagnetic spectrum and offers the capability of being able to generate white light via careful control of the dopant ions and their respective concentration.

White light emitting nanomaterials are garnering significant attention as they find potential uses and applications involving flat panel displays and light emitting diodes (LEDs). The latter is especially important in replacing the existing incandescent and fluorescent bulbs. In general, white light can be generated by careful mixing of red, green, and blue emissions. Most of the previously reported work on white light emitting materials has relied on UV excitation to generate the desired emission [14–18]. However, the use of short-wavelength excitation could result in photodegradation of the phosphor material resulting in a loss of luminescence efficiency and in turn, performance of the device [19, 20]. The problems associated with UV excitation can be circumvented by using Ln³⁺-doped materials, which can be used to generate white light following long-wavelength excitation through a multiphoton process known as upconversion. This important process converts two (or more) low energy near-infrared (NIR) photons into a single visible (or UV) photon [21–25]. One of the benefits of developing Ln³⁺-doped nanomaterials that can emit white light is that they can be easily excited with low power and inexpensive commercially available NIR diodes. This is possible since the 4f energy states possess long lifetimes thereby eliminating the need for ultrafast NIR lasers.

Few reports are available on Ln³⁺-doped nanocrystals producing white light through upconversion [4, 5, 8, 24]. However, to the best of our knowledge white upconversion luminescence has not been observed in vanadate nanocrystals. Ln³⁺-doped vanadate crystals are known to have excellent luminescent properties and are widely used commercially as phosphors and laser materials [21, 23]. For any intended applications, it is of paramount importance to develop highly intense and efficient luminescent materials. Several factors contribute to the luminescence intensity such as dopant ion concentration and host composition, which must be optimized. Moreover, it has also been shown that the incorporation of lithium ions (Li⁺) in Ln³⁺-doped materials plays a significant role in the enhancement of the emission intensity [26–28]. To date, no studies have reported on the enhancement of white light emission in Ln³⁺ -doped nanomaterials, especially, for white light achieved through upconversion.

In this work, we report for the first time the generation of white light via upconversion in Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals. The effect of Li⁺-doping and subsequent enhancement of the upconverted white light is also investigated and reported.

2. Experimental details

2.1 Chemicals

Gd(NO₃)₃∙6H₂O, Tm(NO₃)₃∙5H₂O, Ho(NO₃)₃∙5H₂O, Yb(NO₃)₃∙5H₂O, LiNO₃, NH₄VO₃, and poly(ethylene glycol) (PEG, M_w = 8000) of the highest available purity were purchased from Aldrich and used without further modification. Citric acid was purchased from Caledon. Milli-Q water (18 MΩ) was used in all experiments.

2.2 Preparation of GdVO₄ Nanocrystals

Briefly, 0.5 mmol of Gd(NO₃)₃∙6H₂O, the corresponding dopant lanthanide nitrates (Tm³⁺, Yb³⁺, and Ho³⁺), and stoichiometric amounts of LiNO₃ were dissolved in 0.1 M nitric acid. Citric acid, twice the molar ratio of the metal ions and NH₄VO₃ (0.073 g) were added to the nitrate solution followed by 4.0 g of PEG and heated at 90 °C for 45 minutes with stirring. The mixture was then heated at 90 °C for 24 h followed by calcination at 800 °C for 16 h at a rate of 4 °C/min.

2.3 Transmission Electron Microscopy (TEM) measurements

The TEM measurement was performed with a JEOL JEM-2000 FX microscope operating at 80 kV equipped with a CCD camera (Gaten). Roughly, 1 mg of the sample was dispersed in 5 mL ethanol and sonicated for 2 h. One drop of the resulting suspension evaporated on a carbon film supported on a 300 mesh copper grid.

