Abstract
Red, green, and blue-emitting Li-doped 
(
, 
, or 
) film phosphors were coated on sapphire substrates using a sol-gel technique in which yttrium nitrate, metal nitrate (
, 
, or 
), lithium carbonate, and citric acid were used as the precursor materials. The effect of annealing temperature on 
thin films 
and thick films 
on sapphire substrates was studied with respect to the structural and cathodoluminescence (CL) properties of the 
film phosphors. The structural and optical trends observed as a function of annealing temperature differed between the thin and thick films. The 
(
, 
, or 
) red, green, and blue thick films 
annealed at 
showed the highest volume and crystallinity of the 
phase and CL intensity of Li-doped 
red, green, and blue phosphors. These results for thick-film phosphors are discussed on the basis of the optical properties, such as the CL spectrum, color coordinates, brightness, and longevity.
Export citation and abstract BibTeX RIS
Considerable attention has recently been focused on 
(
, 
, or 
) film phosphors, because they are the well-known red, green, and blue (RGB) components of oxide RGB triads.1, 2 Film phosphors have also attracted increased interest following reports establishing that carbon nanotubes (CNTs) are useful as emitters in field emission displays (FEDs). The contamination of emitter tips by outgassing from powder phosphors is a serious drawback for FED applications. One way of addressing this issue could be to replace powder phosphors with thin-film phosphors. However, despite the advantages of thin-film phosphors, including reduced outgassing, superior adhesion, high image resolution, good heat resistance, and long-term stability, their emission levels are low because of their low internal and extraction quantum efficiencies,3–6 which suggests that they are not promising candidates for use in FEDs. However, there is evidence that the internal quantum efficiencies of thin-film phosphors are as important for the external quantum efficiency as their extraction efficiencies. Therefore, our recent research efforts have focused on improving both the internal and extraction quantum efficiencies of film phosphors.3–6 The crystallinity of a thin-film phosphor is strongly correlated with its luminescence efficiency. In general, increases in the deposition temperature, the annealing temperature, and flux addition can be used to improve the crystallinities and luminescence intensities of thin-film phosphors. Increasing the optical volume of thin-film phosphors is another important approach to obtaining maximum brightness. As previously reported, the photoluminescence (PL) intensity increases with increases in the thickness of the film up to a certain thickness, beyond which the PL intensity is constant.5 In this study, we focused on the effects of varying the annealing temperature and film thickness on the CL properties of 
(
, 
, and 
) film phosphors.
Thin-film phosphors can be fabricated with various vacuum techniques such as chemical vapor deposition,7 sputtering deposition,8 pulsed laser deposition,9 and atomic layer deposition,10 which all require expensive and complicated equipment; it is also difficult with these methods to increase the film thickness above 
. Rao recently reported a simple and economical method for fabricating high-quality 
film phosphors that can be used instead of these vacuum techniques.11 Highly homogeneous films can be obtained with sol-gel syntheses at low temperatures through mixing and reacting of precursors on the atomic scale. Among the various available sol-gel techniques, we selected a sol-gel process with inorganic salt (metal nitrate) precursors, which is an inexpensive, nontoxic, and nonhygroscopic method for fabricating 
thin-film phosphors.12–14 We have also tested a solution-based material process, the sol-gel coating process, with a view to fabricating efficient 
thick-film phosphors.15, 16 In these publications, we reported methods for optimizing the structural properties and film thicknesses of red 
thin films in order to enhance their luminescence efficiencies. We have previously studied the Li concentration dependence of the internal efficiency and extraction efficiency contributions to the enhanced PL brightness of Li-doped red 
thin films.17 Further, we also recently investigated the CL brightness of sol-gel derived, Li-doped 
(
, 
, or 
) RGB thin-film phosphors to assess their usefulness in oxide RGB triads for FEDs.18 However, although the CL properties of such Li-doped 
(
, 
, or 
) RGB thin-film phosphors have been determined, there has been little quantitative analysis of the effects of varying the annealing temperature on their CL brightness. In this paper, we focus on the effects of varying the annealing temperature and film thickness on the CL and structural properties of these sol-gel derived, Li-doped 
(
, 
, or 
) RGB film phosphors on sapphire substrates. We also investigated the CL properties of 
(
, 
, or 
) RGB thick-film phosphors prepared with the sol-gel spin-coating method under E-beam irradiation of 
in order to assess their suitability for CNT-FED applications operated at moderate voltages.
