BLUE LIGHT EMITTING NANOMATERIALS AND SYNTHESIS THEREOF

CROSS REFERENCE TO RELATED APPLICATION This application claims the benefit of the earlier filing date of U.S. provisional application No. 61/057,982, filed June 2, 2008, which is incorporated herein by reference in its entirety.

FIELD The present technology relates to blue light emitting nanoparticles, and their use in electroluminescent devices. The technology more specifically relates to blue light emitting nanomaterials, the synthesis and use thereof.

BACKGROUND Blue is one of the primary colors used in white light and hence materials that are sources of blue light are technologically important. There is interest in using blue and ultraviolet light emitting diodes (UV LEDs) as light sources, as well as blue and white organic molecule light-emitting diode (OLEDs). Unfortunately known devices have low efficiencies and poor stability. Blue light emission has been observed from Mg -, Tm - and As-doped

GaN films and nanowires. For example, Lee and Steckl (D. S. Lee, A. J. Steckl, Appl. Phys. Lett. 2003, 83, 2094.) observed enhanced blue emission from Tm ³⁺ - doped AlGaN electroluminescent devices (EL). As GaN is a very robust material, there was hope that GaN film-based devices would provide the characteristics needed for LEDs. However, these devices cannot be conveniently fabricated, and therefore, fabrication costs are very high.

There are a few reports of blue emission from GaN-based nanomaterials. For example, thermal decomposition of an amido precursor [Ga ₂ (NMe ₂ ) _O ] in NH ₃ resulted in the formation of a polymeric intermediate, which on reaction with NH ₃ resulted in the GaN nanoparticles. Recently, van Patten et al. ( G. Q. Pan, M. E.

Kordesch, P. G. Van Patten, Chem. Mater. 2006, 18, 3915.) discovered that the same precursor can be used to produce GaN nanocrystals without using NH ₃ .

The same group reported a room-temperature synthesis of GaN from Li ₃ N and GaCl ₃ that emits at 320 nm, which shifted to 365 nm upon annealing at 310 ⁰ C. Solvothermal decomposition Of GaCl ₃ and NaN ₃ mixture and in situ thermal decomposition of cyclotrigallazane incorporated into a polymer results in the formation of GaN nanoparticles which exhibited blue emission near 426 and 475 nm, respectively. However, the origin of the blue emission is mainly attributed to the presence of impurity or defect levels, which was not desirable, as it is difficult to reproduceably produce the desired result. Moreover, many of the GaN nanoparticles synthesis involve using azides and other organometallic reagents as precursors for gallium and nitrogen. These organometallic precursors and azides are highly explosive, very toxic, and extremely sensitive to air, which requires that reactions be performed with extreme care in a glove box.

Recently, InN nanomaterials have attracted increasing attention because of their potential applications in building optoelectronic nanodevices. Indium nitride (InN) is an important semiconductor of the group- 13 (also known as group-Ill) nitrides with high electron mobility, low band gap, and low toxicity. However, it remains relatively less studied compared to other group- 13 nitrides due to its low thermal stability, low dissociation temperature, and high equilibrium vapor pressure. Generally, InN thin films are made through high-temperature processes, such as reactive magnetron sputtering, metalorganic vapor phase epitaxy (MOVPE), and molecular-beam epitaxy (MBE). The development of InN-based nanomaterials enables the possibility of separation between the high-temperature synthesis and the formation of the emission layer.

To date, there have been several efforts to prepare nano-sized InN semiconductors; however, investigations on blue electroluminescence (EL) from InN-based nanoparticles have not been reported.

SUMMARY

The present technology provides a number of approaches for the production of blue light emitting nanoparticles, nanomaterials, nanocomposites and electroluminescent devices. The products of the methods have high efficiencies and

good stability. The nanocomposites and nanomaterials emit blue light with a maximum at about 420 nm to about 500 nm.

The foregoing and other objects, features, and advantages of the invention will become more apparent from the following detailed description, which proceeds with reference to the accompanying figures.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic providing one embodiment of a method for making Mg ⁺ -doped GaN nanoparticles in an inert matrix. FIG. 2 is a normalized emission spectra of Mg ⁺ -doped GaN nanoparticles before (dashed trace) and after (solid trace) removal of the Eu ³⁺ -doped La ₂ O ₃ matrix.

FIG. 3 provides experimental (dotted trace) and calculated (solid trace) X- ray diffraction patterns of Mg ²⁺ -doped GaN nanoparticles after removal of the La ₂ O ₃ inert matrix and MgO, where the small peaks at left, from left to right, were cristobalite, most likely from the quartz tube in the furnace, and some remaining La ₂ O ₃ , respectively.

FIG. 4 provides electroluminescence (EL) spectra of GaN:Mg nanocrystals- based EL device with a configuration of ITO/PEDOT:PSS/GaN:Mg/Ca/Al.

FIG. 5 provides a comparison of photo luminescence (PL) and EL spectra obtained from an ITO/PEDOT:PSS/GaN:Mg/Ca/Al device. EL spectrum of a control device, ITO/PEDOT:PSS/Ca/Al, is shown for comparison, where the applied voltages for PEDOT:PSS//GaN:Mg and PEDOT:PSS were 14 V and 8 V, respectively.

FIG. 6 are PL spectra of Zn-doped (solid trace) and undoped (dotted trace) GaN nanoparticles.

FIG. 7 provides PL spectra of GaN nanoparticles incorporated with different amounts of (A) Mg , and (B) Zn ions.

FIG. 8 is a TEM image of silica-coated, Mg ²⁺ -doped GaN nanoparticles. FIG. 9 is a schematic representation of one disclosed embodiment of a method for making Eu -doped GaN/SiO ₂ nanocomposites.

FIG. 10 provides TEM images of the (a) Eu ³⁺ -doped Ga ₂ (VSiO ₂ and (b) Eu ²⁺ -doped GaN/SiO ₂ nanocomposites, where the arrows indicate the attachment of small GaN nanoparticles onto the silica surface, and with the inset showing the enlarged view of a single Eu -doped GaN/SiO ₂ nanocomposite for clarity. FIG. 11 are PL emission spectra collected from (a) Eu -doped GaN/SiO ₂ nanocomposites, (b) Eu ³⁺ -doped Ga ₂ θ ₃ @SiO ₂ after heating in air at 900 ⁰ C, and (c) bare SiO ₂ nanoparticles after heating at 900 ⁰ C in NH ₃ (λg _X = 285 nm), where the insert is an the excitation spectrum collected from the nanocomposites (λ _em = 450 nm). FIG. 12 is an EPR spectrum of Eu ²⁺ -doped GaN/SiO ₂ nanocomposites, recorded at 135 K in the X-band (9.44 GHz).

FIG. 13 provides (a) EL spectra collected from an ITO// Eu ²⁺ -doped GaN/SiO ₂ //Ca//Al EL device and from two control devices such as ITO//GaN/SiO ₂ //Ca//Al and ITO//Eu ³⁺ -doped Ga ₂ O ₃ @SiO ₂ //Ca//Al, with the PL spectrum of Eu -doped GaN/SiO ₂ being shown, along with (b) Current (A)-voltage (V) characteristics from an ITO// Eu ²⁺ -doped GaN/SiO ₂ //Ca//Al EL device.

FIG. 14 is (a) a TEM image, and (b) a digital photograph collected from the PMMA coated Eu ²⁺ -doped GaN/SiO ₂ nanocomposites, with an enlarged TEM image of one polymer-coated nanocomposite being shown in the inset. FIG. 15 is a TEM image of InN@SiO ₂ nanomaterial.

FIG. 16 provides emission spectrum of (A) InN@SiO ₂ nanomaterial, (B) bare silica particle after nitridation, (C) In ₂ θ ₃ @SiO ₂ heat treated in Argon, and (D) In ₂ O ₃ and SiO ₂ mixed prior to nitridation, where the inset shows the excitation spectrum collected from InN@SiO ₂ nanomaterial, and where spectra B, C, and D were multiplied by 3 for clarity.

FIG. 17 provides (a) absorption (dotted) and photo luminescence (solid) spectra of InN@SiO ₂ nanomaterials, where photoluminescence spectra were measured with a 450 W Xe arc lamp excitation in KBr pellet, and (b) a TEM image and a schematic representation (insert) of InN@SiO ₂ nanomaterials. FIG. 18 provides PL and EL spectra of InN@SiO ₂ nanoparticles, with the applied voltages for EL being (a) 14 V, (b) 10 V, and (c) 9 V, respectively, and

where (d) EL OfIn ₂ OsZSiO ₂ control nanoparticles at a driven voltage of 18 V. (Insert) CIE color coordinates of the resulting blue EL emission and a photo taken from the working device.

FIG. 19 are XRD patterns of intermediate Mg -doped products, where the relative Mg/Ga percentage concentration measured with EDX is indicated beside each pattern, and the solid sticks indicate the reference pattern Of Ga ₂ MgO ₄ and dashed sticks the one of Ga ₂ O ₃ .

FIG. 20 are XRD patterns of intermediate Zn ⁺ -doped products, where the relative Zn/Ga percentage concentration measured with EDX is indicated beside each pattern, and the solid sticks indicate the reference pattern Of Ga ₂ ZnO ₄ and dashed sticks are for Ga ₂ O ₃ .

FIG. 21 are XRD patterns of the final Mg ²⁺ -doped products, where the relative Mg/Ga percentage concentration of the initial mixture is indicated beside each pattern, the solid sticks indicate the reference pattern of GaN, and the dashed sticks the for MgO.

FIG. 22 are XRD patterns of the final Zn ²⁺ -doped products, where the relative Zn/Ga percentage concentration of the initial mixture is indicated beside each pattern, and where the solid sticks indicate the reference pattern of GaN and dashed sticks are for ZnO. FIG. 23 is a TEM image of 2.9 % doped Zn:GaN.

FIG. 24 are PL spectra of 4.3% Zn doped (solid trace) and undoped (dotted trace) GaN nanoparticles.

FIG. 25 are PL spectra of 24.7% Mg ²⁺ - (dotted) and 2.6% Zn ²⁺ -doped (solid) GaN nanoparticles, where the inset shows the Gaussian fittings of the PL from Zn ²⁺ - doped GaN nanoparticles

FIG. 26 are PL spectra of GaN nanoparticles incorporated with different nominal amounts of (A) Mg ²⁺ and (B) Zn ²⁺ ions, where the order in the legend box is inverted because it is consistent with the initial concentrations.

FIG. 27 are Raman spectra of 24.7% Mg ²⁺ -, 2.6% Zn ²⁺ -doped and undoped GaN nanoparticles.

FIG. 28 are digital images of direct and inverse opals: a) 1Ox magnification of a direct opal made of PBs; b) 4Ox magnification of Eu -doped silica inverse opal in transmission mode; c) 4Ox magnification of Eu ²⁺ -doped silica inverse opal in reflection mode; and d) 10Ox magnification of FIG. 19c. FIG. 29 are transmission spectra of the direct and inverse opals made of 400 nm PBs, with the assignment of the stop bands based on the planes responsible for them being indicated.

FIGS. 30 are SEM images of the Eu ²⁺ -doped inverse opal, where FIG. 30a shows the surface with a 90° angle of incidence, and FIGs. 30 b, c and d are at a 60° angle and show the thickness of the inverse opal, with the scale bar measuring 2, 20, 3, and 2 μm in FIGS. 30 a, b, c and d, respectively.

FIGS. 31 are a) transmission spectra, b) emission spectra of sample (solid line) and references (dotted line), with the reference sizes of the initial PBs used for the preparation of the samples being indicated on the transmission spectra. FIGS. 32 are transmission spectra, where a) is Eu lifetimes at different wavelengths, λex: 355 nm, and b) is the Ratio between the reference and the sample, with the transmission spectrum of the sample being superimposed.

FIG. 33 is an Electron Paramagnetic Resonance spectra of a blue light emitting nanoparticle comprising Eu ²⁺ . FIG. 34 are PL spectra of 24.7% Mg ²⁺ - (dotted) and 2.6% Zn ²⁺ -doped (solid)

GaN nanoparticles, where the inset shows the Gaussian fittings of the PL from Zn ²⁺ - doped GaN nanoparticles.

FIG. 35 are PL spectra of GaN nanoparticles incorporated with different nominal amounts of (A) Mg and (B) Zn ions, and where the order in the legend box is inverted because it is consistent with the initial concentrations.

DETAILED DESCRIPTION I. Definitions

The following definitions are provided solely to aid the reader. These definitions should not be construed to provide a definition that is narrower in scope than would be apparent to a person of ordinary skill in the art.

Nanomaterial - Any nanoparticle, nanocrystal, or nanocomposite. In general a nanomaterial is a material with at least one length scale below 100 nm.

Nanoparticles - Includes nanoparticles and nanocrystals. Often nanoparticle and nanocrystal are used interchangeable, but a nanocrystal has to be crystalline, a nanoparticle not necessarily crystalline.

Nanocrystal - A nanoparticle that is necessarily crystalline.

Nanocomposite - A composition that has at least one nanomaterial, nanoparticle or nanocrystal in it.

Controllable defects - Defects in a nanoparticle that can be controlled by the amount of M ⁺ doping, such as Mg ⁺ and Zn ⁺ doping, of the nanoparticles. These differ from defects that arise in thin layers. The latter tend to be uncontrollable.

@ - indicates that the nanomaterial is grown on an interface material, such as SiO ₂ or TiO ₂ .

II. Blue Light Emitting Nanoparticles

The present technology concerns blue-light emitting doped nanomaterials, and a method for making such as nanomaterials. For example, doped GaN nanoparticles have been made that exhibit blue emission around 410 nm. The nanomaterials comprise Group 13 (M ¹³ ) elements, particularly gallium and indium. The Group 13 elements are doped with charged metal ions, particularly metal ions having a 2+ charge (M ²⁺ ). The nanomaterials are exemplified herein by GaN and InN nanoparticles, and the charged metal ion dopants are exemplified by Cu , Eu , Mg ²⁺ , Mn ²⁺ , Ni ²⁺ and Zn ²⁺ ions, with Eu ²⁺ , Mg ²⁺ , and Zn ²⁺ being most commonly used for disclosed embodiments. Disclosed nanomaterials are doped with an effective amount of a dopant, which may be from greater than 0 atom percent to at least 10 atom percent, and more typically is from about 1 atom percent to about 5 atom percent.

Certain disclosed nanomaterials satisfy a formula

M ¹³ N _1-X -M ²⁺ with 1 -x being close to one, and where M and M are as stated above. If M replaces the M ¹³ element, such as Ga ³⁺ , then nitrogen vacancies may compensate for

the charge difference of the cation. This is not the only way charge neutrality can be achieved, and so it stating a more precise numerical value for x is difficult.

Disclosed embodiments also may include an interface material, such as SiO ₂ and TiO ₂ . These materials often satisfy the general formula M ¹³ N _1+x -(Si0 ₂ . _y or TiO ₂ . _y ) where 1+x is slightly greater than one and 2-y is slightly smaller than 2. This result is based on elemental analysis, which shows that there is more nitrogen than is strictly needed for InN, and not enough oxygen for SiO ₂ .

Doped materials having an interface typically have a formula M ¹³ Ni _+x :M ²⁺ @(SiO ₂ . _y or TiO ₂ . _y ) where 1+x is slightly greater than one and 2-y is slightly smaller than 2. Again, this result is based on elemental analysis, which shows that there is more nitrogen than is strictly needed for M ¹³ N, and not enough oxygen for SiO ₂ .