2.4 Spectroscopic measurements

The upconverted emission spectra were collected using a Coherent fiber-coupled F6 series 980 nm laser diode, coupled to a 100 μm (core) fiber, with a maximum power of 800 mW at 1260 mA. The nanocrystal sample was packed in a 1 mm diameter capillary tube and fixed to a metal stand. The visible emission was collected at an angle 90° from the incident beam and then dispersed by a 1 m Jarrell-Ash Czerny-Turner double monochromator. The resolution of the monochromator was ~1 nm with silt widths of 100 μm. The emissions from the sample were detected by a thermoelectrically cooled Hamamatsu R943-02 photomultiplier tube and the signals were processed by a Standard Research Systems (SRS) model SR440 preamplifier. An SRS model SR 400 gated photon counter data acquisition system was used as an interface between the spectroscopic equipment and the computer running the SRS SR 465 data acquisition software. The lifetimes were measured by exciting the sample directly with the third harmonic generated from a pulsed Nd:YAG laser (Spectra Physics, Quanta-Ray INDI). The output from the photomultiplier tube was fed into a digital oscilloscope (Tektronix, TDS 520A).

2.5 Fourier transform infrared (FTIR) measurements

The FTIR spectrum was collected using a Thermo Electron Corporation, Nicolet 6700 FTIR instrument. Approximately 0.5 mg of the GdVO₄ nanocrystal was mixed with 5 mg of KBr and pressed into a pellet. The spectrum is an average of 108 scans collected at a resolution of 4 cm⁻¹.

2.6 X-ray diffraction (XRD) measurements

XRD patterns were measured using a Scintag XDS-2000 Diffractometer equipped with a Si(Li) Peltier-cooled solid state detector, CuKα source at a generator power of 45 kV and 40 mA, divergent beam (2 mm and 4 mm) and receiving beam slits (0.5 mm and 0.2 mm). Scan range was set from 20 to 80° 2θ with a step size of 0.02° and a count time of 2 sec. The sample was measured using a quartz “zero background” disk.

3. Results and discussion

The Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals with an average size of 25 nm (TEM image shown in Fig. 1A ) were prepared via the Pechini sol-gel method. The TEM image shows that the nanocrystals are slightly sintered due to the high temperature used in the ultimate step of the synthesis. The formation of the vanadates was substantiated by the strong absorption centered at approximately 800 cm⁻¹ along with a sharp peak at approximately 450 cm⁻¹ attributed to the Gd-O stretching in the FTIR spectrum (Fig. 1B). In addition, XRD was used to confirm the formation of the tetragonal GdVO₄ phase (Fig. 1C). The sharp peaks indicate the high crystallinity of the material.

 

Fig. 1 (A) TEM image, (B) FTIR spectrum, and (C) XRD pattern of the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals prepared by the Pechini method.

Download Full Size | PDF

Figure 2A shows the upconversion emission spectrum of the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals following excitation with 980 nm (500 mW). The upconversion emission peaks observed at 475 nm and 802 nm were ascribed to the ¹G₄ → ³H₆ and ³H₄ → ³H₆ transitions of Tm³⁺ ions, respectively. The peaks observed at approximately 550 nm and 650 nm were assigned to the (⁵F₄, ⁵S₂) → ⁵I₈ and ⁵F₅ → ⁵I₈ transitions emanating from the Ho³⁺ ions, respectively. It should be noted that while the peak at 650 nm is predominantly a result of the Ho³⁺ emission, there is, however, a minor contribution from the ³F₂ → ³H₆ transition of Tm³⁺. The transitions appearing in the visible region of the spectrum (400 – 700 nm) contribute to the observed bright white light (Fig. 2A, inset). The Commission internationale de l'éclairage (CIE) coordinates, which gives the actual color the human eye perceives, is calculated from the upconverted emission spectrum (at 55.5 W/cm²) of the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals. The values are found to be x = 0.34; y = 0.32 and fall near the center of the white region. The CIE diagram indicating the position of the calculated coordinates is shown in Fig. 2B for the optimized dopant ion concentration of 1.0, 22, 1.8% for Tm³⁺, Yb³⁺, Ho³⁺ respectively. Due to the non-linear nature of the upconversion process, the upconversion emission intensity does not change linearly with the laser power. This might lead a shift in the CIE color coordinates. This is verified by measuring the upconversion emission at various laser pump powers and the corresponding CIE color coordinates were observed to only vary slightly suggesting that these nanocrystals can be used for solid state lighting applications. The x and y values of the CIE color coordinates at different power densities are 0.337, 0.322 (47.2 W/cm²); 0.331, 0.315 (42.7 W/cm²) and 0.32, 0.312 (38.8 W/cm²).