Experimental
Stoichiometric amounts of yttrium nitrate, europium nitrate (or terbium nitrate or thulium nitrate), and lithium carbonate were dissolved in 2-methoxyethanol, and then the appropriate amounts of citric acid were added. The ratio of metal cations to citrate anions was 1:2. As previously reported,17 
was also added as flux to enhance the optical properties of the 
(
, 
, or 
) films. It has been demonstrated that adding 
to conventional 
thin films optimizes their PL brightness. This improvement arises for complex reasons, such as enhanced internal factors (crystallinity, grains, and substitution of interstitial oxygen) and the increased optical volume (thickness), as well as the reduction of photon trapping in the high-index guiding layers of the 
thin films due to the increased surface roughness. Thus, all the 
film samples in this study were doped with 
. After complete mixing, a homogeneous, transparent solution was obtained. Formaldehyde 
was added as a drying control chemical additive (DCCA) to the solutions.17 In the preparation of each coating, a few drops of semigel solution were placed on a 
sapphire substrate [KMT Corporation, (0001) plane, two-side polished] and spun at 
for 
. Each substrate was then dried in air at 
and baked at 
. To fabricate thicker 
(
, 
, or 
) film phosphors, the whole process was repeated, i.e., the number of coatings was increased. Thin and thick films were obtained with 5-layer and 30-layer coatings. Finally, the 
(
, 
, or 
) film-coated sapphire substrates were annealed at various temperatures, namely, 800, 1000, 1200, and 
, for 
in air. According to our previous results for the variation of CL with dopant concentration, the optimum dopant concentration was fixed at 
for 
,17 
for 
, and 
for 
. These optimum concentrations were obtained by studying the relationships between PL intensity and activator concentration. Above these optimum values, concentration quenching occurred in our experimental systems.
The room-temperature PL emission spectra of the samples were obtained with a spectrophotometer (PSI Co., Ltd., Darsar). The CL was measured at a beam current density of 
and an irradiation energy of 
using a PSI Co., Ltd., electron-gun system. The demountable electron-beam source was installed in an ultrahigh vacuum chamber equipped with an in-house CL spectrophotometer. The CL brightness was measured instantly in the reflected direction using a Topcon BM-7 luminance colorimeter. The variation of longevity with electron dosing time of the Al-coated 
thick-film samples was measured in the transmitted direction under 
electron excitation with an average current density of 
. Thermal gravity analysis (TGA) and differential temperature analysis (DTA) data were obtained to determine the appropriate baking temperature for the dried semigel samples. The TGA/DTA analysis of each gel was carried out from room temperature to 
in air at a heating rate of 
using a TA Instruments model STA1500/DSC-SP TGA/DTA analyzer. The crystal structures of the thin-film phosphors were determined from their X-ray diffraction (XRD) patterns. The XRD patterns of the thin-film phosphors were obtained with a Philips model PW1800 X-ray diffractometer with Cu 
radiation. The diffraction patterns were obtained over the range 
with a scan rate 
of 
. The thicknesses of the films were measured with a field-emission type scanning electron microscope (FE-SEM) (Hitachi model S-4700) operated at 
, and their surface morphologies were determined with an atomic force microscope (AFM) (PSIA model XE-100) operating in contact mode.
Results and Discussion
Preparation of the 
(, , or ) films by spin coating
To fabricate homogeneous, crack-free Li-doped 
(
, 
, or 
) RGB film phosphors, a small quantity of formaldehyde was added as a DCCA to the sols before gelation. As is consistent with our results for 
red films,16, 17 the degree of crack density was dramatically reduced by fabricating the Li-doped 
(
or 
) green and blue films in the presence of 2% formaldehyde by volume. As previously reported by Rao,11 the addition of formaldehyde reduces the drying stress, which is a function of the pore size and the rate of evaporation of the solvent. TGA/DTA data were used to characterize the heat-treatment history of the dried semigel coated samples in this experiment. Figure 1 shows that the 
Li-doped 
xerogels go through three successive weight loss steps during their conversion to Li-doped 
. The exothermic transitions observed at 430, 469, and 
probably correspond to the elimination of residual citric acid, the conversion of hydroxide to oxide, and the decomposition of carbonate, respectively.13, 16 We selected 
as the minimum annealing temperature for Li-doped 
(
, 
, or 
) film formation.