Disclosed embodiments concern nanoparticles that were prepared by the nitridation of suitable substrate materials, such as nitridation OfMg ²⁺ - and Zn ²⁺ - doped gallium oxide nanoparticles in an ammonia atmosphere, at an effective temperature, such as greater than 500 ⁰ C, and more typically greater than 750 ⁰ C, such as from about 950 ⁰ C to 1,200 ⁰ C, depending on the melting point of the dopant. The high temperature employed in certain preparations led to the sintering of GaN nanoparticles, thus hindering the post chemical treatment to improve their processability in organic medium. To circumvent this problem, a method has been developed in which the precursor 2+-doped Group 13 oxide nanoparticles were first diluted in an inert matrix before the nitridation reaction. This was achieved by mixing the precursor nanoparticles with Eu ³⁺ -doped La ₂ O ₃ matrix in the ratio 1 :10. Eu ⁺ was used as an optical probe to determine if any changes occurred in the matrix during the nitridation step. After the nitridation, the doped GaN nanoparticles were separated from the matrix by dissolving the matrix. For Mg ⁺ -doped nanoparticles, some examples formed MgO completely with 10% aqueous HNO ₃ . The optical properties of the Mg ⁺ -doped GaN nanoparticles were not affected by the nitric acid treatment.

Certain disclosed embodiments concern coating nanoparticles with a coating material. The nanoparticles can be coated for a variety of purposes, such as to improve their dispersibility. The coating material typically includes two components. A first component, such as a phosphorous atom, that is useful for binding to the surface of the nanoparticle. A second component, typically an aliphatic organic component, is selected to increase the dispersibility of the nanoparticles. For example, the organic component can be one or more aliphatic chains, as exemplified by alkyl chains having a chain length of up to at least 10 carbon atoms. Particular disclosed embodiments used trioctylphosphine oxide as the coating material.

A hybrid polymer-GaN:Mg structure electroluminescence (EL) device utilizing these GaN:Mg nanocrystals as light-emitting material also has been fabricated. The GaN:Mg nanocrystal-based EL device exhibited a white EL emission from GaN:Mg nanocrystals (NCs). Bright blue luminescence (~ 425 nm) from nanoparticles produced according to the present invention, as exemplified by both Mg ⁺ -and Zn ⁺ -doped GaN nanoparticles, has been accomplished. The effect of the doping concentrations of these ions on the structural and optical properties was systematically studied using photoluminescence (PL), Raman and X-ray photoelectron spectroscopy (XPS). While the Raman spectroscopic analysis confirmed the distortion of the GaN lattice due to the incorporation of these ions, XPS suggests that the most likely location of the magnesium ions was the gallium lattice. Blue light emission arises from defects, and these defects can be controlled by controlling the amount of Mg or Zn . Finally, a thin shell of silica was coated on the GaN nanoparticles to improve their dispersability.

A magnesium-doped, gallium-nitridenanocrystals-(GaN:Mg NCs)-based electroluminescence (EL) device with a hybrid organic/inorganic structure of indium tin oxide (ITO)/poly(3, 4-ethylene dioxythiophene) doped with poly (styrenesulphonic acid) (PEDOT:PSS)/GaN:Mg NCs/Ca/Al has also been fabricated. The conducting polymer, PEDOT:PSS layer, was used to enhance hole injection from the ITO electrode. Current-voltage characteristics of the GaN:Mg

nanocrystal-based EL device show a diode-like behavior. White electroluminescence was observed from the device and a voltage-dependent phenomenon of EL emission spectra was found and investigated. A good correlation between the EL and photoluminescence emission peaks suggested that electron-hole recombination indeed occur in the GaN:Mg nanocrystals layer.

Eu ²⁺ -doped GaN/SiO ₂ composite nanomaterials also have been made via a simple solid state reaction. The synthetic strategy was to grow a shell of Eu ³⁺ -doped Ga ₂ O ₃ on the surface of silica nanop articles, followed by nitridation with a nitrogen source, such as NH ₃ . This material exhibited a blue emission when excited in the ultraviolet region. The origin of the blue emission was attributed to the presence of europium ions in the +2 oxidation state, probably at the interface of GaN and silica. This was supported by several control experiments. The nitridation performed in ammonia atmosphere not only assisted the GaN formation over silica but also reduced Eu ³⁺ to Eu ²⁺ . These nanocomposites were dispersible in toluene after coating with a thin layer of polymer, which was advantageous for the fabrication of polymer-based LEDs. The presence of GaN on silica was advantageous in improving the semiconductor property of the materials and potentially makes the growth of p- and n- type doped GaN materials possible. These advantages were lacking for GaN materials coated with silica. The Eu ²⁺ -doped GaN/SiO ₂ nanocomposites were characterized by TEM, EDS, XRD, FT-IR, EPR, and photoluminescence analyses.

Blue photoluminescence (~450 nm) has also been demonstrated from InN@SiO ₂ nanomaterials. The InN@SiO ₂ nanomaterials were prepared by a precipitation reaction followed by a solid-state reaction. Various control experiments demonstrated that the interface between the InN and SiO ₂ seemed to play an important role in the origin of the blue emission from the InN@SiO ₂ nanomaterial. The InN@SiO ₂ nanomaterial was characterized using analytical methods such as TEM, XRD, Raman, XPS, and photoluminescence spectroscopy, confirming the existence of InN on SiO ₂ with a small excess of nitrogen relative to indium.

III. Photonic Crystals A. Introduction

In the past, the most common way to control the characteristics of the electromagnetic radiation was to choose the appropriate vibronic energy levels structure of the material, which was the source of radiation. Lately, photonic materials are being pursued. Photonic materials are able to modify the radiation by acting on the structure of the photonic levels. Photonic crystals are promising materials from these points of view. They are systems in which the periodic modulation of the dielectric constant over the structure of the material generates a forbidden gap of photonic states, in a similar way as a periodic lattice of atomic potentials determine a forbidden electronic gap in semiconductor crystals. The combination of Bragg scattering from the periodicity of the structure and Mie scattering resonance leads to the complete exclusion of electromagnetic modes over a continuous range of wavelengths. If the periodicity of the system is not perfect or the contrast in the dielectric constant inside the structure is low, instead of a photonic band gap only a reduction of Density of States (DOS) is observed, which is normally called a Stop Band (SB).

The simplest photonic crystals are mono-dimensional: the periodicity of the system and hence the gap occurs just in one direction or at different wavelengths for different directions. They are employed as coatings on lenses or mirrors to modulate the reflectivity, as color changing paints and inks, etc. Two dimensional photonic crystals are used to design optical waveguides, nano-cavities, optical fibres, and as low threshold lasers. Three dimensional photonic crystals are hard to achieve, but they will probably open the door to optical computing. These materials are attracting more and more interest also in the hope that, in the near future, photons will be able to replace electrons as information carriers in integrated micro circuits. Photons present several advantages with respect to electrons: they can travel through dielectric materials much faster, they can carry a larger amount of information, and the energy losses are also reduced because as bosons they are not as strongly interacting as electrons. So far, photonic crystals have mainly been used as a tool to control the propagation of light through the material: drive it along

particular directions and stop it along others. The challenge now is to understand what happens at the wavelengths on the sides of the stop bands. There is still some theoretical disputation about this, but the most credited mechanism is that the reduction of the DOS within the SB range is accompanied by an increase of DOS on the sides of the stop band. This redistribution would have determinant consequences for the design of new devices able to purify and intensify the emission at certain wavelengths.

The present technology concerns coupling blue light emitting materials, particularly nanoparticles, with photonic crystals to further tune the emission characteristics of the blue light emitters. For example, this is exemplified by using Eu ²⁺ as the responsible emitter in a hybrid material based on GaN in SiO ₂ , which has an intense and fairly broad emission with the maximum in the blue, but tailing into the green. Such material was shaped into an inverse opal (air voids in silica doped matrix), in which the size of the holes in the different samples was varied between 300 and 540 nm, in order to tune the SB in different positions with respect to the Eu ²⁺ emission. Besides the modifications of the lifetime, a decrease of spontaneous emission in the range of the photonic stop band and an increase around the low energy edge of such incomplete gap was observed, in agreement with the modification of the DOS predicted by the theory. This effect is presumed to also occur on the high energy side, but in our sample this is probably obscured by the fact that the emission decreases very steeply.

Constancy in quantum yield (QY) is observed over the whole range of the spontaneous emission. This means that for an emitter placed inside a photonic crystal the reduction of the local density of states not only determines a redistribution of emission along different directions, but also redistribution to different wavelengths. This behavior can be referred to as a "smart filter": it demotes certain wavelengths and promotes others without wasting intensity, as an ordinary filter would do. A change in color coordinates due to this redistribution has indeed been observed. Such an enhancement of spontaneous emission, especially towards higher energies, allows production of more efficient devices, exploiting radiation at shorter wavelengths.

B. Theory

In order to describe the behavior of light in a photonic crystal the solutions of Maxwell equations in a periodic dielectric medium should be considered. After simple manipulations they can be rewritten in the form:

Equation 1

In which ε{r) is the dielectric function, H(r) the magnetic field of the photon, ω the frequency and c the speed of light. It is possible to demonstrate that the operator

1 on the magnetic field is ηermitian θ = Vx- . Therefore, equation 1 can be

seen as an eigenvalue equation and the diagonalization of θ can be performed completely by considering the strength and the symmetry properties of ε{r) . In such a way a band structure is spanned with regions of allowed wave-vectors and forbidden gaps. An important difference with respect to the case of electrons in solids is that for photons there is nothing like a Bohr radius. Therefore, the energy of the system is scalable with respect to the size and the dielectric constant, without any modifications on the shape of the spectrum. This treatment can be extended to two and three dimensions to determine the photonic band structure depending on the periodicity of the system. A rough approximation of the wavelength of the SB can be determined considering a modified version of the Bragg 's law, combined with Snell's law.

Equation 2

Where λ is the wavelength, S is a shrinkage factor, which takes into accounts the eventual shrinkage that a structure undergoes during its formation (vide infra), a is the cell's parameter, m is the order of Bragg 's diffraction, m and n. ₂ are the refractive indexes of the materials constituting the structure, and φ is the volume fraction of

one of them, the other being the complementary (l - φ) . In the case of direct or inverse opals, where the periodicity of the system is distorted by defects, SBs are relatively broad and equation 2 is a good prediction of the SB position.

If an emitter is inserted into a photonic crystal, its spectroscopic properties are modified by the presence of the photonic band gap. The best way to describe this is considering Fermi's golden rule:

W = 2πh V _t P (E.)

Equation 3 Where JF is the transition rate, fi is the reduced Planck constant, V _β is the matrix element of the potential that operates between the initial and final value, and p\E _fl ) is the DOS at the energy of the transition. Equation 3 shows that the probability of a transition depends on the DOS of the system. Therefore, inside the range of the photonic SB, where the DOS is reduced, W will be lower, producing a decrease in emission intensity and a lengthening of the lifetime. Outside this range or along different directions the transition probability will not be affected, but on the edges of the SB, if the DOS really increases, it would lead to an increase of the emission intensity and a decrease in the lifetime of the fluorophore located inside the structure. Disclosed embodiments of the present invention proved the theoretical prediction that in the case of an emission overlapping with the photonic stop band, the intensity is redistributed at different wavelengths. This prediction has two major consequences: i) the total QY remains the same; and ii) the intensity increases just outside the band gap. For one disclosed embodiment, Eu ²⁺ is the responsible emitter in a hybrid material based on GaN on silica, which has a fairly broad emission with its maximum at 500 nm. The GaN and Eu ²⁺ were placed inside an inverse opal of silica (air voids in silica matrix). The size of the holes in the different samples was varied between 300 and 600 nm, in order to tune the stop band in different positions with respect to the Eu ²⁺ emission. The measured quantum yield was constant for the different samples at about 5 %, the lifetime of the Eu ²⁺ increased in the forbidden

range, and its emission intensity was squeezed towards the side of the stop band, with a concomitant decrease of the lifetime. The enhancement of the emission intensity at a certain energy range opens new possibilities for the design of more efficient devices, providing color purification and intensification at whichever wavelength is needed.

IV. Examples

The following examples are provided to illustrate certain features of working embodiments of the present invention. A person of ordinary skill in the art will appreciate that the scope of the invention is not limited to the particular features exemplified by the working examples.

Example 1

A. Materials Ga(NOs) ₃ xH ₂ 0, Mg(NOs) ₂ 6H ₂ O, La(NOs) ₃ 6H ₂ O, Eu(NOs) ₃ 5H ₂ O, glycine, and trioctylphosphine (90%) were purchased from Aldrich and used as received. Aqueous ammonium hydroxide (28-30%) was purchased from Merck. The anhydrous ammonia gas (99.999%) used for the nitridation was purchased from Praxair. Milli-Q™ water with resistance greater than 18 Mω was used in all our experiments.

B. Preparation of Mg ²⁺ -doped gallium oxide nanoparticles

Mg ⁺ -doped gallium oxide nanoparticles were prepared by the glycine-nitrate combustion method. Stoichiometric amounts of Ga(NO ₃ ) ₃ ^' xH ₂ O (1.25 mmol, assuming x = 8), Mg(NO ₃ ) ₂ 6H ₂ O (0.20 mmol), and glycine were dissolved in 25 ml of water by keeping a glycine to metal ion ratio of 1.2. The solution was slowly evaporated at 120 ⁰ C, until a transparent residue was obtained. This was then heated to 220 ⁰ C. The combustion reaction took place and a brownish-yellow coloured product was obtained. The resulting solid was then heated in a flowing air at 650 ⁰ C for 5 hours to get white Mg ²⁺ -doped gallium oxide nanoparticles.

C. Preparation of Eu ³⁺ (5%)-doped La ₂ O ₃ powder

An aqueous solution was prepared by dissolving corresponding amounts of La(NOs) ₃ ^' 6H ₂ O and Eu(NO ₃ ) ₃ 5H ₂ O. This solution was added drop wise to the flask containing 10 ml of 28% NH ₄ OH solution. The precipitate was washed well with Milli-Q™ water and dried in vacuum and then converted to Eu ³⁺ -doped La ₂ O ₃ by heating in air at 950 ⁰ C for 12 hours.

D. Preparation of Mg ²⁺ -doped GaN nanoparticles

In a typical example, nanoparticles of Mg ²⁺ -doped gallium oxide were mixed with Eu ³⁺ -doped La ₂ O ₃ powder in the ratio 1 :10 (w/w) and grounded well for proper mixing. This oxide mixture was taken in a quartz crucible and placed inside a quartz furnace. The nitridation was performed in an ammonia atmosphere. The temperature of the furnace was increased to 95O ⁰ C at a rate of 5°C/minute. This temperature was maintained for 3 h before it was cooled down to room temperature (RT) at the same rate in the NH ₃ atmosphere. The NH ₃ flow was maintained at 10 SCCM (cubic centimeter per minute at standard temperature and pressure (STP). The resultant light yellowish white product was digested in 10% HNO ₃ solution for 1 hour to remove all the La ₂ O ₃ and MgO. The residue was washed with water, and methanol followed by drying under vacuum. The yield of GaN was essentially quantitative.

E. Coating Mg ²⁺ -doped GaN nanoparticles with TOPO

The dried Mg -doped GaN nanoparticles (15 mg) were mixed with 1.50 g of trioctylphosphine oxide (TOPO) and refluxed for 24 hours at 220 ⁰ C in an argon atmosphere. The resulting solid was washed well with methanol to remove any uncoordinated TOPO. The nanoparticles were finally dispersed in absolute ethanol.

