Engineered structure for solid-state light emitters

Engineered structure for solid-state light emitters

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to solid-state optical devices, and more particularly, to a design structure of a semiconductor film including luminescent centers for use in a solid-state light emitter.

The next generation of solid state lighting seeks to deliver advances in brightness, efficiency, color, purity, packaging, scalability, reliability and cost savings. The creation of light emitting devices from silicon-based materials on which the modern electronics industry is built has been the subject of intensive research and development around the world. A major obstacle has been the indirect energy gap in bulk silicon, which limits efficiency to a very low degree. However, one specific technique based on silicon nanoparticles formed through various techniques, for example, nanocrystals, could overcome this difficulty.

US Patent No. 7,081,664 for "Doped Semiconductor Powder and Preparation Thereof," issued July 25, 2006 in the name of Hill; US Patent No. 7,122,842 for "Solid State White Light Emitter and Display Using Same", issued Oct. 17, 2006 to Hill; US Published Patent Application No. 2004/ titled "Broadband Optical Pump Source for Optical Amplifiers, Planar Optical Amplifiers, Planar Optical Circuits and Planar Optical Lasers Fabricated Using Group IV Semiconductor Nanocrystals," published August 5, 2004 in the name of Hill 151461; US Published Patent Application No. 2004/214,362, entitled "Doped Semiconductor Nanocrystal Layers and Preparation Thereof," published Oct. 28, 2004 to Hill et al.; and in U.S. Published Patent Application No. 2004/252,738 entitled "Light Emitting Diodes and Planar Optical Lasers Using IV Semiconductor Nanocrystals," published December 16, 2004 in the name of Hill, incorporated herein by reference. Light emitting devices of the prior art are silicon dioxide (SiO<sb>2</sb> The use of silicon-rich silicon oxide (SRSO) composed of silicon nanoparticles embedded in a matrix (or glass) reduces many of the problems associated with bulk silicon, and rare earth ions from the excited nanocrystals It has been demonstrated that, due to the high efficiency of the energy transfer process to the cells, it can exhibit efficient room temperature rare earth luminescence when doped with erbium or other rare earth materials. SRSO thus provides an alternative to thin-film electroluminescent materials. Silicon nanoparticles act as a traditional photosensitive system that absorbs incident photons or electrons, and then transfers energy to rare earth ions, which have several advantages compared to the direct fluorescence of rare earths, but with infrared or visible light. It emits fluorescence in the wavelength range of the line. First, the absorption cross-section of silicon nanoparticles is larger than the third order size compared to rare earth ions. Second, the incident photon needs to resonate with one of the narrow absorption bands of the rare earth because excitation takes place between the rare earth ions and the carrier in the silicon nanoparticles, either through Auger-type interactions or through the Forster transition process. there is no Unfortunately, approaches to develop these silicon nanoparticle materials have only been successful in producing very low concentrations of rare earth elements, which are not sufficient for many practical applications.

Silicon nanoparticles formed by these techniques have a relatively narrow distribution of photo-luminescent (PL) wavelengths or energies despite their wide size distribution, i.e., the observed energy is only expected from quantum confinement of nanocrystals. It was observed that not very high. The reduced nanoparticle excitation energy affects the efficiency of energy transfer from electrons that conduct electricity when these structures are powered, severely limiting the efficiency of light generation from such films.

In general, the production of type IV semiconductor nanoparticles doped with rare earth elements is done by implanting silicon ions into a silicon oxide layer, followed by high-temperature annealing to form silicon nanoparticles and reduce ion implantation damage. Implantation of rare earth ions into the annealed silicon nanoparticle oxide layer is followed by implantation of silicon ions. The resulting layer is annealed again to reduce ion implant damage and optically activate the rare earth ions.

There are several problems with this method:

i) it reduces the uniformity of the layer surface due to ion implantation;

ii) it requires an expensive ion implantation step;

iii) It fails to obtain a uniform distribution of group IV semiconductor nanoparticles and rare earth ions unless many ion implantation steps are performed.

iv) it requires a balance between reducing ion implant damage by thermal annealing and maximizing optically active rare earths: and

) The thickness of the film is limited because the implanted ions do not penetrate deeply into the film with respect to the actual ion implantation energy.

To reduce the above drawbacks, plasma enhanced chemical vapor deposition (PECVD) has been used to make the group IV semiconductor nanoparticle layer. In order to form the group IV semiconductor nanoparticles and to optically activate the rare earth ions doped in the nanoparticle region, the prepared layers are subjected to a rare earth ion implantation step and a subsequent annealing cycle. Unfortunately, the layers prepared in this way still undergo an implantation step, which results in poor surface uniformity, non-uniform distribution of rare earth elements and limited film thickness.