Figure 1. TGA and DTA data for the Li-doped 
dried semigel samples.
To fabricate thicker Li-doped 
(
, 
, or 
) films, successive coating, drying, and baking processes were carried out to increase the number of coatings. To maintain a homogeneous and uniform coating of sol while adding layers onto a given baked Li-doped 
film, we treated the film with UV exposure in the presence of ozone gas. This treatment converts baked 
films from hydrophobic to hydrophilic,16 which assists the uniform coating of the sol onto the precoated 
layer. As for our red films, we determined the contact angles and AFM surface morphologies of the Li-doped 
green films and the Li-doped 
blue films before and after treatment with 
. These results confirmed that the number of coatings of the solution on the surface of a baked 
film can easily be increased without roughening the surface. By selecting the proper DCCA, baking temperature, and surface treatment, we were able to fabricate Li-doped 
(
, 
, or 
) RGB films thicker than 
with this repeated coating process.
Crystal structures and morphologies of the thin film phosphors
We found that the crystallinities of five-layer 
(
, 
, or 
) thin films are highly dependent on their fabrication conditions, and in particular on the annealing temperature. Figure 2 shows XRD patterns of 
(
, 
, or 
) films on sapphire substrates annealed at 800, 100, 1200, and 
for 
. All the XRD lines obtained from the RGB samples annealed at 
matched the lines specified in the JCPDS files (no. 25-1200) for the 
cubic phase.19 Increases in temperature above 
were found to result in the growth of films with mixed cubic 
and monoclinic 
(YAM) phases.20 These mixed phases are due to solid-state reactions at the interfaces between the 
thin films and the 
substrates at high temperatures. These XRD patterns also indicate the presence for an annealing temperature of 
of mixed 
, YAM, and 
garnet (YAG) phase.21 The appearance of the YAG phase occurs because this yttrium aluminum oxide phase is more stable than the YAM phase at high temperatures.22 The phase evolution processes of the five-layer Li-doped 
(
, 
, or 
) RGB thin films indicated by the XRD patterns are similar and are as follows: a pure 
cubic phase is obtained on sapphire substrates at 
, and mixed 
cubic and yttrium aluminum oxide phases are obtained on the sapphire substrates at postannealing temperatures between 1000 and 
. There is one unknown phase in the 
green thin film annealed at 
. The high temperature of this annealing process and the use of sapphire substrates would definitely be of concern for FED manufacturers, who currently use low-temperature substrates. However, our purpose in the present work is to understand the conditions required to produce bright, efficient 
(
, 
, or 
) RGB thin- and thick-film phosphors.
Figure 2. XRD patterns of the Li-doped 
films after annealing at 800, 1000, 1200, and 
: (a) 
, (b) 
, and (c) 
.
The three-dimensional AFM images indicate that the root-mean-square (rms) roughnesses of the five-layer, Li-doped 
thin films increase with increases in the annealing temperature. Figure 3 shows the dependences of the rms roughnesses of the Li-doped 
(
, 
, or 
) thin films on their annealing temperatures. The rms roughnesses of the 
(
, 
, or 
) RGB films, measured with AFM, were found to range from 
, 
, and 
, respectively, with increases in the annealing temperature. The AFM results confirm that these temperature trends in the rms roughnesses of the three different kinds of 
(
, 
, or 
) RGB films are similar. The increases in the values of the rms roughness values at higher annealing temperatures are attributed to increased grain size. The two zones of the temperature dependences of the rms roughnesses of the Li-doped 
(
, 
, or 
) RGB thin films coincide with those present in the relationships between crystal structure and annealing temperature. These results suggest that the growth speed of grains in 
thin-film phosphors increases more slowly with temperature than that of grains in mixed 
and yttrium aluminum oxide phases. However, the film surfaces are still smooth at the highest annealing temperature, 
, with rms roughnesses of 19.2, 15.5, and 
for 
, 
, and 
, respectively, even for mixed phases. We also conclude that varying the kind of dopant has little effect on the surface morphologies of the sol-gel derived, Li-doped 
thin films.
Figure 3. Plots of the rms roughnesses of the 
(
, 
, or 
) thin films as functions of annealing temperature.