F. X-ray powder diffraction (XRD) studies

Approximately 40-50 mg of the sample was gently stirred in an alumina mortar to break up lumps. The powder was smeared on to a zero-diffraction quartz plate using ethanol. Step-scan X-ray powder-diffraction data were collected over the

2θ range 3 - 100° with CuKa (40 kV, 40 rnA) radiation on a Siemens D5000 Bragg- Brentano θ-2θ diffractometer equipped with a diffracted-beam graphite monochromator crystal, 2 mm (1°) divergence and anti-scatter slits, 0.6 mm receiving slit, and incident beam Soller slit. The scanning step size was 0.04°2θ with a counting time of 2 s/step.

G. AFM measurements.

Atomic force microscopy (AMF) images were recorded in the contact mode using a Thermo microscope AFM scanner having a silicon nitride tip (model MLCT-EXMT-A) supplied by Veeco Instruments. The nanoparticles were dispersed in absolute ethanol and sonicated for an hour before depositing on a thin glass plate (5 x 5 mm ² ) by placing a drop of the dispersion followed by slow drying in air for ca. 1 h to avoid the capillary interactions during the drying process. The measurements were done with a resolution of 500 x 500 pixels per image and an image dimension of 50 x 50 μm ² . For the histogram analysis only particles which were smaller than 100 nm were taken. The reported values were based on the heights of the AFM features.

H. Photo Luminescence (PL) measurements Room temperature PL measurements on the Mg ⁺ -doped GaN samples were performed using a 325nm Omnichrome Series 74 He-Cd Laser by Melles Griot. The laser beam was focused on the Mg ⁺ -doped GaN particles through the microscope. A SpectroPro-500™ monochromator by Acton Research Corporation was used to scan the Photoluminescence signal in the visible range (350nm - 700nm). The signal was amplified by a differential preamplifier and then acquired by the computer. The PL studies of the Eu -doped La ₂ O ₃ samples were carried out using an Edinburgh Instruments' FLS 920 instrument with a 450 W Xe arc lamp and a red-sensitive Peltier-element-cooled Hamamatsu R928P PMT. The measurement was done using a solid sample holder. A KBr pellet was prepared by mixing the sample and the KBr in the ratio 1 :10 and placed in a solid sample holder. The emission from the sample was collected from the reverse side of the pellet at an

angle 30° with respect to source and normal to the sample surface. All spectra were recorded with 1 nm resolution and corrected for the instrument response.

I. Results and Discussion. FIG. 1 illustrates the various steps involved in the synthesis of nanoparticles according to the present invention, as exemplified by synthesis of Mg -doped GaN nanoparticles. First, the Mg ²⁺ -doped gallium oxide nanoparticles prepared by combustion method were mixed with a La ₂ O ₃ matrix at a mass ratio of 1 : 10. These oxide mixtures were then exposed to NH ₃ atmosphere at 950 ⁰ C for 3 hours. This resulted in the formation of Mg ⁺ -doped GaN nanoparticles, which were subsequently separated from the matrix by removing the matrix with 10% nitric acid.

The Mg ²⁺ -doped gallium oxide nanoparticles were prepared using the combustion method. During the combustion synthesis, highly exothermic reaction between the oxidant (nitrate ions) and fuel (glycine) results in the localized heating, thereby forming the nanoparticles without much sintering. XRD pattern of the nanoparticles revealed the formation of Mg ²⁺ -doped gallium oxide and AFM analysis indicates that the average particle size was ~ 6 nm.

FIG. 2 shows the emission spectra for Mg ⁺ -doped GaN nanoparticles before and after removal of the matrix. The emission spectrum of Mg ⁺ -doped GaN before removal of the matrix exhibited a peak at 410 nm along with some sharp emission peaks at 578, 591, and 619 nm. The latter three peaks arose from the Eu ³⁺ ions that were doped inside the La ₂ O ₃ matrix (vide infra). These sharp peaks were assigned to the ⁵ Do- > ⁷ Fo,i, ₂ transitions, respectively. After nitric acid treatment, the sharp emission peaks corresponding to the Eu ⁺ emissions were absent indicating the complete removal of the matrix. This also indicates that no Eu ⁺ ions diffused into the GaN material during the nitridation step. The emission peak at 410 nm was retained with a less intense broad band around 580 nm. The emission observed around 410 nm was from Mg ²⁺ -doped GaN. To verify that the emission was originating due to Mg ⁺ doping in the GaN matrix, prepared GaN nanoparticles were prepared without magnesium doping under identical conditions. This sample

exhibited a sharp emission near 385 nm which was attributed to the band-edge emission. For comparison, the emission spectrum of GaN along with that OfMg ²⁺ - doped GaN nanoparticles were determined The emission spectrum collected from Mg -doped GaN nanoparticles was red-shifted by 25 nm. This shift was attributed to the Mg ²⁺ doping of the GaN matrix. The emission of the Mg ²⁺ -doped GaN nanoparticles was much stronger than of the GaN nanoparticles.

The formation of the wurtzite phase of GaN was indicated by the XRD pattern displayed in FIG. 3. The peaks appearing at the 2θ values 33, 35 and 37 were respectively assigned to (100), (002) and (101) peaks of the nanocrystalline GaN. The lattice parameters [a = 3.18298 (47); c = 5.17706 (93)] for undoped GaN increased to [a = 3.19217 (55); c = 5.18476 (93)] for Mg-doped GaN. This slight increase in the lattice parameter can be attributed to the incorporation of magnesium ions in GaN at a Ga site or an interstitial site. In the Mg -doped GaN samples prepared without the La ₂ O ₃ matrix a 5% MgO (periclase) phase in the XRD pattern was observed. This implies that some MgO had formed as a separate phase. This phase separation happened during the nitridation. The absence of any observation of MgO phase for the Mg ²⁺ -doped GaN nanoparticles prepared with the La ₂ O ₃ matrix indicates that the nitric acid treatment employed to remove the La ₂ O ₃ matrix etches the MgO as well. The formation of the GaN and the complete removal of the matrix (Eu -doped La ₂ O ₃ ) after nitric acid treatment were also substantiated by infrared measurements. The presence of strong stretching at 575 cm ^"1 , characteristic of Ga-N and no peaks corresponding to the lanthanum oxide matrix were observed. The presence of magnesium in the Mg -doped GaN nanoparticles after the nitric acid treatment was verified by EDS analysis. The EDS results indicate that the ratio of Ga to Mg was close to 16:1, suggesting a doping level of ~ 6.0 % Mg in GaN. This ratio was very close to the expected value calculated from the weight percent of magnesium in Mg ²⁺ -doped gallium oxide nanoparticles and taking into account the 5% MgO phase that was formed during the nitridation (vide supra).

The absence of any remarkable change in the La ₂ O ₃ matrix during the nitridation of Mg ²⁺ -doped gallium oxide was confirmed by the emission analysis. No change in the Eu ⁺ emission pattern, which was very sensitive to the

environment, was observed for the as-prepared Eu ³⁺ -doped La ₂ O ₃ and after nitridation in NH ₃ at 950 ⁰ C. This was further supported by the X-ray diffraction studies which indicate hardly any change in the La ₂ O ₃ patterns before and after nitridation at 900 ⁰ C. To study the effect of using matrix in avoiding the sintering of nanoparticles during high temperature nitridation, the Mg ²⁺ -doped GaN nanoparticles after removal of the matrix were coated with TOPO to improve their dispersability in organic medium. TOPO was used in a subsequent step as a coordinating ligand for GaN nanoparticles. The attachment of TOPO to the surface of the GaN was verified by infrared analysis: a strong C-H stretch absorption was observed, in addition to a weak signal at 1125 cm ^"1 for the P=O stretching frequency). The AFM images of the TOPO-coated Mg -doped GaN nanoparticles that were prepared with and without La ₂ O ₃ matrix clearly substantiate that the Mg -doped GaN nanoparticles prepared with matrix show much less sintering compared to the nanoparticles prepared without matrix. Though some aggregation of nanoparticles was observed for the synthesis in the matrix, it was much less compared to the GaN particles prepared without the La ₂ O ₃ matrix. This was substantiated by histogram analysis, that showed that GaN nanoparticles prepared with the matrix had an average nanoparticle size of 20 nm. For GaN particles prepared without the matrix the histogram indicates formation of larger size particles with multiple distributions. Moreover, the AFM image of the Mg ²⁺ -doped GaN nanoparticles prepared without the matrix indicates the formation of larger aggregates (~ 200 nm in size). These results highlight the importance of the La ₂ O ₃ matrix in avoiding the sintering of Mg -doped GaN nanoparticles to a greater extent during high temperature nitridation. Though this matrix method avoids sintering of nanoparticles into larger clusters, sintering of 2 to 3 nanoparticles was inevitable as the void sizes between the La ₂ O ₃ matrix particles were larger than the average size of Mg -doped gallium oxide nanoparticles. This was clear from the increase in the average nanoparticle size to 20 nm after nitridation of the 8 nm precursor Mg ²⁺ -doped gallium oxide nanoparticles. This method is easily optimized by: (i) increasing the mass ratio of La ₂ O ₃ matrix to the precursor oxide nanoparticles; (ii) matching the void sizes

formed between the La ₂ O ₃ particles to the size of the precursor particles; and (iii) developing co-precipitation methods where the precursor nanoparticles were prepared during the synthesis of the matrix materials.

Example 2

The EL devices with a configuration of indium tin oxide (ITO)/poly(3, 4- ethylene dioxythiophene) doped with poly (styrenesulphonic acid) (PEDOT:PSS)/GaN:Mg NCs/Ca/Al were fabricated using a spin-coating method. The average size of the GaN:Mg NCs was estimated using transmission electron microscopy (TEM) image. The NCs had sizes in the range of 15-35 nm. X-ray diffraction (XRD) analysis confirmed the formation of the hexagonal GaN wurtzite phase. Briefly, poly(3,4-ethylene dioxythiophene) doped with poly(styrenesulphonic acid) (PEDOT:PSS) was spin-coated on a clean surface of a patterned ITO covered glass substrate, followed by 15 minutes soft baking at 110 ⁰ C in vacuum (<10 torr). Subsequently, a dispersion of GaN:Mg NCs (4 mg/ml) in iso- propanol: chloroform (1 :1, v/v) was ultra-sonicated for 2 hours, spin-coated on top of the PEDOT: PS S layer, and dried in vacuum (<10 torr) oven at 60 ⁰ C overnight. The ώo-propanol:chloroform mixture did not dissolve the PEDOT:PSS blend. Finally, the metal cathode (20 nm Ca and 150 nm Al) was thermally evaporated onto the NCs layer at <5 x 10 ⁵ torr using a shadow mask to complete the device. The devices having an active area ~7.5 mm ² were connected to a direct current (dc) voltage supply (Keithley 2400 source meter) with the positive terminal attached to the ITO electrode. EL spectrum was measured using an Edinburgh Instruments' FLS 920 fluorescence spectrometer. All the device fabrication and characterization steps were carried out under ambient conditions.

The work functions of ITO, PEDOT:PSS and Ca/Al electrodes were values reported in the literature. The electron affinity of GaN:Mg nanocrystals was taken from previous experimental data of p-type GaN. The energy gap of 3.0 eV was based on the emission at 410 nm. In the EL device, holes were injected from the ITO electrode through PEDOT:PSS layer into the NCs. Similarly, the electrons

were considered to be injected from the Ca/ Al electrode into the NCs layer. The electron-hole pairs combine on nanocrystal layer releasing energy as emission light.

The PEDOT:PSS was chosen as a hole injection/transport layer because of its many advantages, such as a high transparency, excellent thermal stability and high conductivity. The current-voltage (I-V) characteristics of the GaN:Mg EL device exhibited a diode-like behavior. The EL emission was only found under a forward bias, i.e. when a positive voltage was applied to the ITO electrode. As a reference, the threshold voltage of the PEDOT:PSS-only device was measured and found to be around 2.5 V, whereas the threshold voltage of GaN:Mg EL device was increased to 5 V. Increased threshold voltage suggests that there was a higher barrier for charge injection from electrodes into the GaN:Mg layer. For further investigation, /- V curve of another device without PEDOT:PSS, i.e. a device with a configuration of ITO/GaN:Mg/Ca/Al was fabricated, which exhibits only a conductor behavior, indicating that the PEDOT:PSS was necessary to lower the energy barrier for hole injection from ITO to GaN :Mg layer.

EL spectra as a function of applied currents from the device with a configuration of ITO/PEDOT:PSS/GaN:Mg/Ca/Al were displayed in FIG. 4. An interesting aspect was that the spectra showed a voltage-dependent behavior. The emission intensity increased with increased applied voltage with some change in emission wavelength. At lower voltages, the emission appears preferentially at longer wavelengths with a broad maximum wavelength around 570 nm, whereas at higher voltages the emission of 410 nm appears. Only a very weak emission around 530-570 nm was found from a control device consisting of ITO/PEDOT:PSS/Ca/Al (FIG. 5). This indicates that the broad emission probably occurred in the PEDOT:PSS layer at lower voltages. When the voltage was high enough, EL emission from GaN:Mg NCs was clearly observed. EL emission peaks at 70 niA almost cover all the visible regions of the spectrum. The calculated CIE (Commission International de l'Eclairage) coordinates from EL spectrum of 70 mA in FIG. 4 were 0.35 and 0.39, well within the white region of 1931 CIE diagram. This white light can be seen by the naked eye. No noticeable EL emission was found in the absence of the PEDOT :PSS layer.

As a comparison, the PL emission from GaN:Mg nanocrystals excited with a 325 nm He-Cd laser at room temperature is also provided in FIG. 5. The PL spectrum shows a peak at 410 nm with a broad band around 565 nm. The EL spectrum matches well with the PL spectrum of GaN:Mg NCs. The emission observed around 410 nm was attributed to the donor-acceptor pair recombination from GaN:Mg NCs ¹⁸ and the broad emission centered at 565 nm was assigned to the well-known yellow emission from GaN. The origin of the latter emission was, however, still under debate. Without being bound to theory, the broad emission in our devices most likely arose due to the presence of defects at the surfaces, which could in principle be minimized by subsequent surface chemistry or tuning the preparation conditions such that the synthesis leads to reduced defects on the nanocrystal surfaces. If the broad emission near 565 nm is effectively depressed, it is also possible to make a blue light-emitting device with GaN:Mg NCs. The good correlation between the EL and PL emission peaks suggest that electron-hole recombination indeed occurred in the GaN:Mg nanocrystals layer. However, the intensity of long wavelength emission peak was enhanced due to the EL from PEDOT:PSS in the long wavelength of- 550 nm. The EL spectrum in FIG. 5 can be fit by the PL emission of the GaN:Mg NCs and EL spectrum of PEDOT:PSS. This reveals that the EL was both from the nanocrystals and the polymer layers, due to the recombination in both components, but the EL emission of GaN:Mg nanocrystals was the dominate one at higher voltages.

Example 3 A. Materials Ga(NO ₃ ) ₃ xH ₂ O, Mg(NO ₃ ) ₂ 6H ₂ O, Zn(NO ₃ ) ₂ 6H ₂ O, tetraethylorthosilicate

(TEOS), and glycine were purchased from Aldrich and used as received. The anhydrous ammonia gas (99.999%) used for the nitridation was purchased from Praxair. Milli-Q™ water with resistance greater than 18 Mω was used in all our experiments.