 

Fig. 2 (A) Upconversion emission spectrum of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals. Inset shows the digital image of the white light, and (B) CIE diagram with the calculated color coordinates, which fall in the white region (black dot).

Download Full Size | PDF

To understand the mechanisms by which the excited luminescent states are populated, the upconversion emission intensities (I) of the blue (¹G₄ → ³H₆), green (⁵S₂ → ⁵I₈) and red (⁵F₅ → ⁵I₈) transitions were measured as a function of pump power (P). The slope of the logarithmic plot of I versus P was determined to be 1.63, 1.16, and 1.38 for the blue, green and red emissions, respectively. The results obtained are lower than the expected theoretical values for three-, two-, and two-photon processes involved in the blue, green and the red upconversion luminescence, respectively, and are indicative of a saturation of the upconversion process as previously observed [29]. A schematic representation of the upconversion mechanisms involved in the white light emission of the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals is shown in Fig. 3 .

 

Fig. 3 Schematic representation of the upconversion mechanisms responsible for the generation of white light from Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals.

Download Full Size | PDF

Briefly, the blue emission is obtained by three successive energy transfers from excited Yb³⁺ ions, in close proximity, to populate the ¹G₄ emitting state of Tm³⁺ ions (three-photon process). Similarly, the green and red emissions are generated by successive energy transfers by excited Yb³⁺ ions to populate the ⁵F₄, ⁵S₂ and ⁵F₅ excited states of Ho³⁺, respectively (two-photon process). It should be noted that the Yb³⁺ ion is used to sensitize the Tm³⁺ and Ho³⁺ ions in the upconversion process due to its favorable electronic energy level, which has only one excited state (²F_5/2) that is resonant with the 980 nm pump wavelength and also possesses an inherently large absorption coefficient.

In an effort to enhance the intensity of the upconverted white light emission, the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals were doped with the lithium ion (Li⁺). To optimize the effect of Li⁺-doping in order to obtain maximum upconversion emission, a series of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ samples with different Li⁺ concentrations were prepared under identical experimental conditions. The upconversion emission spectra of the Li⁺-doped GdVO₄ nanocrystals along with the similar sample without Li⁺ are shown in Fig. 4 . It is quite clear that all the emission peaks are enhanced for all the lithium doped GdVO₄ nanocrystals, though only a slight increase in the luminescence intensity was observed for the 2 wt% Li⁺-doped GdVO₄ sample. The maximum enhancement (in absolute scale) is observed for the NIR peak at 802 nm for the 5 wt% Li⁺ sample (vide supra), whereas the maximum enhancement for the blue emission at 475 nm is observed for the 8 wt% Li⁺ sample. To isolate the effect of Li⁺ on the different optical transitions, each emission peak in the Li⁺-doped Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ samples was normalized with the corresponding peak in the sample without lithium (bearing same dopant concentrations). Thus calculated enhancement values are plotted against the Li⁺ concentration and shown in Fig. 5A .

 

Fig. 4 Upconversion emission spectra of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals with and without Li⁺-doping (5%).

Download Full Size | PDF

 

Fig. 5 (A) Enhancement of the upconversion emission intensities for different Li⁺ ion concentrations in Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals. (B) XRD patterns of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals with and without Li⁺-doping (5%).

Download Full Size | PDF

It is clear that the intensity of the blue, green, and NIR emissions increase up to 8% Li⁺ doping then exhibit a decrease in the enhancement, except for the NIR peak, which displays a reduction in the emission intensity at a Li⁺-doping concentration above 5%. The observation of initial enhancement and subsequent decrease in intensity has been previously noted in Ln³⁺-doped phosphor materials [24–26]. Moreover, it is clear from the graph in Fig. 5A that the emission peaks (i.e. blue, green and NIR) are enhanced to different extents for the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ samples as the Li⁺ concentration is varied.