CL properties of the thin-film phosphors
Figure 4 shows the CL emission spectra of five-layer, Li-doped 
(
, 
, or 
) RGB thin-film phosphors annealed at various temperatures between 800 and 
. The CL emission spectra of Li-doped 
(
, 
, or 
) RGB thin-film phosphors annealed at 
under electron-beam excitation at 
match the RGB spectra of 
(
, 
, or 
) phosphors reported in previous publications,1, 2 because their major phase is the 
cubic phase. The emission lines of Li-doped 
thin-film phosphors annealed at 
, shown in Fig. 4a, correspond to transitions from the excited 
level to the 
levels of the 
ion.23 The most intense line (at 
) corresponds to the hypersensitive transition between the 
and 
levels of the 
ion. As shown in Fig. 4b, the emission spectrum of a Li-doped 
thin-film phosphor annealed at 
consists of four bands located at about 489, 541, 589, and 
; these four bands are due to the 
(
, 5, 4, and 3) transitions, respectively.2 As shown in Fig. 4c, the major peak of a Li-doped 
thin-film phosphor annealed at 
centered at 
corresponds to the transition 
, while the transitions 
, 
, and 
of the 
ion correspond to the emission peaks at 364, 680, and 
, respectively.24 Small emission peaks are present at 
for 
, 
for 
, and 
for 
for thin-film phosphors annealed between 1200 and 
. It is thought that these lines at 608, 555, and 
correspond to the emission lines of the 
, 
, and 
phosphors, respectively. The characteristics of the emission spectra of five-layer, Li-doped 
thin-film phosphors on sapphire substrates fired above 
clearly indicate that the phase containing rare earth ions consists of a mixed 
cubic and yttrium aluminum oxide crystal structure. Figure 4 also shows the relationship between the CL emission spectra of the five-layer Li-doped 
(
, 
, or 
) RGB thin-film phosphors and the annealing temperature: the annealing of the sol-gel derived films results in some improvement in their brightness. Figure 4a shows that 
is the annealing temperature that produces the maximum brightness of the 
red phosphors. At this annealing temperature, the effects of the secondary phase on the CL emission are insignificant compared to those of the emission from the cubic 
phase. At annealing temperatures above 
, the detrimental effects on the CL red emission of the 
activators due to the growth of yttrium aluminum oxide phases increase. As is consistent with the CL results for the 
red phosphors shown in Fig. 4c, 
was found for 
to be the annealing temperature that resulted in maximum brightness and peak intensity, despite the formation of secondary yttrium aluminum oxide phases. For the 
red phosphor and the 
blue phosphor, the 
cubic phase is a more efficient phase than yttrium aluminum oxide phases for CL emission. However, Fig. 4b shows that the annealing temperature that produces the highest CL intensity in 
green phosphors is 
. Thus, the CL intensity of the 
-doped yttrium aluminum oxide phosphors is not lower than that of 
green phosphors. According to Fig. 2a, 2b and 2c, the type of dopant does not influence the phase formation or crystallinity of the phosphor films. Therefore, it is possible to obtain different optimum temperatures for different dopants in 
(
, 
, or 
) phosphors, because the CL performances of YAM:Re (
, 
, or 
) and YAG:Re (
, 
, or 
) phosphors are different to those of 
-based phosphors.
Figure 4. CL emission spectra of the 
thin-film phosphors after annealing at 800, 1000, 1200, and 
: (a) 
, (b) 
, and (c) 
.
Preparation and optical properties of the thick film phosphors
Increasing the volume of the Li-doped 
(
, 
, or 
) film phosphors is another important approach to increasing the CL brightness. As previously reported, the CL brightness increases with increases in the thickness of the film up to a certain thickness beyond which the CL brightness is constant.4 To produce thicker Li-doped 
(
, 
, or 
) films, we increased the number of coating processes up to 
. Figure 5 shows FE-SEM images of cross-sectional views of 5- and 30-layer 
blue film phosphors annealed at 
. The thicknesses of the 5- and 30-layer 
films are 
and 
, respectively. The thicknesses of the three kinds of 30-layer Li-doped 
(
, 
, or 
) RGB films are similar at the same frequency and annealing temperature. Figure 5 also shows that the grain sizes of Li-doped 
films with different thicknesses are similar to each other at the same annealing temperature. The FE-SEM images confirm that thick Li-doped 
(
, 
, or 
) RGB thin-film phosphors are obtained with the sol-gel spin-coating process.
Figure 5. Cross-sectional SEM views of the 
thin- and thick-film phosphors: (a) five-layer coating and (b) 30-layer coating.