B. Preparation of oxide precursors

The Mg ²⁺ - and Zn ²⁺ -doped gallium oxide precursors were prepared by the combustion method, which has been reported elsewhere. Briefly, stoichiometric amounts of Ga(NO ₃ ) ₃ xH ₂ O (1.25 mmol, assuming x-8), Mg(NO ₃ ) ₂ 6H ₂ O (0.20 mmol) or Zn(NO ₃ ) ₂ 6H ₂ O (0.20 mmol), and glycine were dissolved in 25 ml of water by keeping a glycine-to-metal ion ratio of 1.2. The solution was slowly evaporated at 120 ⁰ C, until a transparent residue was obtained. This was then heated to 220 ⁰ C. The combustion reaction took place and a brownish-yellow coloured product was obtained. The resulting solid was then heated in a flowing air at 650 ⁰ C for 5 hours to get white Mg ²⁺ -doped gallium oxide or Ga ₂ ZnO ₄ nanoparticles.

C. Preparation of Mg ²⁺ - and Zn ²⁺ -doped GaN nanoparticles

Nanoparticles of Mg ²⁺ -doped gallium oxide or Zn ²⁺ -doped gallium oxide in a quartz crucible were placed inside a quartz furnace. Nitridation was performed in an ammonia atmosphere. The temperature of the furnace was increased to 950 ⁰ C at a rate of 5°C/minute. This temperature was maintained for 3 hours before it was cooled down to RT at the same rate in the NH ₃ atmosphere. The NH ₃ flow was maintained at 10 SCCM (cubic centimeter per minute at STP).

D. Silica coating on the GaN nanoparticles

To coat a thin layer of silica over the GaN nanoparticles (M ⁺ doped, such as Mg - or Zn -doped), roughly 25 mg of the corresponding nanoparticles were dispersed in 75 ml ethanol and sonicated for 2 hours. After sonication, the supernatant was removed after the dispersion was undisturbed for an hour. To the above mixture, 23.5 ml water, 1.5 ml NH ₄ OH, and 50 μl of TEOS were added and stirred overnight.

E. X-ray powder diffraction (XRD) measurements

The XRD patterns of the Mg ²⁺ - and Zn ²⁺ -doped GaN nanoparticles were collected using a Rigaku Miniflux™ X-ray diffractometer with a Cr K _α (30 kV, 15 mA) radiation source. The nanoparticle samples were gently crushed before being

smeared on to a clean quartz slide. The powder diffraction patterns were collected over the 2θ range 20 to 140 ° with a scan speed and sampling width of 1 °/minute and 0.02 °, respectively.

F. Raman spectroscopic measurements

Raman spectra were collected by exciting the sample with 632.8 nm He-Ne

Laser by Melles Griot. The solid sample was evenly spread over a clean glass slide.

Each spectrum was collected over 30 seconds and were expressed as an average of 6 scans.

G. X-ray photoelectron spectroscopy (XPS) measurements

X-ray photoelectron spectra were collected using a Leybold Max200™ spectrometer, using monochromatic Al K _α X-ray source (1486.6 eV). The pass energy for the survey and narrow scans were 192 and 48 eV, respectively. Photoelectrons were collected at 90 ° from the surface. All binding energies (BE) were reported relative to the CIs peak (BE: 285.0 eV).

H. Transmission Electron Microscopic (TEM) measurements

The TEM images were collected using a Hitachi H-7000 tungsten filament up to 125 kV Transmission Electron Microscope. TEM specimens were prepared by dipping a copper grid (600 mesh), which was coated with an amorphous carbon film into the ethanol dispersion of the silica-coated GaN:Mg nanoparticles, followed by slow drying at room temperature.

I. Photo Luminescence (PL) measurements

Room temperature PL measurements on the Mg ²⁺ - and Zn ²⁺ -doped GaN samples were performed using a 325 nm Omnichrome Series 74 He-Cd Laser by Melles Griot. The laser beam was focused on the Mg ⁺ -doped GaN particles through the microscope. A SpectroPro-500 monochromator by Acton Research Corporation was used to scan the Photo luminescence signal in the visible range

(350nm - 700 nm). The signal was amplified by a differential preamplifier and then acquired by the computer.

J. Results and Discussion The Mg - and Zn -doped GaN nanoparticles were prepared by a solid state reaction, involving nitridation of the corresponding oxides, such as Mg -doped gallium oxide or Zn ²⁺ -doped gallium oxide nanoparticles, which were produced by the glycine combustion method to yield small nanoparticles typically in the size range of from about 5 to about 10 nm. The XRD pattern of the Zn ²⁺ -doped GaN nanoparticles clearly confirms the crystalline GaN with wurtzite structure. Both Mg - and Zn -doped GaN nanoparticles exhibit a bright blue emission at 425 nm when excited at 325 nm. This PL peak position is red-shifted by roughly 50 nm with respect to the undoped GaN nanoparticles. For example, FIG. 6 displays the emission spectra of Zn ²⁺ (3%)-doped GaN nanoparticles and the undoped GaN nanoparticles. The shift in the emission was attributed to the doping of Zn ⁺ ions into the GaN matrix. Upon excitation in the UV region, most of the electrons from the conduction band non-radiatively relaxed to the defect level from where the PL originates. It was also possible that few electrons from the conduction band edge relaxed directly to the ground state. This led to the band edge emission near 375 nm appearing as a shoulder. In addition to these peaks, there was also a broad band observed around 600 nm, which was more pronounced in the case of undoped GaN nanoparticles compared to Zn -doped nanoparticles. This was normally denoted as the "yellow emission" and assigned to the certain type of defects in the GaN lattice (vide infra). To compare the optical characteristics of Mg ⁺ - and Zn ⁺ -doped GaN nanoparticles, samples comprising a doping level of 3% were prepared under identical conditions. The PL spectra of the Mg ²⁺ (3%)-doped GaN nanoparticles compared with the corresponding Zn -doped nanoparticles showed that the main PL peak for both Mg - and Zn -doped GaN nanoparticles appeared at 425 nm; however, the emission spectrum of the Zn ²⁺ -doped GaN nanoparticles showed additional shoulder towards longer wavelength. Moreover, the Gaussian fitting of

the emission peaks of Zn ²⁺ -doped nanoparticles shows two peaks with peak maximum centering at 427 and 458 nm. Moreover, the Gaussian fitting of the emission peaks of Zn ²⁺ -doped nanoparticles showed two peaks with peak maximum centering at 427 and 458 nm. Without being bound to a theory of operation, the multiple peaks could be attributed to the presence of additional defect states created by the incorporation of zinc ions into the GaN matrix. The sharp emission at 427 nm may arise from the Znoa, whereas the origin of the emission near 458 nm was attributed to the zinc ions at the nitrogen vacancy. This difference in the luminescence was possibly due to the difference reactivity (solubility) of the Mg ²⁺ and Zn ²⁺ ions during the growth of GaN nanoparticles.

The difference in the optical properties between the Mg - and Zn -doped GaN nanoparticles was not only limited to the difference in the shape of the PL peak but also on the defect-related yellow emission as well as on the observed PL trend for different doping levels (see below). As was observed for Zn and undoped GaN, the observed yellow emission near 580 nm was predominant in the Mg ⁺ - doped GaN compared to Zn ²⁺ -doped nanoparticles. Without being bound to a theory of operation, Mg ⁺ doping appears to create more defects on the GaN lattice compared to the Zn ²⁺ -doped GaN nanoparticles most likely closer to the surface of the nanoparticles. The latter was true as the size of the nanoparticles was roughly in the range of 12-15 nm. The presence of more lattice defects in Mg ²⁺ -doped GaN nanoparticles was substantiated with the Raman analysis (see below).

FIG. 7 shows the PL tread observed in Mg - and Zn -doped GaN s with different doping concentrations. Mg ²⁺ -doped samples showed a slight reduction in the intensity with the decrease in the Mg ⁺ concentration, whereas the opposite effect was observed for the Zn -doped GaN nanoparticles (FIG. 7b). The preferred location for the Mg ⁺ in GaN was magnesium in the gallium site (Mg _Ga )- These results suggest that increasing the doping concentration increases the probability of incorporation of Mg into GaN lattice. Thus the increase in the concentration of the Zn ²⁺ might increase the probability for the creation of additional defect states. Though this explanation is tentative, the observed PL behavior is attributed to the different origin of the emission.

To get more insight on the structural characteristics, Raman and XPS analyses were performed. All spectra displayed a strong peak at 570 cm ^"1 , which was the characteristic E ₂ phonon frequency of the host GaN. Additionally, the spectra showed few other weak bands at 530 cm ^"1 , 650 and 727 cm ^"1 . The peak at 530 cm ^"1 was assigned to the A ₁ optical modes and the latter two bands were attributed to the disorder-activated Raman scattering. These disorder-activated bands were originally Raman inactive modes and their appearance was ascribed to the distortion in the lattice structure. As it was clear from the Raman spectra the peak at 727 cm ^"1 was more intense in the Mg-doped GaN compared to the Zn-doped samples and it was quite weak in the undoped GaN nanoparticles. This clearly implies that there was more lattice rearrangement in the case of Mg ²⁺ -doped GaN compared to Zn ²⁺ - and undoped GaN samples.

XPS analysis was performed to obtain details on the elemental composition. The Ga 3d and NIs region of the XPS spectra showed that the Ga 3d peak appeared at a binding energy (BE) of 20.1 eV and the nitrogen Is at 397.2 eV. This was consistent with the value obtained for GaN nanorods. There was hardly any change in the BE positions of Ga 3d or N Is after incorporation of Mg ⁺ , which might be due to the small amount OfMg ²⁺ incorporation. To get a clearer picture on the effect of increase in magnesium doping concentration, the atomic ratios of Mg-to-N, Ga- to-N, and Mg-to-Ga were plotted against the doping amounts of the magnesium. The graph indicated that both the Mg/Ga and Mg/N ratios increased with the increasing Mg concentration, whereas the Ga/N ratio dropped for 3% doping and slightly increased before decreasing again. Without being bound to theory, the slight decrease in the ratio of the Ga/N for all Mg ⁺ -doped GaN samples compared to the undoped GaN nanoparticles along with the increase in the Mg/N and Mg/Ga ratio suggests that the most likely location of the Mg ²⁺ ions in the GaN lattice was the gallium site. Though a steady increase has been observed in the Mg/N and Mg/Ga ratios with increase in the doping level a similar decrease was not observed with the Ga/N ratio. Without being bound to theory, this is probably because with the increase in the Mg ²⁺ concentration, the Mg ²⁺ ions prefer to occupy the interstitial sites, as the size of the Mg ²⁺ is slightly larger than the Ga ³⁺ ion.

Having studied the structural and optical properties of the Mg ²⁺ - and Zn ²⁺ - doped GaN nanoparticles, applications of the technology were examined. One of the applications was the exploitation of these GaN nanoparticles in the development of EL-based devices. The other possibility was the integration of these nitride-based semiconductor nanoparticles into biological systems. Any bio-related applications demand the need of dispersible nanoparticles. As a proof of concept, the dispersa bility of the nanoparticles was increased by coating a thin layer of silica over the GaN nanoparticles via the well-known Stόber method. In principle, silica coating not only improves the dispersability but provide a biocompatible surface. The TEM image shown in FIG. 8 indicates a thin bright region over a dark spheres. This clearly suggests the formation of a very thin silica shell over the GaN nanoparticles. The coating of the silica shell was substantiated by the dispersability of the core/shell materials in ethanol over a period of few weeks. These core/shell materials can be utilized for attachment of bio-linkers or to grow another semiconductor shell.

Example 4 A. Chemicals

Tetraethyl orthosilicate, aqueous ammonium hydroxide (28-30%) gallium nitrate (99.98%), europium nitrate (99.99%), methylmethacrylate, urea, potassium bromide, and ethanol (99.9%) were used as received from Aldrich. The anhydrous ammonia gas (99.999%) used for the nitridation was purchased from Praxair. Milli- Q ^TM _wa ter with resistance greater than 18 Mω was used in all examples.

B. Preparation of bare silica nanoparticles

The silica nanoparticles with an average size of 50 nm were synthesized using a literature procedure. Briefly, 3.8 ml of tetraethyl orthosilicate was added to a mixture containing 114 ml of ethanol and 5.7 ml of ammonium hydroxide (28- 30%) while vigorous stirring. Stirring was continued overnight, which results in the formation of silica nanoparticles with average size of 50 nm. These silica particles were used as prepared in all the following examples.

C. Preparation of Eu ³⁺ (5%)-doped Ga ₂ O ₃ ZSiO ₂ nanocomposites

Eu (5%)-doped Ga ₂ O ₃ ZSiO ₂ nanocomposites were prepared by adapting a procedure reported for Gd ₂ O ₃ :Eu ³⁺ @Si0 ₂ . Briefly, 0.6 g of urea dissolved in water was added to a flask containing 15 ml of 50 nm silica nanoparticles. To this mixture an aqueous solution containing Ga(NO ₃ ) ₃ and Eu(NO ₃ ) ₃ was added dropwise and vigorously stirred at 85 ⁰ C for 3 hours. The resulting white precipitate was purified by centrifugation and redispersion in water for 3 times followed by drying in vacuum. Finally, the white product was heated to 800 ⁰ C for 2 hours.

D. Preparation of Eu ²⁺ (5%)-doped GaNZSiO ₂ nanocomposites

Approximately 100 mg of the Eu ³⁺ -doped Ga ₂ O ₃ ZSiO ₂ nanocomposites was taken in a quartz crucible and placed it in an electric furnace (Lindberg). The furnace was heated to 900 ⁰ C for 3 hours in NH ₃ atmosphere at a rate of 5 ⁰ C per minute. The sample was left at the final temperature for 3 hours before it was cooled to room temperature in NH ₃ atmosphere at a rate of 5 ⁰ C per minute. The ammonia flow was maintained at 10 SCCM (cubic centimeter per minute at STP). The sample color changed from colorless to yellow during this reaction process.

E. Preparation of PMMA coated Eu (5%) doped GaNZSiO ₂ nanocomposites Roughly 0.5 ml of methylmethacrylate was added to 15 mg of the Eu ²⁺ - doped GaNZSiO ₂ nanocomposites and sonicated for 2 hours. After sonication, approximately 1 mg of azobisisobutylnitrile (AIBN) as initiator was added to the above mixture. The mixture was thermally polymerized at 72 ⁰ C for 1 hour. The resulting mixture was dispersed in toluene.

F. Transmission Electron Microscopic (TEM) measurements

A Hitachi H-7000 tungsten filament up to 125 kV Transmission Electron Microscope was used to collect the TEM images. TEM specimens were prepared by dipping a copper grid (600 mesh), which was coated with an amorphous carbon film into the ethanol dispersion of the nanomaterial composites, followed by drying at room temperature.

G. Energy Dispersive X-ray Spectroscopy (EDS)

The EDS analysis was done using an Oxford Instruments Link ISIS EDX X- ray microanalysis system on the Hitachi S-3500N Scanning Electron Microscope. The samples were gently crushed before depositing on a double sided tape.

H. Fourier Transform Infrared (FTIR) measurements

FTIR measurements were done using a Perkin Elmer FTIR spectrometer 1000 machine. A KBr pellet was made by mixing dried KBr and the sample approximately in the ratio 10:1. All spectra were an average of four scans and recorded with a resolution of 2 cm ^"1 .