The exact reason for the enhancement of the upconversion following Li⁺-doping is still a subject of debate [30, 31]. It is suggested that the Li⁺ ion, with a smaller ionic radius occupies the interstitial sites and reduces the site symmetry around the Ln³⁺ dopant ions thereby positively affecting the optical transitions. Furthermore, incorporation of Li⁺ ions has been suggested to decrease the Ln³⁺ inter-ionic distances [27, 30]. XRD and fluorescence lifetime measurements were carried out to gain additional insight on the effect of Li⁺-doping on the upconversion emission intensity. The XRD patterns of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals with 0 and 5% Li⁺ are shown in Fig. 5B and are both in good agreement with the reported XRD pattern for bulk GdVO₄. No significant changes were observed in the XRD pattern prior to and following Li⁺-doping confirming that the structural features are not significantly affected by incorporation of Li⁺ ions inside the GdVO₄ nanocrystals. These results suggest that the most likely location for the Li⁺ ions in the GdVO₄ matrix might be the interstitial sites. This explanation is plausible as the size of Li⁺ (0.76Å) is smaller than that of Gd³⁺ (0.938Å). Moreover, the charge mismatch between Li⁺ and Gd³⁺ makes it less likely to replace the Gd³⁺ ions. For the same reasons discussed above, the possibility of Li⁺ occupying the vanadium site is also ruled out. Thus, it is likely that the Li⁺ ions could occupy one of the interstitial sites formed by the Gd³⁺ and four oxygen atoms.

The lifetime values of the energy states responsible for the blue, NIR (both from Tm³⁺) and green (Ho³⁺) emissions of the Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ samples were measured prior to and following Li⁺-doping and are summarized in Table 1 . The lifetimes of the red emission were not considered as it has contributions from both Tm³⁺ and Ho³⁺ ions. It is clear from Table 1 that the lifetimes of the emitting states increase upon Li⁺-doping. The increase in the emission intensity is substantiated with the increase in the lifetimes of the associated energy levels and follows closely the trend observed for the enhancement of the emission intensity (Fig. 5A). Recently, similar increase in the lifetime of the emitting states has been observed in Li⁺-doped Y₂O₃:Er where an increase in the lifetime was observed up to 5% Li⁺ and remain unchanged thereafter [31]. In our study with GdVO₄ nanocrystals, an increase in the lifetimes of the blue and green emissions was observed up to a 5% Li⁺-doping concentration. In contrast, the increase in the NIR emission reaches a maximum at 8% doping. Further increase in Li⁺ concentration results in a decrease of the lifetimes of the emitting states. As previously mentioned, it is likely that the incorporation of Li⁺ in the local environment surrounding the Ln³⁺ ions is modified slightly resulting in a reduction of the site symmetry and consequently causing the enhancement of the cross-section of the particular intra-4f transitions. At higher concentrations of Li⁺ ions (ca. >8%), it is reasonable to assume that Li⁺ induces significant distortion to the lattice, which affects the spatial distribution of the Ln³⁺ ions and induces concentration quenching thus reducing the emission intensity.

Table 1. Lifetimes of the excited states associated with the blue, green, and NIR emissions observed from Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals. The Ln³⁺ dopant concentrations are 1, 22, and 1.8% for the Tm³⁺, Yb³⁺, and Ho³⁺ dopants, respectively.

View Table

4. Conclusions

In this work we have investigated the optical properties of Tm³⁺/Yb³⁺/Ho³⁺-doped GdVO₄ nanocrystals, which exhibit strong white light upconversion emission following 980 nm excitation. The calculated CIE color coordinates fall well within the center of the white region. The intensity of the white upconversion emission was further enhanced by doping with Li⁺ ions in the GdVO₄ matrix. The enhancement of the intensity depended strongly on the concentration of the Li⁺ ions and was corroborated with an increase in the lifetimes of the associated energy levels. The generation and enhancement of the white upconversion emission in a superior host such as GdVO₄ is interesting for many applications and suggests that this material may be suitable for display and lighting technologies.