As mentioned above, the annealing temperatures for optimum CL emission of the 5-layer coated thin films 
were found to be 
for the 
and 
film phosphors, and 
for 
film phosphors. At higher annealing temperatures, the presence of yttrium aluminum oxide phases at the interfaces between the 
substrates and the 
thin films dominates the CL properties of the five-layer coated, Li-doped 
thin films. Figure 6 shows the CL emission spectra of 30-layer 
, Li-doped 
(
, 
, or 
) RGB thick-film phosphors annealed at 1000 and 
. These spectra show that the CL emission due to the cubic phase plays a more important role in the CL emission of all three RGB Li-doped 
(
, 
, or 
) phosphors, even at higher annealing temperatures 
. To confirm this phenomenon, we measured the XRD patterns of five-layer and 30-layer Li-doped 
blue films. Figure 7 shows the XRD patterns of the thin five-layer 
films and the thick 30-layer 
films on sapphire substrates annealed at 
for 
. These results show that the YAM phase present at the interfaces between the 
substrates and the 
films is one of the main structural phases in the thin films 
but is only a minor phase in the thick films 
. Thus, the optical and crystal contribution of the pure 
cubic phase is more dominant in thick films than that of the yttrium aluminum oxide phase, even at a high annealing temperature 
. The temperature that optimizes the CL brightness of 
thick-film phosphors is mainly determined by the dependence of the crystal quality of the 
-based films on the annealing temperature. Thus, the annealing temperature for optimum CL emission of the 30-layer Li-doped 
thick films 
was found to be 
.
Figure 6. CL emission spectra of the 
thick-film phosphors after annealing at 1000 and 
: (a) 
, (b) 
, and (c) 
.
Figure 7. XRD patterns of the Li-doped 
films annealed at 
.
According to previous research1 into the conditions for optimizing the CL intensity of 
blue phosphors prepared with spray pyrolysis, ambient 
results in an increase by a factor of two relative to air. It was speculated that one possible method for enhancing the CL intensity is the conversion of unwanted 
to 
during heat-treatment under ambient 
.1 We annealed the 30-layer Li-doped 
blue phosphor under a flow of 
at 
to test the effects on the CL intensity of our blue phosphors. Figure 8 shows the CL spectra of 30-layer Li-doped 
thick films annealed at 
under different ambient conditions: annealing under ambient 
improves the CL intensity by a factor of over 1.4 compared to air annealing. Therefore, we tested 30-layer Li-doped 
and 
film phosphors annealed at 
under air, and 
phosphors annealed at 
under 
as RGB specimens to check their suitability as moderate voltage CL phosphors, because their thickness 
and annealing conditions are enough to produce bright CL emission.
Figure 8. CL emission spectra of the Li-doped 
thick films annealed at 
under different ambient conditions.
CL properties of the thick-film phosphors for FED application
Figure 9 shows the CL brightness-voltage characteristics of 30-layer Li-doped 
(
, 
, or 
) RGB thick-film phosphors, measured in the reflected direction. As previously reported,4, 17 the relationship between CL brightness and applied voltage depends on several factors such as the impurity density at the surface, the penetration depth of the e-beam, the angle of emission measurement, and the thickness of the thin-film phosphor. As is also found for conventional thin-film phosphors, the CL brightnesses of the sol-gel derived, Li-doped 
(
, 
, or 
) RGB thick-film phosphors increase with increases in the applied voltage between 2.0 and 
. At an annealing temperature of 
, the 30-layer, Li-doped 
, 
, and 
thick-film phosphors exhibit CL luminances of 337, 525, and 
, respectively, under dc excitation with 
and 
. The moderate voltage CL brightnesses of Li-doped 
(
, 
, or 
) RGB thick-film phosphors reported in this work confirm that sol-gel derived 
(
, 
, or 
) film phosphors could be useful as RGB phosphors in CNT-FEDs and lighting systems. However, the sol-gel derived, Li-doped 
thick-film phosphors still exhibit 50% less brightness than commercialized powder phosphors under the same excitation conditions (voltage and current density), because not only does less light escape from the thick-film phosphors at the interface between the phosphor film and air, but the film phosphors also have smaller emission volumes and less crystallinity. Therefore, further research into enhancing the extraction efficiency and the internal efficiency of Li-doped 
(
, 
, or 
) RGB thick-film phosphors is needed, if they are to be used in display devices.
Figure 9. Plots of CL brightness vs voltage for the sol-gel derived, Li-doped 
(
, 
, or 
) thick films annealed at 
.