I. Powder X-ray diffraction analysis

The solid sample was first gently crushed to break the larger clumps, and then after smeared on to a zero-diffraction quartz plate using ethanol. Step-scan X- ray powder diffraction data were collected over the 2θ range 15 to 80 ° with CuK _α (40 kV, 40 mA) radiation on a Siemens D5000 Bragg-Brentano θ-2θ diffractometer equipped with a diffracted-beam graphite monochromator crystal, 2 mm (1°) divergence and anti-scatter slits, 0.6 mm receiving slit, and incident beam Soller slit. The scanning step size was 0.04° 2θ with a count time of 2s per step.

J. Electron paramagnetic resonance (EPR) measurements

The EPR spectrum was recorded on a Bruker EMX EPR instrument operating in the X-band (9.44 GHz). The temperature was 135 K. The collected spectrum was an average of 100 scans.

K. Photoluminescence (PL) measurements

PL measurements were carried out using an Edinburgh Instruments' FLS 920 instrument with a 450 W Xe arc lamp and a red-sensitive Peltier-element-cooled Hamamatsu R928P PMT. The measurement was done using a solid sample holder.

A KBr pellet was prepared by mixing the sample and the KBr in the ratio 1 :10 and

placed in a solid sample holder. The emission from the sample was collected from the reverse side of the pellet at an angle of 30° with respect to source and normal to the sample surface. All spectra were recorded with 1 nm resolution and corrected for the instrument response. The filters used were 320 nm and 435 nm respectively, for the collection of emission and the excitation spectra. The Eu ³⁺ emissions were recorded by exciting the sample at 464 nm with 5 ns pulses at a frequency of 10 Hz, using a Vibrant tunable laser system (Model 355 lib) with a Quantel Nd-YAG nanosecond pump laser. The lifetimes of the Eu ²⁺ emission (at 450 nm) were collected using a nanosecond flash lamp. The lifetime decays were fitted by two- exponentials based on the following equation: ψ- = B _{0 +} ±A, ∞p(-t/τ,)

Intensities down to 1% of the initial intensities were included in the lifetime analysis and the chi-square (χ ² ) was 1.36.

L. Quantum yield measurements

The absolute quantum yield measurements were determined using an integrating sphere (Edinburgh instruments, 150 mm in diameter coated with barium sulfate). The Eu -doped GaN/SiO ₂ nanocomposites were coated with a thin layer of silica to enhance their dispersability in ethanol. Similarly, a GaN/SiO ₂ composite without europium ions was also coated with silica and used as a reference sample.

The emission spectra collected from the nanocomposites before and after the silica coating were identical with respect to shape and position. Absolute error of 2% was based on duplicate measurements.

M. Electroluminescence (EL) device fabrication using Eu ²⁺ -doped GaN/SiO ₂ nanocomposites.

Devices with a configuration of ITO//(Eu -doped) GaN/SiO ₂ //Ca//Al were fabricated as follows. First, the nanocomposites were dispersed in ώo-propanol by sonicating the mixture for 6 hours. After sonication, the nanocomposites were spin- coated onto a clean indium tin oxide (ITO) glass substrate at 1500 rpm for 30

seconds under air-free conditions. The resulting film was then dried overnight in a vacuum oven at 60 ⁰ C. Finally, the metal cathode (20 nm thick Ca and 150 nm Al) was thermally evaporated onto the nanocomposite layer at < 5 ^χ 10 ^~5 torr using a shadow mask to complete the device. The EL was characterised using the Edinburgh Instruments' FLS 920 fluorescence spectrometer (vide supra). The DC power was supplied with a Keithley 2400 source meter.

N. Results and Discussion.

FIG. 9 is a schematic representation of the various steps involved in disclosed working embodiments of the preparation of Eu ²⁺ -doped GaN/SiO ₂ nanocomposites. Silica nanoparticles of 50 nm were prepared using a literature procedure (H. Hiramatsu, F. E. Osterloh, Langmuir 2003, 19, 7003.) These silica nanoparticles were formed by the hydrolysis and condensation of tetraethyl orthosilicate in the presence OfNH ₄ OH as catalyst. These silica colloids were used as templates to grow a thin shell of Eu ³⁺ -doped Ga ₂ O ₃ over them. This was achieved by first coating a hydroxide layer followed by heating at 800 ⁰ C. These oxide-coated silica nanoparticles were then converted into nitrides by nitridation in an ammonia atmosphere at 900 ⁰ C for 3 hours, which resulted in the formation of small nanoparticles of Eu ²⁺ -doped GaN on the silica nanoparticles. The formation of the GaN was indicated by a color change of the sample from colorless to yellow after nitridation. Moreover, the Eu -doped GaN/SiO ₂ nanocomposites were coated with a thin layer of PMMA to tune their dispersability in organic medium.

The formation of a thin shell of Ga ₂ O ₃ on the silica surface was substantiated by the TEM image of the sample. This is shown in FIG. 1OA. The image shows dark spheres coated with a lighter shell. The large dark regions correspond to the silica nanoparticles and the thin shell corresponds to Eu -doped Ga ₂ O ₃ . The image also represents some dense regions, which could be due to formation of isolated gallium oxide nanoparticles which seem still attached to the silica surface. Nitridation of the oxide precursor led to the formation of small GaN nanocrystals on the surface of silica nanoparticles as seen in TEM measurements (FIG. 10B). The image shows small dark regions of GaN nanocrystals (~ 2-3 nm) on the surface of

large silica nanop articles. The image also suggests the absence of a homogeneous coating of GaN nanoparticles on the silica surface.

The formation of the hexagonal GaN on the surface of the silica nanoparticles was confirmed by XRD analysis. The XRD pattern obtained from Eu ²⁺ -doped GaN/SiO ₂ nanocomposites showed peaks appearing at the 2θ values 37.8, 40.4 and 43.1. These were assigned, respectively, to (100), (002) and (101) peaks of the crystalline GaN. The lattice parameters for the hexagonal GaN [a = 3.187(1) and c= 5.183(3)] were in agreement with those reported in the literature for the bulk GaN. A broad hump observed around 25° (2θ) was from the amorphous silica. The exact nature of the silica material (nitride or oxynitride) was not clear due to the broadening of the peak but there likely was oxynitride formation (vide infra). In addition, the XRD analysis showed no evidence of any Ga ₂ O ₃ present in the nanocomposites after nitridation indicating the complete conversion of the Ga ₂ O ₃ present on the silica surface into GaN, although a tiny amount of crystalline or amorphous Ga ₂ O ₃ cannot be ruled out. However, the controls argue against Ga ₂ O ₃ as the cause of the blue emission.

EDS analysis of the Eu ²⁺ -doped GaN/SiO ₂ showed the presence of Si, Ga, Eu, O and N. The peaks appearing at 1.098, 1.740 and 5.846 KeV correspond to the characteristic X-ray energies of GaL _αl , SiK _αl and EuL _αl , respectively. The atomic ratio of Ga to N was 1 :2. The nitrogen concentration was twice that expected for the formation of GaN, indicating the formation of some silicon oxy nitride compound. This was also consistent with the slight decrease in the Si:O ratio from 1 :2.0 to 1 :1.6.

The infrared spectrum collected from the Eu ²⁺ -doped GaN/SiO ₂ nanocomposite confirmed the presence of Si-O stretching and bending vibrations near 1100 cm ^"1 and 450 cm ^"1 . The presence of the peak near 650 cm ^"1 was assigned to the Ga-N stretching. In order to verify this assignment, the sample was exposed to 10% HF to etch the silica present in the nanocomposites. After exposure to HF solution for 5 hours the sample was washed well with water before drying in vacuum. The infrared measurement after etching clearly indicated the presence of a strong band at 650 cm ^"1 which was a characteristic stretching frequency of Ga-N.

O. Optical Properties of Eu ²⁺ -doped GaN/SiO ₂

The Eu ²⁺ -doped GaN/SiO ₂ nanocomposite when excited in the UV region exhibited a bright blue emission centered at 450 nm, as shown in FIG. 11a. The calculated CIE colour co ordinates were X = 0.183 and Y = 0.226, which falls well within the blue region. The excitation spectrum collected from the Eu -doped GaN/SiO ₂ nanocomposites displayed broad absorptions from 250 to 300 nm (shown in the inset of FIG. 11). The measured lifetimes were 1.38 μs (80%) and 0.38 μs (20%) for the emission at 450 nm were consistent with the lifetimes reported for Eu ²⁺ ions. Fast photocycles are advantageous for increased brightness for LEDs, displays, etc. The absolute quantum yield calculated using an integrating sphere for the 450 nm emission from the nanocomposites was 23 ± 2 %. The broad shape and the short lifetime of the emission peak infer that the peak does not correspond to an intra-4/ transition. This was because the intra-4/ ^" electronic transitions exhibited sharp optical emissions with lifetimes in the millisecond range due to the shielding of the 4/" energy levels from the surroundings by the filled 5 s ² and 5p ⁶ energy levels and the forbidden nature of the intra 4/ transition. In addition, only a weak Eu ³⁺ emission was observed when excited with a laser (464 nm, absorption line of Eu ) whereas the corresponding Eu ³⁺ -doped Ga ₂ Ch @SiO ₂ precursor exhibits strong Eu ³⁺ characteristic emissions. The narrow peaks appearing at 578, 591 and 612 nm were characteristic of the Eu ³⁺ emission originating from the ⁵ Do (excited state) to ⁷ Fj (ground state) 4/electronic transitions. All the above observations indicate that the Eu ions present in the Ga ₂ θ3@SiO ₂ have been reduced to Eu during the nitridation step. To confirm this several control experiments were performed. First, the precursor Eu ³⁺ -doped as well as undoped Ga ₂ θ ₃ @SiO ₂ were subjected to heating at 900 ⁰ C in air to verify the presence of any defect-related emission from Ga ₂ O ₃ or Ga ₂ O ₃ @SiO ₂ as observed by others. Only a very weak emission was observed from these materials in the blue region and the Eu -doped Ga ₂ O ₃ @SiO ₂ displayed only the characteristic Eu ⁺ emissions when excited in the UV region. Second, only a very weak emission was observed from the silica nanoparticles after nitridation at 900 ⁰ C in NH ₃ (FIG. 1 Ic). Finally, a GaN/SiO ₂

nanocomposite was prepared by doping 5% yttrium ions instead of europium ions under similar experimental conditions. The Y ³⁺ mimics the Eu ³⁺ doping but was optically silent and could not be reduced to Y . The sample did not display any emission in the blue region but exhibited a broad spectrum centered at 560 nm, which could be attributed to the yellow emission from GaN. This signifies the importance of Eu in the generation of the blue emission and rules out the possibility of band-edge or any defect-related emission from the GaN. From the control experiments discussed above it was clear that the emission observed at 450 nm was from Eu ²⁺ ions and that most of the Eu ³⁺ ions that were present in Ga ₂ θ ₃ @Siθ ₂ had been converted into Eu ²⁺ ions during the nitridation process.

In order to get definite proof for the presence of Eu in the europium-doped GaN/SiO ₂ nanocomposites, EPR analysis was performed on the powder sample. The EPR spectrum displayed in FIG. 12 confirms the presence of Eu in the europium-doped GaN/SiO ₂ nanocomposites. The observed spectrum is similar to that obtained for the Eu ²⁺ -doped heavy-metal fluoride glasses (ZBLAN).

The emission from the Eu ²⁺ ions is assigned to the electronic transition occurring between the 41^-5(I ¹ excited state and ⁸ S _7/2 ground state of the Eu ²⁺ . This transition is an allowed transition due to mixing of 5d states with 4f states in contrast to the parity forbidden transitions within the 4f energy levels of the lanthanides. The excited state configuration 41^-5(I ¹ is very sensitive to the host lattice (i.e. crystal field) and can occur in any part of the visible region of the electromagnetic spectrum. Earlier, Kim and Holloway (J. H. Kim, P. H. Holloway, J. Appl. Phys. 2004, 95, 4787.) observed the presence of some Eu ²⁺ in Eu ³⁺ -doped GaN thin films prepared by radio-frequency sputtering, the origin of which was attributed to the result of preferential sputter removal of nitrogen. T. Maruyama, S. Morishima, H. Bang, K. Akimoto, Y. Nanishi, (J. Cryst. Growth 2002, 237, 1167.) observed the valence transition of Eu to Eu ions in GaN material. They found that the valence transition occurs to those Eu ⁺ ions that were present near the surface of GaN. To understand the location of the Eu ²⁺ ions and to verify that the emission originating from the nanocomposite formed by GaN and silica in our Eu ²⁺ -doped GaN/SiO ₂ nanocomposites we performed a control experiment where two samples, i.e.,

Eu ₂ O ₃ @SiO ₂ and Ga ₂ O ₃ nanoparticles were prepared separately. These two oxide samples were mixed well and subjected to nitridation at 900 ⁰ C in NH ₃ atmosphere. The emission analysis of the mixed oxides after nitridation shows a weak broad emission near 500 nm. The shape and position of this emission peak is quite different from that obtained from the Eu -doped GaN/SiO ₂ nanocomposites. In addition, the mixed oxides show a sharp emission at 612 nm due to the presence of

Eu ions which was not observed for the Eu -doped GaN/SiO ₂ nanocomposites. Finally, Eu ³⁺ -doped GaN nanoparticles mixed with silica particles show only the characteristic Eu ³⁺ emissions but no blue emission. These results strongly support the fact that the blue emission arises from the Eu ²⁺ that are present in the nanocomposites formed by GaN and SiO ₂ . Due to the large size of the Eu ⁺ ions compared to Ga ³⁺ ions it is reasonable to presume that they are present at the interface of GaN and SiO ₂ .

In order to take advantage of the presence of Eu -doped GaN nanoparticles on the surface of the silica nanoparticles and as a proof of principle, especially in the fabrication of thin film electroluminescence (device)(TFEL) devices, an LED device was fabricated using the Eu -doped GaN/SiO ₂ nanocomposites as light-emitting material. The EL spectrum of the nanocomposites measured at 18 V was shown in FIG. 13. A broad peak centered at 485 nm dominates the EL spectrum, which was strong enough to be observed with the naked eye. The EL peak is slightly red-shift compared to the PL peak. This behavior has been observed before for nanocrystal- based EL. The reason was attributed to the local heating from the large current flux and the poor thermal conductivity of the inorganic light emitting layer. Control experiments were also performed under the identical conditions using the Eu - doped Ga ₂ O ₃ @SiO ₂ as well as an undoped and Y ³⁺ -doped GaN/SiO ₂ nanocomposites. The absence of any emission from these devices (EL measured at 18 V for GaN/SiO ₂ and Eu ³⁺ -doped Ga ₂ O ₃ @SiO ₂ are shown in FIG. 13) clearly indicate that the blue emission is originating from the Eu ²⁺ -doped GaN/SiO ₂ nanocomposites in the LED device. The current- voltage (I-V) of a device consisting of IT 0//Eu -doped GaN/SiO ₂ //Ca//Al exhibits an exponential diode characteristics with a threshold voltage around 5 V (shown in FIG. 13b). The observation of bright

blue EL proves the advantage of the semiconducting material such as GaN outside the silica surface. However, to achieve better stability, efficiency, and to make use of these nanocomposites for fabrication of polymer-based devices, the device structure needs to be optimized. In order to show a feasible route, the Eu -doped GaN/SiO ₂ nanocomposites were coated with a thin layer of polymethyl methacrylate. The TEM image shown in FIG. 14 confirms the formation of a thin layer of PMMA coating over the Eu ²⁺ - doped GaN/SiO ₂ nanocomposites. The image also shows that the GaN nanoparticles attached to the silica surface were not affected by the PMMA coating. This was supported by the observation of a blue emission from the PMMA coated GaN/SiO ₂ nanocomposites when excited at 285 nm. The emission spectrum collected from the Eu ²⁺ -doped GaN/SiO ₂ nanocomposites matches exactly the emission spectrum of nanocomposites before PMMA coating indicating that the optical properties of the Eu ²⁺ -doped GaN/SiO ₂ have not been affected by the thin polymer coating over them. A digital photograph of the blue emission from the Eu ²⁺ -doped GaN/SiO ₂ dispersion in toluene is also shown in FIG. 14.

P. Conclusions

The synthesis of Eu ²⁺ -doped GaN/SiO ₂ nanocomposites has been demonstrated via a simple solid state reaction. The nanocomposites show a bright blue emission when excited in the ultraviolet region. The origin of the blue emission was attributed to the existence of europium ions in the +2 oxidation state. Thus, Eu ²⁺ could be a potential alternative for other blue light-emitting Ln ³⁺ ions such as Tm ⁺ or Dy ⁺ . Various control experiments showed that the blue emission originates from the Eu ²⁺ ions present in the nanocomposite matrix made with GaN and SiO ₂ , and it was most likely that the Eu ²⁺ ions were located at the interface between GaN and silica. A bright blue electroluminescence was also observed from Eu ²⁺ -doped GaN/SiO ₂ nanocomposites. In addition, the Eu ²⁺ -doped GaN/SiO ₂ nanocomposites were coated with a thin layer of polymer to make them dispersible in organic medium. These polymer-coated Eu -doped GaN/SiO ₂ nanocomposites

were envisioned to have practical significance in the fabrication of polymer-based electronic devices.

Example 5 A. Chemicals

Tetraethyl orthosilicate, aqueous ammonium hydroxide (28-30%), indium nitrate (99.98%), urea, potassium bromide, and ethanol (99.9%) were used as received from Aldrich. The anhydrous ammonia gas (99.999%) used for the nitridation was purchased from Praxair. Milli-Q ^™ water with resistance greater than 18 mω was used.

B. Preparation of monodispersed silica nanoparticles

The silica nanoparticles with an average particle size of 50 nm were synthesized using a literature procedure. Briefly, 3.8 ml of tetraethyl orthosilicate was added to a mixture containing 114 ml of ethanol and 5.7 ml of ammonium hydroxide (28-30%) while vigorous stirring. Stirring was continued overnight, which resulted in the formation of silica nanoparticles with an average particle size of 50 nm. These silica particles were used as prepared, i.e. without isolation.

C. Preparation of In ₂ θ ₃ @SiO ₂ nanomaterial

In ₂ O ₃ (S)SiO ₂ nanomaterial was prepared by adapting a procedure reported for Gd ₂ O ₃ :Eu@SiO ₂ . Briefly, 0.6 gram of urea dissolved in water was added to a flask containing 15 ml of 50 nm silica nanoparticles. To this mixture, a 0.2 M aqueous solution containing In(NO ₃ ) ₃ .xH ₂ O (x = 5, as provided by the supplier) was added drop wise. The resulting mixture was vigorously stirred at 85 ⁰ C for 3 hours. The excess urea was removed by centrifugation and the precipitate was dispersed in water 3 times. The resulting white product was dried in vacuum before heated to 600 ⁰ C for 9 hours.

D. Preparation of InN@SiO ₂ nanomaterial

Approximately 100 mg of the In ₂ O ₃ (S)SiO ₂ nanomaterials was taken in a quartz crucible and placed in an electric furnace (Lindberg HTF553222A with a CC58114PA controller). The furnace was heated to 700 ⁰ C at a rate of 5 ⁰ C per minute in an NH ₃ atmosphere. The sample was left at the final temperature for 2 hours before it was cooled down to room temperature in NH ₃ atmosphere. The ammonia flow was maintained at 100 SCCM (cubic centimeter per minute at STP). The formation of the InN was indicated by a color change of the sample from colorless to grayish black.

E. Photoluminescence measurements

Photoluminescence (PL) measurements were carried out using an Edinburgh Instruments' FLS 920 instrument with a 450 W Xe arc lamp and a red-sensitive Peltier element-cooled Hamamatsu R928 PMT. The measurement was done using a solid sample holder. A KBr pellet was prepared by mixing the sample and the KBr in a weight ratio 1 :10 and placed in a solid sample holder. All spectra were recorded with 1 nm resolution and were corrected for the instrument response. The filters used were 320 nm and 435 nm for the collection of emission and the excitation spectra, respectively.

F. X-ray powder diffraction (XRD) measurements

The XRD pattern of the InN@SiO ₂ nanomaterial was collected using a Rigaku Miniflex X-ray diffractometer with a Cr K _α (30 kV, 15 mA) radiation source. The nanoparticle samples were gently crushed to break down big lumps and a thick paste was made using ethanol. This thick paste was evenly spread onto a clean quartz slide and the ethanol evaporated at 85 ⁰ C. The powder diffraction patterns were collected over the 2θ range 20 to 140 ° with a scan speed and sampling width of l°/min and 0.02 °, respectively.

G. Raman spectroscopy

Raman spectra were collected by exciting the sample with 632.8 nm from a He-Ne Laser by Melles Griot. Roughly, 20 mg of the solid sample was placed on a clean glass slide and spread over the slide evenly using a spatula. The spectrum is an average of 3 scans with a 30 seconds collection time for each scan.

H. Transmission Electron Microscopic measurements

A Hitachi H-7000 tungsten filament up to 125 kV Transmission Electron Microscope was used to collect the TEM images. TEM specimens were prepared by dipping a copper grid (600 mesh), which is coated with an amorphous carbon film into the ethanol dispersion of the InN@SiO ₂ nanomaterial and dried at room temperature.

I. X-ray Photo-electron Spectroscopy X-ray photoelectron spectra were collected using a Leybold Max200 spectrometer, using monochromatic Al K _α X-ray source (1486.6 eV). The pass energy for the survey and narrow scans were 192 and 48 eV, respectively. Photoelectrons were collected at 90° from the surface. All binding energies (BE) were recorded relative to the CIs peak (BE: 285.0 eV).

J. Results and Discussion

The InN@SiO ₂ nanomaterials were prepared via a simple precipitation followed by a solid-state reaction. Briefly, an aqueous solution containing indium nitrate and urea was added drop wise to 50 nm spherical silica nanoparticles followed by heating at 85 ⁰ C for 3 hourse. This produces an In-urea complex over the silica nanoparticles. This resulting white precipitate was washed 3 times with water to remove any excess urea and followed by drying under vacuum. Subsequently, the white powder was heated to 600 ⁰ C in air which resulted in a yellowish- white powder due to the formation of an In ₂ O ₃ on silica. The yellowish- white powder was nitridated under nitrogen which results in the formation of a black material. The black color indicates the formation of the InN. To understand the

structure of the black material TEM analysis was performed. The TEM image shown in FIG. 15 shows the formation of small nanostructures of InN that are attached to the surface of the silica particles. In order to confirm the formation of the InN we carried out an XRD analysis. The XRD pattern collected from InN@SiO ₂ nanomaterial shows that the observed pattern corresponds well with that reported for crystalline InN. The broad diffraction peak at 20° (2θ) corresponds to the amorphous silica phase. Additionally, the formation of the InN was confirmed by Raman analysis. The room temperature Raman spectrum for InN@SiO ₂ nanomaterial was conducted. The peak at wave number of about 590 cm ^"1 is assigned to A ₁ [longitudinal optical (LO)] phonon peak of InN. The characteristic Raman Si-O-Si bending vibration appearing at 430 cm ^"1 is not clearly seen in the spectrum as it is overshadowed by the strong Raman band from InN.

K. Optical Properties of InN@SiO ₂ The photo luminescence (PL) spectrum collected from InN@SiO ₂ nanomaterial displays a blue emission centered at 450 nm when excited in the UV region. This blue emission was easily seen by the naked eye. FIG. 16A shows the emission spectrum along with the excitation spectrum in the inset. The material has a strong absorption in the UV with a shoulder close to 275 nm and an onset at -300 nm. The emission peak is strongly blue-shifted compared to band gap emission (600-700 nm) reported for the InN nanomaterials.

As quantum confinement effects are not expected in this case, due to irregular shapes of the nanomaterials, various control experiments were performed to understand the origin of the blue emission. First, to verify whether the emission was coming from the indium nitride, InN material alone was prepared using a similar procedure. No emission was observed from this sample. Second, the silica particles alone were heated in NH ₃ under identical conditions. Only a very weak emission was observed from the SiO ₂ after nitridation (FIG. 16B). This weak emission could have come from carbon related impurities present in silica, which has been observed earlier. Thirdly, to confirm whether the emission was coming from In ₂ O ₃ present in a very small amount (as In ₂ O ₃ shows broad emission in the

blue region due to oxygen deficiencies), and to understand the need for the nitridation, the precursor nanoparticles, i.e. In ₂ Os(S)SiO ₂ material, was heated at 700 ⁰ C in an argon atmosphere. The absence of any characteristic emission from this sample (FIG. 16C) indicates the importance of the presence of InN@SiO ₂ in producing the blue emission. In order to confirm this, another control sample was made by mixing the In ₂ O ₃ and silica particles together followed by nitridation under the same conditions used for the preparation of InN@SiO ₂ . The resulting mixture displays only a very weak emission (FIG. 16D), which highlights the importance of the InN growing on the surface of a SiO ₂ nanostructure for the blue emission. The above control experiments clearly verify the importance of both InN and silica in producing the blue emission from InN@SiO ₂ nanomaterial. Moreover, the absence of any characteristic emission from the hybrid mixture (i.e. In ₂ O ₃ and silica separately made and mixed together) after nitridation substantiates the importance of growing the In ₂ O ₃ on the silica template (In ₂ O ₃ (S]SiO ₂ ). This observation indicates that the interface between the InN and silica plays an important role in bringing the blue emission. Without being bound to theory, one possibility could be formation of a different material, such as In ₂ Si ₂ O ₇ , at the interface. Though In ₂ Si ₂ O ₇ has characteristic luminescence it mostly falls in the ultraviolet region. Moreover, if it were In ₂ Si ₂ O ₇ , the precursor In ₂ O ₃ (S]SiO ₂ would have exhibited the emission after heating but it only shows a weak emission (see above). The second possibility is the formation of In-O-N complex as observed by Liu et ah, in their In-O-N nanospheres. The third possibility, which is closely related to the second, is the widening of the optical bandgap of InN with incorporation of oxygen. This is reported for the polycrystalline InN where the bandgap was increased from 1.55 eV to 2.27 eV with increasing oxygen concentration from 1% to 6%. These options are tentatively ruled out as no band gap emission was observed in InN alone, made under identical conditions. To get more insight, XPS analysis was performed on the InN(S]SiO ₂ as well on the precursor nanomaterials. The results indicated that for the precursor In ₂ O ₃ (S]SiO ₂ , the expected ratios of 1 : 1.5 and 1 :2 were observed for In:O and Si:O, respectively. However, InN(S]SiO ₂ indicates an In/N ratio of 0.66 and Si/0 ratio of

2. These results indicate that the material is richer in N than is strictly needed for InN.

Thus, the synthesis of a hybrid InN@SiO ₂ has been demonstrated, which exhibit a blue emission when excited in the ultraviolet region. The structural characterization of the nanomaterials indicates formation of crystalline InN nanomaterial over the amorphous silica particles. Various control experiments support that the interface between the InN and silica plays an important role in the origin of the blue emission.

Example 6

FIG. 17a shows the absorption and photo luminescence spectra of InN@SiO ₂ nanoparticles measured at room temperature. These nanoparticles exhibit a blue emission centered at 445 nm when excited in the UV region, but no luminescence was observed corresponding to the most quoted band gap value, 1.89 eV (656 nm), or the recently reported band gap ~ 0.70 eV (1771 nm). The blue luminescence was most likely not due to the quantum confinements in the InN nanostructure because of their irregular shapes observed on the surface of SiO ₂ nanoparticles (FIG. 17b). A control sample such as InN-SiO ₂ hybrid material prepared by directly mixing the same amount OfIn ₂ O ₃ and pure SiO ₂ nanoparticles and treated under identical nitridation conditions did not show the blue emission in the visible region. An EL device with a configuration of ITO//InN@SiO ₂ //Ca//Al was fabricated by simply spin-coating the nanoparticles onto an indium tin oxide (ITO) covered glass substrate. After drying, the metallic cathode (Ca protected by Al) was thermally deposited onto the emission layer with a shadow mask. The current- voltage (I- V) characteristic of the InN@SiO ₂ EL device showed a clear diode-like behavior, with blue emission observed only in the forward bias, i.e., when a positive voltage was applied to the ITO electrode (FIG. 18). The logl /logV plot showed an ohmic behavior at low voltages (< 4 V), following the I oc V ^1'4 relation, which indicates that current was limited by the InN@SiO2 layer. At higher voltages (> 4 V), the I-V data was fit well to the trapped-charge-limited (TCL) current model, i.e., I oc V ^m relation with m ~ 4. This suggests that the filling of the limited traps result

in a rapid increase (power-law increase) in current at higher forward bias. QD-based LEDs and organic light-emitting devices have been reported to exhibit an I x V m relation with variations in m values. Here, m was the signature of the charge condition mechanism and was related to the type of the emitting materials, the temperature, and the trap energy.

The EL device started to emit blue light, which was visible to the naked eye, at 9 V, when a dc voltage was applied to the device. As shown in FIG. 18, the intensity of EL emission increased with increased applied voltage with no change in emission wavelength. The EL occurs ~ 460 nm, i.e., exhibiting ~ 15 nm red-shift versus the PL peak at 445 nm. The red-shift of the EL, relative to the PL, was well known in nanocrystal-based EL. In the present technology, the reason for the red- shift of the EL could be explained by local Joule heating from the large current injection and relatively poor thermal conductivity of the emitting layer. Because there was no additional hole or electron transfer layers into which the recombination zone can extend, all the EL emission originated from InN@SiO ₂ layer. This results in excellent color purity. Control devices fabricated using In ₂ O ₃ (S)SiO ₂ or pure SiO ₂ as emission layers did not show any blue EL (FIG. 18), substantiating that the blue EL occurs from the InN@SiO ₂ nanomaterials. The calculated CIE (Commission International de l'Eclairage) coordinates for the EL spectrum in FIG. 16 were 0.18 and 0.23, which fall well within the blue region of 1931 CIE diagram. The calculated brightness of the blue emission was ~ 0.3 cd/m at 18 V with a luminous efficiency of 0.2 mcd/A at a current density of 150 mA/cm ² . The reason for the low brightness could be attributed to the poor charge injection efficiency on the inorganic nanomaterials surface. Though the observed brightness seems to be low, it was observed from an un-optimized device. Optimization of the devices could be achieved by the following ways (i) formation of a close packed monolayer for nanomaterials, (ii) combination with electron/hole transporting layers, and (iii) optimization of the device structures.

In conclusion, blue electroluminescence from InN@SiO ₂ nanomaterials has been observed when the EL device was driven at a forward bias voltage. The

control experiments and the comparison analysis of PL and EL showed that electron-hole recombination does occur in the InN@SiO ₂ nanomaterials layer.

Example 7 A. Materials

Ga(NO ₃ ) ₃ xH ₂ O, Mg(NO ₃ ) ₂ 6H ₂ O, Zn(NO ₃ ) ₂ 6H ₂ O, tetraethylorthosilicate (TEOS), and glycine were purchased from Aldrich and used as received. The anhydrous ammonia gas (99.999%) used for the nitridation was purchased from Praxair. Milli-Q water with resistance greater than 18 Mω was used in all examples.