The Commission International de l'Eclairage (CIE) chromaticity coordinates of the CL spectra of the Li-doped 
(
, 
, or 
) RGB thick-film phosphors are shown in Fig. 10. The chromaticity coordinates are another important property for moderate excitation voltages in full-color FED applications, because these coordinates affect the ratios of the three basic colors. The CIE chromaticity coordinates of the 
, 
, and 
thick-film phosphors are (0.66, 0.34), (0.32, 0.58), and (0.14, 0.06), respectively. The color coordinates of the sol-gel derived 
thick-film phosphors are close to those of commercialized 
CRT red powders (0.65, 0.35). The coordinates of the 
-doped green film are not as good as those obtained (0.29, 0.31) for the ZnS:Cu,Al CRT green phosphor used in cathode ray tubes. Note also that the 
-doped blue phosphors produce a similar saturated blue to ZnS:Ag,Cl CRT blue phosphors, which have color coordinates of (0.14, 0.06). The color gamut is defined as the area of the triangle between the RGB color coordinates. Figure 10 shows the NTSC color gamut and that calculated from the color coordinates of the Li-doped, 
-based RGB thick-film phosphors. The color gamut of the Li-doped, 
-based RGB thick-film phosphors is 
of the NTSC color gamut. Although the color gamut of the Li-doped, 
-based RGB thick-film phosphors is smaller than the NTSC value, their color coordinates do approach those required for FEDs.
Figure 10. CIE chromaticity coordinates of the Li-doped 
(
, 
, or 
) thick-film phosphors. Color gamuts for NTSC and a 
RGB phosphor set.
Figure 11 shows the variations with time of the normalized CL intensities of Al-coated, 30-layer, Li-doped 
(
, 
, or 
) thick-film phosphors under 
electron excitation with an average current density of 
. The total electron dose for these curves corresponds to 
. At the completion of the accelerated aging experiments, the final relative CL intensities of the Li-doped 
(
, 
, or 
) RGB thick-film phosphors were found to be 42.1, 71.9, and 41.1% of their original values, respectively. It is apparent that the 
red and 
blue thick-film phosphors have similar aging rates, while the 
green thick-film phosphors have much better aging properties than the red and blue phosphors. These results indicate that different dopants have different effects on the aging rates of Li-doped 
thick-film phosphors.
Figure 11. Normalized CL intensity at 
and 
of the 
(
, 
, or 
) thick films as a function of time.
Conclusion
Efficient Li-doped 
, 
, and 
RGB thin- 
and thick-film 
phosphors were fabricated using various annealing temperatures on sapphire substrates with 5 and 
, respectively, of a sol-gel coating process. The crystal structures of the Li-doped 
thin-film phosphors on sapphire substrates were found to be divided into two phase blocks: a pure 
cubic phase for annealing below 
, and 
aluminum oxide mixed phases for annealing above 
. The aluminum oxide phases have greater effects on the CL properties of Li-doped 
thin films 
than on Li-doped 
thick films 
. The dependences of the CL properties of sol-gel derived, Li-doped 
thin-film phosphors on annealing temperature indicate that 
is the optimum temperature range for annealing Li-doped 
thin films, and 
is the optimum temperature for annealing 
thick films. The 30-layer, Li-doped 
, 
, and 
thick-film phosphors annealed at 
were found to exhibit CL luminances of 337, 525, and 
and CIE coordinates of (0.66, 0.34), (0.32, 0.58), and (0.14, 0.06), respectively, under E-beam excitation with 
and 
. The CL aging study found that the film phosphors have superior CL aging curves, confirming their lower surface areas, i.e., the 
(
, 
, or 
) RGB thick films have improved phosphor stability under electron-beam irradiation. Although the blue phosphor is less intense than the red and green phosphors, the CL properties of these efficient Li-doped 
(
, 
, or 
) RGB thick-film phosphors suggest that Li-doped, 
-based RGB films are possible candidates for use as RGB phosphors in FED applications.
Acknowledgments
This work was supported by grant no. 2005-02522 of the Nano R&D Program and grant no. R11-2005-048-00000-0 of the ERC program of the Korean Ministry of Science and Technology. This work was also supported by grant no. 10583 from the Seoul Research and Business Development Program. Professor Huh also thanks the Brain Korea 21 project.
Kookmin University assisted in meeting the publication costs of this article.