B. Preparation of the oxide precursors

The Mg ²⁺ - and Zn ²⁺ -doped gallium oxide precursors were prepared by the combustion method which has been reported elsewhere. Briefly, stoichiometric amounts of Ga(NO ₃ ) ₃ xH ₂ O (1.25 mmol, assuming x = 8), Mg(NO ₃ ) ₂ 6H ₂ O (0.20 mmol) or Zn(NO ₃ ) ₂ 6H ₂ O (0.20 mmol), and glycine were dissolved in 25 ml of water by keeping a glycine-to-metal ion ratio of 1.2. The solution was slowly evaporated at 120 ⁰ C, until a transparent residue was obtained. This was then heated to 220 ⁰ C. The combustion reaction took place and a brownish-yellow colored product was obtained. The resulting solid was then heated in a flowing air at 650 ⁰ C for 5 hours to give white product consisting Of Ga ₂ O ₃ , and Ga ₂ O ₃ :Mg ²⁺ /Zn ²⁺ .

C. Preparation of Mg ²⁺ - and Zn ²⁺ -doped GaN nanoparticles

In a typical preparation, the oxide precursor material was put in a quartz crucible and placed inside a quartz furnace. Nitridation was performed in an ammonia atmosphere. The temperature of the furnace was increased to 950 ⁰ C at a rate of 5 °C/minute. This temperature was maintained for 3 hours before it was cooled down to RT at the same rate in the NH ₃ atmosphere. The NH ₃ flow was maintained at 10 SCCM (cubic centimetre per minute at STP).

D. X-ray powder diffraction (XRD) measurements

The XRD patterns of the oxide and nitride nanoparticles were collected using a Rigaku Miniflux X-ray diffractometer with a Cr K _α (30 kV, 15 rnA) radiation source. The nanoparticle samples were gently crushed before being smeared on to a clean quartz slide. The powder diffraction patterns were collected over the 2θ range 20 to 140 ⁰ C with a scan speed and sampling width of 1 °/minute and 0.02 °, respectively.

E. Transmission Electron Microscopy (TEM) TEM was performed on the various nanoparticle and composite samples using a Hitachi H-7000 microscope operating at 75 kV. The microscope was equipped with a charge-coupled device (CCD)-camera. The samples were prepared by dispersing ~1 mg of nanoparticles in 1 gram of ethanol. These nanoparticle dispersions were drop-casted on a formvar/carbon film supported on a 300 mesh copper grid (3 mm in diameter) and allowed to dry in air at room temperature.

F. Raman spectroscopic measurements

Raman spectra were collected by exciting the sample with 632.8 nm He-Ne Laser by Melles Griot. The solid sample was evenly spread over a clean glass slide. Each spectrum is an average of 6 scans, each collected over 30 seconds.

G. Photoluminescence (PL) measurements

Room temperature PL measurements on the Mg - and Zn -doped GaN samples were performed using a 325 nm Omnichrome Series 74 He-Cd Laser by Melles Griot. The laser beam was focused on the Mg -/Zn -doped GaN particles through the microscope. A SpectroPro-500 monochromator by Acton Research Corporation was used to scan the PL signal in the visible range (350nm - 700 nm). The signal was amplified by a differential preamplifier and then acquired by the computer.

H. Energy dispersive X-ray spectroscopy

Energy dispersive X-ray spectroscopy was done using a Hitachi S-3500N scanning electron microscope, operated at 20 kV and a resolution of 102 eV. Dry powdered samples were attached to the substrate using a double-sided carbon tape, and mounted onto the sample holder. Each reported value is the result of three measurements.

I. Results and Discussion

The Mg ²⁺ - and Zn ²⁺ -doped GaN nanoparticles were prepared by a solid state reaction. These nanoparticles were prepared by the nitridation of their corresponding oxides, which were produced by the glycine combustion method to yield small nanoparticles typically in the size range 5 to 10 nm. Three series of samples have been prepared, in which the percentage ratio X/Ga (X = Mg, Zn) in the initial mixture was 3%, 7.5%, and 15%. For preliminary work on the synthesis and characterization of the Mg ²⁺ - doped GaN nanoparticles the XRD patterns of the intermediate products were refined only taking into account the presence of Ga ₂ MgO _4. A careful re-examination of the XRD furnished more insight on the formation of the intermediate products and the presence of Ga ₂ O ₃ cannot be excluded. Both Ga ₂ MgO ₄ and Ga ₂ ZnO ₄ present diffraction patterns which are very similar to Ga ₂ O ₃ and the broadness of the peaks in nano-crystalline samples might as well hide both contributions. As shown in FIGS. 19 and 20, the reference patterns for Ga ₂ MgO ₄ and Ga ₂ ZnO ₄ should present some peaks at high and low 2θ that are not close to any other peak Of Ga ₂ O ₃ and should be indicative of the presence of the double oxides (122°, 130° for Ga ₂ MgO ₄ , and 27°, 109°, 120°, 129° for Ga ₂ ZnO ₄ ). However, the intensity of these peaks is extremely low or absent; therefore, Ga ₂ MgO ₄ and Ga ₂ ZnO ₄ , if present, might constitute a very small contribution. It is thus more reasonable to infer that Mg ⁺ and Zn ⁺ are already in the Ga ₂ O ₃ as dopants.

The relative amounts OfMg ²⁺ and Zn ²⁺ in the intermediate products is evident from the EDX analyses, which are reported in Table 1, below.

Table 1

X/Ga molar ratios (X = Mg or Zn) are reported for each sample. They are weighted in the initial mixture (second row) and measured with EDXs in the intermediate oxides (third row), and in the doped nitrides (fourth row).

The concentrations in the intermediate oxides are proportional to the amounts in the initial mixture, but they are a little bit higher, especially for the lower concentrations. This occurs because part of the gallium does not reach the intermediate products. The relative amount of X also decreases with the concentrations; this suggests that the loss of gallium is proportional to the amount of gallium that does not form the intermediate product.

Despite the possibility of having various intermediate products, the second step of the reaction leads without any doubt to the formation of Mg - or Zn - doped GaN, as the XRD patterns in FIGS. 21 and 22 clearly show the wurtzite structure of GaN. In the case of Mg - doped GaN, FIG. 21 shows an emerging peak at 65° and a shoulder at 99°, which increase with the concentration, denoting the increasing presence of some MgO. In FIG. 22 no observable presence of ZnO can be discerned. The reason for this could be the low boiling point of zinc (907.0° C), which is reached during the nitridation. Hence zinc may be lost during the nitridation. The nitridation is done in a reducing environment (i.e. NH ₃ ), so the reduction of Zn ⁺ to Zn metal is possible. It could also be so small in size or amorphous and thus unobservable with XRD. On the other hand, the boiling point of magnesium is 1107.0° C, which makes magnesium evaporation occur more slowly. The doping levels of Zn ⁺ and Mg ⁺ in the GaN nanoparticles can again be appreciated with the EDX data. Table 1 shows the relative percentage amounts of dopants with respect to gallium. The concentration of dopants is higher in the case of the samples with magnesium. For the zinc doped samples the concentration of

the dopant in GaN is also inversely proportional to the amount OfZn(NOs) ₂ -OH ₂ O in the initial reaction mixture. The reason for this could be the low boiling point of zinc (907.0° C), which is reached during the nitridation. Hence zinc may be lost during the nitridation. The nitridation is done in a reducing environment (i.e. NH ₃ ), so the reduction of Zn ⁺ to Zn metal is possible. On the other hand, the boiling point of magnesium is 1107.0 ⁰ C, which makes this evaporation more slowly. The size of the nanoparticles is between 2 and 5 nm in all cases, it was determined by TEM, and is reported for each sample in Table 2, below.

Table 2

Second row: concentration of the dopants in the GaN nanoparticles. Third row: size of the doped nitrides nanoparticles estimated from the TEM images.

FIG. 23 shows a representative TEM image of one of the samples (2.9 % doped Zn:GaN) and shows roughly spherical particles (see supporting information for TEM images of the other samples).

Both Mg - and Zn -doped GaN nanoparticles exhibit bright blue emission at 425 nm when excited at 325 nm. This PL peak position is red-shifted by roughly 50 nm with respect to the undoped GaN nanoparticles. For example, FIG. 24 displays the emission spectra of 3% Zn -doped GaN nanoparticles and the undoped GaN nanoparticles. The shift in the emission is attributed to the doping of Zn ⁺ ions into the GaN matrix. However the shoulder at 395 nm in the spectrum of the sample denotes the presence of some residual undoped GaN. Similar blue emission has been observed and well-studied for the Mg - and Zn -doped GaN in films. These peaks are attributed to donor-acceptor pair recombination, preferentially from the conduction band to the shallow acceptor states (ca. 0.5 eV above the valence band) created by the incorporation OfMg ²⁺ or Zn ²⁺ . Upon excitation in the UV region,

most of the electrons from the conduction band non-radiatively relax to the defect level from where the PL originates. In addition to these peaks there is also a broad band observed around 600 nm, which is more pronounced in the case of undoped GaN nanoparticles compared to Zn ²⁺ -doped nanoparticles. The origin of this peak is still an unresolved problem in the GaN literature. This is normally denoted as the "yellow emission" and assigned to the certain type of defects in the GaN lattice. Size dependent effects between different samples likely can be excluded, because the size of the nanoparticles observed with TEM is very similar for all the samples (Table 2). The size distribution may cause inhomogeneous broadening of the PL bands. FIG. 25 shows the PL spectra of the 3% Mg ²⁺ -doped GaN nanoparticles with the corresponding Zn ²⁺ -doped nanoparticles. Though the main PL peak for both Mg and Zn -doped GaN nanoparticles appears at 425 nm, at a closer look, the emission spectrum of the Zn ²⁺ -doped GaN nanoparticles show an additional shoulder towards longer wavelength. This is clear from FIG. 25. Moreover, the Gaussian fitting of the emission peaks of Zn ²⁺ -doped nanoparticles shows two peaks with peak maximum centering at 427 and 458 nm, respectively (Inset of FIG. 25). The multiple peaks could be attributed to the presence of additional defect states created by the incorporation of zinc ions into the GaN matrix. It has been discussed in the literature that Zn ²⁺ ions not only occupy the gallium site (Znoa) but also zinc at nitrogen sites (ZnN, Zn ^~ N or Zn ² N)- The latter scenario leads to additional luminescence bands at lower energies. The sharp emission at 427 nm can arise from the Zno _a whereas the origin of the one near 458 nm is attributed to the zinc ions at the nitrogen vacancy. This difference in the luminescence is possibly due to the different reactivity (solubility) of the Mg ⁺ and Zn ⁺ ions during the growth of GaN nanoparticles.

The difference in the optical properties between the Mg ²⁺ - and Zn ²⁺ -doped GaN nanoparticles is not limited to the difference in the shape of the PL peak but also the defect-related yellow emission as well as the observed PL trend for different doping levels (see below). The observed yellow emission near 580 nm is more pronounced in the Mg -doped GaN compared to Zn -doped nanoparticles (FIG. 25). Similar results have been observed for Mg ²⁺ - and Zn ²⁺ - doped GaN crystals.

As stated above, though the origin of this yellow emission has been debated for long time, it is understood that C, N, O, H, etc. play an important role in the creation of certain type of defects. It is not only limited to these elements but other elements as well, and the preparation condition might also play a role. Recently, Kudrawiec et al. have demonstrated that the yellow emission is associated with surface-related defects. This was studied using Eu ⁺ -doped GaN powders, where they observed that the intensity of the yellow emission is dominant in nanosized grains compared to the micrometer-sized grains. Considering the above observation, Mg ²⁺ doping creates more defects on the GaN lattice compared to the Zn -doped GaN nanoparticles, most likely closer to the surface of the nanoparticles. The presence of more lattice defects in Mg -doped GaN nanoparticles is substantiated by the Raman analysis (see below).

FIG. 26 shows the PL trends observed in Mg - and Zn -doped GaN samples with different doping concentrations. In both cases the intensity is proportional to the measured concentration; however, due to intrinsic errors in the measurements the EDX data should be considered only to identify a trend, rather than an absolute correspondence between concentration and PL intensity.

The insertion into the lattice of ions with different size and charge in the place of gallium, produces structural modifications of the host site and at the level of the second sphere of coordination due to charge compensation mechanisms. Therefore, in order to obtain more insight on the structural characteristics, we performed Raman analyses. The Raman spectra collected from the Mg - and Zn - doped and undoped GaN nanoparticles are shown in FIG. 27. For comparison, the spectrum of the undoped GaN nanoparticles is also displayed. All three spectra display a strong peak at 570 cm ^"1 which is the characteristic E ₂ phonon frequency of the host GaN. Additionally, the spectra show a few other weak bands at 530, 650 and 727 cm ^"1 . The peak at 530 cm ^"1 is assigned to the A ₁ optical modes and the latter two bands are attributed to the disorder-activated Raman scattering. These disorder-activated bands are originally Raman inactive modes and their appearance is ascribed to the distortion in the lattice structure. As is clear from the Raman spectra the peak at 727 cm ^"1 is more intense in the Mg ²⁺ -doped GaN compared to the

Zn ⁺ -doped samples and it is quite weak in the undoped GaN nanoparticles. This clearly implies that there is more lattice rearrangement in the case of Mg ²⁺ -doped GaN compared to Zn - and undoped GaN samples.

J. Conclusions

The structural and optical characteristics of GaN nanoparticles have been studied by doping different amounts of Mg ⁺ and Zn ⁺ ions. Both nanoparticles exhibit a blue emission. The distortion of the lattice due to the incorporation of these ions into GaN host matrix is confirmed by the presence of additional modes present in the Raman spectra.

Overall, the optical characteristics of both Mg - and Zn -doped GaN nanoparticles display high reproducibility, which is very much essential to envision the application of these nanoparticles for the development of future optoelectronic devices. Example 8

A. Chemicals

All the chemicals were used as received from Aldrich: tetraethyl orthosilicate, gallium nitrate (99.98%), europium nitrate (99.98%). The anhydrous ammonia gas (99.999%) used for the nitradation was purchased from Praxiar. Milli- Q water with resistance greater than 18 Mωwas used in all experiments.

B. Sample Preparation

The samples were prepared starting from the vertical deposition of polystyrene beads (PBs) (Bangs Laboratories) of different sizes, depending on the desired position of the SB. A new quartz slide (Chemglass CGQ-06040-10) previously etched overnight with chromic acid, was vertically soaked in a 0.3 weight % water dispersion of PBs, and heated at 60 ⁰ C for about 14 hours, until complete evaporation. The self-assembly of the spheres into a face centred cubic (fee) lattice led to the formation of an opal. These opals acted as templates for the preparation of the inverse opal sample: they were infiltrated with a dispersion of 50 nm silica nanoparticles formed by the hydrolysis and condensation of tetraethyl orthosilicate

in 1 :5 ethanol solution. The pH was set at 1 with the addition of few drops of IN HCl. On the surface of these silica colloids a thin shell OfEu ³⁺ doped Ga ₂ Os was grown infiltrating the opal with Eu(NOs) ₃ and Ga(NOs) ₃ in a 1 :10 ratio in water solution, the concentration of Eu(NOs) ₃ was 1 rnM, while the cocentration of Ga(NO ₃ ) ₃ was 10 rnM. For the infiltration the quartz slide was horizontally immersed two times, for three minutes each time, in a 1 :5, by volume, mixture of the nitrates solution and the silica nanoparticles dispersion. The homogenization of the mixture provides a simultaneous infiltration of the silica and the dopants and ensures the homogeneity of the distribution of the europium all over the structure. The formation of Ga ₂ θ ₃ , the nitradation to GaN and the reduction OfEu ³⁺ to Eu ²⁺ was performed in an electric furnace (Lindberg) in a single thermal cycle. The temperature was raised to 650 ⁰ C in 6 hours, dwelled there for 2 hours and raised further to 950 ⁰ C in 6 hours. At this point NH ₃ was fluxed at a rate of 10 SCCM (cubic centimeter per minute at STP) keeping the temperature for 2 hours. After that, the temperature was decreased to room temperature in 6 hours. During the first step at 650 ⁰ C the nitrates are transformed into oxides. During the second step the oxides are transformed into nitrides and the Eu ³⁺ is reduced to Eu ²⁺ . This process leads to the formation of GaN nanoparticles on the surface of the silica. In this case, the silica was not in the form of nano-particles, but instead was a solid inverse opaline structure. At the end, the inverse opaline structure has replaced the empty spaces of the direct opal and the polystyrene beads have been completely removed, leaving spherical voids filled with air.

The two references were grown from 300 and 540 nm PBs, respectively. Such references resemble the scattering properties of the sample but their stop bands cannot affect the photonic properties of the active ion. For the quantum yield (QY) measurements study, beside this reference, yet another control sample also is needed, which mimics the scattering characteristics of sample and reference at the excitation wavelength, but which doesn't contain an optical emitter or absorber. Such control sample was synthesized following the same procedure, but without inserting Eu ³⁺ as a dopant. Both were tried without any dopant and using Y ³⁺ as a dopant. The results did not change. All the samples, references, and control

samples compared in this work were rigorously grown in the same batch and in the same thermal cycle. SEM measurements were performed with a Hitachi S35000N scanning electron microscope operating at 15 kV. All the samples and references were characterized by Energy Dispersive X-ray spectroscopy (EDX) using a Hitachi S-3500N scanning electron microscope, operated at 20 kV and a resolution of 102 eV. Dry powdered samples were attached to the substrate using a double-sided carbon tape, and mounted onto the sample holder. Each reported value is the result of three measurements in different positions. The standard deviation was calculated on the average

C. Photoluminescence measurements

Photoluminescence measurement were recorded with an 'Edinburg Instruments' FLS 920 fluorimeter. For all the measurements the detector employed was an R928P Hamamatsu PMT, and the resolution due to the slits' aperture was 1 nm. The emission spectra OfEu ²⁺ were measured exciting with a 450 W Xe arc lamp. Transmission spectra were measured with the same fluorimeter in a 180° geometry obtained by driving the light out of the fluorimeter with two optical fibres and two objectives to focus the light in and out of the sample with two Olympus 1Ox objectives. The lifetimes of the same ion were collected exciting with the 355 nm harmonic line of a Quantel Nd: YAG nanosecond laser and collected with a multi channel scaling card, whose time resolution was 5 ns. The lifetime values were calculated as effective lifetimes by using the formula:

\ t - I - dt τ = ±

Equation 5 where τ is the lifetime, t is the time and I the intensity. Each decay was considered until the intensity reaches 1% of the initial intensity. The color coordinates were determined from the Tristimulus values calculated by the integration of the spectra after the application of the color matching functions, which account for the human

eye sensitivity. The absolute QY was determined using an integrating sphere (Edinburgh instruments, 150 mm in diameter coated with barium sulfate). All the samples and references were placed in a cuvette inside the integrating sphere. The geometry of the measurement was 90° and a baffle was placed beside the sample on the emission monochromator side, in order to avoid that light directly scattered from the sample could be collected without bouncing on the walls of the sphere at least once. The reason is that a direct reflection could compromise the measurement because the samples could scatter light differently along different directions. The QY was calculated using the formula:

Equation 4

QY is the quantum yield, l∑ is the emission intensity, Is-c and Iss are the intensities of the light scattered from the control sample and from the sample, respectively and v is the wave-number. Hence equation (4) expresses the ratio between the number of photons emitted (numerator) versus number of photons absorbed (denominator). The Electron Paramagnetic Resonance (EPR) spectrum was collected on a Brucker EMX instrument operating in the X-band (9.443 GHz) at 115 K.

D. Results and Discussion The growth method for the samples was optimized to maximize the quality of the opal. For instance, the concentration of the Polystyrene Beads (PBs) dispersion and the temperature of the oven, were adjusted to obtain a uniform coating of the quartz slide. In particular a slow evaporation (2 days) was preferable. The concentration was proportional to the thickness of the coating. A thick coating was generally uneven and not functional for optical transmission measurements. On the other hand a too thin coating was not uniform over the area of the slide and the quality of the face centred cubic (fee) structure was low. The best compromise was provided by adjusting the concentration to just cover the slide completely. It is currently believed that effective thicknesses are from greater than 0 to at least about

10 micrometers, and more likely from about 1 to about 3 micrometers. Any attempt to increase the thickness decreased the quality of the structure, producing a wavy profile of the layer and a thickness step at about half of the slide's height due to a natural temperature oscillation and a sudden concentration increase. The quality of the opal also was influenced by the size of the beads: the smaller the beads, the easier it was to get a good quality structure. In particular it was fairly easy to obtain good quality opals with PBs smaller than 800 nm. If the size of the beads exceeded 900 nm the structure of the opal and the intensity of the SB decreased sensitively. FIG. 28a is a picture of an opal made from 400 nm PBs taken from an optical microscope on a 10x magnification, some longitudinal lines are observed on the layer, but the quality of the fee structure is confirmed by the intensity of the stop band in the transmission spectrum measured perpendicular to the 111 planes of the lattice (FIG. 29, dotted line). The barycentre of the stop band in the spectrum matches the expected value calculated from equation. λ = ^1Sa [φn _{ι +} (l - φ)n ₂ ] niλlh + k + 1

Equation 2

Where λ is the wavelength, S is a shrinkage factor, which takes into accounts the eventual shrinkage that a structure undergoes during its formation (vide infra), α is the cell's parameter, m is the order of Bragg 's diffraction, n _γ and n ₂ are the refractive indexes of the materials constituting the structure, and φ is the volume fraction of one of them, the other being the complementary (l - φ) . Other stop bands corresponding to the 200, 220 and 311 series of planes are much less intense and negligible for our purposes.

For preparing the inverse opals the number and duration of each immersion was optimized at three minutes after many attempts, in order to have the most copious infiltration and the lowest possible damage to the lattice. Infiltration was driven by capillary forces in the convex structure and it was particularly effective as proved by the result shown in FIG. 30a. A constant 2 μm thickness of the sample can be appreciated from the Scanning Electron Microscopy (SEM) images in FIGS. 30 b, c, d, where the number of layers can be counted at the extremities of the

sample and in the middle (FIG. 30a and 30b respectively). These figures also confirm the quality of the infiltration, which doesn't lead to any accumulation of silica on top of the inverse opal structure. The SEM images are slightly hazy because of some charging during the measurements. The opalescent effect on the coloration of these samples is very strong. In the regions where they are not damaged they assume the color corresponding to the position of the stop band, when the light incides from the same direction of the observer, and the complementary color of the stop band, when the light comes from the opposite direction of the observation. FIG. 28b-d shows the intensity of the effect. Obviously the infiltration of the opal damaged the coatings a bit: in some parts it came off completely, and generally it was fragmented in "small islands" of roughly 0.1 X 0.2 mm. The reduced intensity of the stop band of the inverse opals observed in the transmission spectra is thus due to this fragmentation combined with the size of the beam (~0.2 mm) used in the measurements. The transmittance of 70 % observed in FIG. 29, is hence an average transmittance over the whole irradiated area of the slide, and therefore an underestimation of the real effect in these materials, due to the fact that some light is not passing through the sample. The shift in the position of the stop band passing from the opal to the inverse opal, according to equation (2), is mainly due to a shrinkage occurring during the thermal cycle. The voids are normally between 30 and 40% smaller than the PBs employed. Factor S in equation (2) takes into account this shrinkage.

For this example, two references were needed: one to check the effect of the stop band on the Eu ⁺ emission; and one to be used as control sample for the determination of the absolute QY. The best one to study the effect of the stop band is an inverse opal grown in the same way as the sample, doped with Eu ²⁺ , but starting from a different size of PBs in order to have the stop band outside the Eu ²⁺ emission. FIG. 31a shows the transmission spectra of the inverse opals compared with its two references. The sample grown from 400 nm PBs and the references grown from 300 and 540 PBs present stop bands at 477, 325, 662 nm, respectively. The resulting shrinkage is 40% for the sample and the 325 nm reference and 35%

for the 662 nm reference. The shrinkage seems to be higher for smaller particles/holes; this trend was confirmed in all the samples measured.

Each sample and reference underwent elemental analysis by Energy Dispersive X-ray Spectroscopy (EDX). The ratio between europium and gallium was 9.4±0.7 %, 10.3±0.8 and 9.8±0.7 in the sample, the reference grown from 300 nm PBs, and the one grown from 540 nm PBs, respectively. The amount of gallium with respect to silica was about 10%. These ratios are very close to the ratios used in the experiments.

The clearest evidence of the presence OfEu ²⁺ in the sample comes from the broad and intense emission corresponding to the de-excitation of the t2g level of the 4f65d configuration to the ground level 8S7/2. This band is centered at around 500 nm, but its position changes sensitively changing the NH3 flux during the nitradation. As a result, whenever samples of different size and different position of stop band are compared, these samples have been grown together undergoing the infiltrations in the same Petri dish and the same thermal cycle. Otherwise, differences could be due to slightly different pressures of NH ₃ in the cylinder from one day to another, etc.

Another confirmation of the presence of the Eu ²⁺ came from the Electron Paramagnetic Resonance (EPR) measurements (FIG. 33). The key feature is the 'step' that occurs between 320-340 mT. A step that occurs between 320 and 340 mT is typical OfEu ²⁺ , and its broadness is due the random orientation of the sample. The presence of the stop band was readily observed in transmission spectra, but in order to be able to observe it in emission the quality of the structure had to be refined. Part of the reason is due to the angle of measurement: in transmission spectroscopy the geometry of the measure is 180° and perpendicular to the 111 series of lattice planes, which is the direction in which the effect of the stop band is more intense. On the other hand, in emission spectroscopy the geometry of the measure was 90°, and the angular information was lost because a big portion of the emitted light was collected by the lens on the excident beam. For this reason, the stop band seen in emission is just an average effect over several different angles.

FIG. 3 Ib clearly shows the presence of the stop band as a dent at about 447 nm. The

position of the stop band is reproducible in different samples because it depends on the size of the initial polystyrene beads. On the other hand the position of the emission maximum changes from growth to growth because it depends on the ammonia flux. In each example, the ammonia flows for more than 8 hours and small uncontrollable differences in the intensity of the flux from one example to the other lead to a different position of the emission maximum of Eu , resulting in different relative position with respect to the stop band. Attempts were made to optimize the characteristics of the sample having the stop band on the long energy side of the emission maximum; nevertheless, the best sample that was obtained presented the stop band on the high energy side. This was probably due to the fact that the most intense stop band is obtained starting from small beads.

The emission of the reference grown from 300 nm PBs is superimposed on the same graph, the emission of the reference grown from 540 nm PBs is not reported because is essentially similar to this one. The parallelism with FIG. 3 Ia on top confirms that the stop bands on the references are outside the Eu ²⁺ emission band and, as expected, they do not affect its density of states. In order to facilitate the comparisons between the positions of the bands' maxima, the spectra have been normalized. Therefore, the relative intensities cannot be compared, but just the shape of the spectra. The original intensities before normalization couldn't be compared either, because the amount of material excited in different samples is not necessarily the same. In the literature a number of papers can be found dealing with the orientational redistribution of light by photonic crystals. However, it is hard to find clear evidence of the redistribution in wavelength, which is however expected from theory. The shift in the emission maximum from 487 nm to 496 nm, shown in FIG. 3 Ib with respect to the samples with the stop band outside of the Eu ²⁺ emission, confirms this expectation. Another confirmation of this shift comes from the color coordinates calculated from the emission spectra of the sample (x = 0.242, y = 0.371) and the 300 nm (x = 0.238, y = 0.367) reference. Even more interesting is the confirmation coming from the QY measurements where the QY measured for the sample and for the references are 5.13 ± 0.2% and 4.99 ± 0.2%, respectively. The error on the QY value has been determined by repeating each measurement 10

times. These QYs are smaller than the ones reported in a previous work by Mahalingam et al. The emitter is Eu ²⁺ albeit in a slightly different crystal field as evidenced by the different emission maxima. The constancy of the QY validates the theory that the wavelength redistribution is achieved without reducing the emission yield. Of course the angular redistribution is still present, but from examples where all the light is collected out of the sample inside the integrating sphere this effect is completely averaged. As a further confirmation of this intensity redistribution at different wavelengths, what is actually happening on the stop band was checked and on its edges by monitoring the lifetime, in order to have an actual picture of the reduction and increase of density of states. The decays of Eu ²⁺ were not exactly mono-exponential, because of the non-homogeneity of the crystal field around the Eu ²⁺ . These experiments were carried out in a 90° geometry, the angular effect cannot be considered and the stop band is again an average over several directions. This is another non-negligible contribution to the broadness of the stop band. If the density of states decreases, according to equation (3) the lifetime becomes longer, on the other side if the density of states increases the lifetime becomes shorter.

W = 2πh V ₄ P fe)

Equation 3

With reference to equation 3, W is the transition rate, fi is the reduced Planck constant, V _β is the matrix element of the potential that operates between the initial and final value, and P\E _fi ) is the DOS at the energy of the transition.

FIG. 32a shows the behavior of the lifetimes at the wavelengths around the stop band. From the comparison with the same measurements on the references the lengthening within the range of the stop band is clearly seen. On the edge of the stop band the value at 480 nm definitely confirms the expected behaviour. FIG. 32b is a ratio between the lifetime of the reference grown from 300 nm PBs and the sample (the ratio with the other reference is not reported because it gives analogous values). When the ratio is close to one, the stop band has no effect on that particular wavelength. The increase and decrease of the ratio reflects the behavior of the density of states .

Employing photonic crystals in such a way it is possible to concentrate the emission intensity of an emitter on a desired range of wavelengths. By having a constant QY, the reduction in DOS in the range of the SB is accompanied by an increase of DOS and hence QY on the edge of the SB, which determines an increase of color purity of the emission. In the same way, by tuning the stop band on the low energy side of the emission band of an emitter, it would be possible to increase the efficiency of the device in the high energy range, which would be especially attractive in the ambit of the current quest for an intense emitter in the blue and at higher energies, extremely useful for a large number of applications, like data storage, high energy lasers, photodiodes, etc.

E. Conclusion

In this example, a shift in the wavelength of the emission of an emitter by the action of a photonic crystal modifying its density of states was observed. Its effect resembles a smart filter capable of modifying the emission intensity without any loss of QY. This not only confirms the expectations coming from the theory, but it also opens the possibility to mold the intensity profile of an emitter. In particular, tuning the stop band on the long wavelength side with respect to the maximum of the emission, it is possible to increase the efficiency of a device on the blue side of the spectrum and at higher energy.

In view of the many possible embodiments to which the principles of the disclosed invention may be applied, it should be recognized that the illustrated embodiments are only preferred examples of the invention and should not be taken as limiting the scope of the invention. Rather, the scope of the invention is defined by the following claims. We therefore claim as our invention all that comes within the scope and spirit of these claims.