Other deposition methods that have been used to obtain doped group IV semiconductor nanoparticle layers include co-sputtering group IV semiconductors and rare earth metals, particularly in oxygen plasma. In this method, a group IV semiconductor and a rare earth metal were placed on a target substrate, which was then placed into a vacuum chamber and exposed to an argon ion beam. The argon ion beam sputtered group IV semiconductors and rare earth metals, which were then deposited onto a silicon wafer. The newly formed film on the silicon wafer was then annealed to grow the nanoparticles and optically activate the rare earth ions. Doped group IV semiconductor nanoparticle layers made in this way have the following disadvantages: i) the layer does not have a very uniform distribution of nanoparticles and rare earth ions; ii) suffer an up conversion efficiency loss due to rare earth clustering in the film; and iii) the concentration of rare earth ions in the film is limited to about 0.1%.

It is desirable that the concentration of one or more rare earth elements in the semiconductor nanoparticle layers be as high as possible because the level of response of the film to an external stimulus, such as an optical stimulus for photoluminescence, is proportional to the concentration. A problem encountered when high concentrations of rare-earth elements are present in the semiconductor layer is that when two rare-earth metals are brought into close proximity to each other, a quenching relaxation interaction that reduces the level of the optical reaction occurs. The concentration of the rare earth element in the semiconductor film is therefore balanced to be as high as possible to provide maximum fluorescence, but small enough to limit the quenching action.

Silicon nanoparticles formed by these techniques generally have a relatively wide size distribution and a similarly wide spatial distribution, i.e., the spatial distance between the nanoparticles, which prevents them from conducting electrons when these structures are energized. affects the energy transfer efficiency of The average distance between the nanoparticles in the direction of electrical conduction must be large enough so that the electrons get sufficient energy from the electric field between the nanoparticles to excite the luminescent object and generate photons of the correct color. However, since the spatial distribution is isotropic, the total concentration of nanoparticles in these films should be very low (~5x10<sp>18</sp> cm<sp>-2</sp>). Unfortunately, with such low nanoparticle concentrations and nanoparticle size and spacing distributions, severe limitations are placed on the efficiency of photogenerating capabilities from these films with embedded nanoparticles.

When the rare earth ions are introduced into the film, it is desirable to place the rare earth ions in close proximity to the nanoparticles to promote efficient energy transfer from the excited nanoparticles to the rare earth ions. However, ion implantation or in situ deposition techniques introduce a random distribution of rare earth ions. In particular, the production of white light requires that multiple types of rare earth ions be introduced into the film because different types of rare earth ions provide different colors. It is impossible to ensure that the correct rare-earth ions are located near nanoparticles of an appropriate size so that the energy of the excited nanoparticles matches the emission wavelength of the rare-earth ions. In other words, it is very likely that the nearest rare earth ion will radiate at a wavelength that is too short. That is, they cannot be excited by nanoparticles of too long wavelength, that is, energy is wasted in the excitation process. Even if sufficiently high concentrations of rare earths are used to avoid such (at least in part) mismatch of the excited nanoparticle energy to the radiative emission wavelength of the rare earth ions, losses still occur from interactions between the rare earths when they are closely spaced.

In films formed such that nanoparticles have significant variations in size and spacing distance, excess silicon atoms and dopants such as rare earth are introduced uniformly throughout the film, but not locally. Therefore, it is possible that some of the excess silicon atoms may be located far from any nucleation sites and thus not participate in the nanoparticles, but instead still remain distributed within the silicon dioxide host matrix. In addition, some rare earth ions may not be located close enough to the nanoparticles. Finally, if a significant carbon content must be introduced into the nanoparticles to increase the excitation energy, the carbon atoms are required to be located close to the nanoparticles. It is generally observed that silicon nanocrystals with a diameter of 2 nm without carbon incorporation should have excitation energy on the order of 2.3 eV from the quantum confinement effect, but only emitting in the 1.4-1.8 eV range. If impurities such as excess silicon atoms, rare earth ions and carbon atoms remain in the oxide matrix, they can very adversely affect the physical properties of the oxide, especially the breakdown field and therefore the reliability and lifetime of the device. .

It is an object of the present invention to overcome the shortcomings of the prior art by providing a multilayer design structure in which wide bandgap semiconductor or dielectric buffer layers are disposed adjacent to a very thin active light emitting layer designed to emit light at a specific wavelength. The buffer layers provide the correct distance in the direction of electrical conduction so that electrons gain sufficient energy from the electric field as they pass through the buffer layers to excite the light emitting center in the active layers to produce photons of the correct color through either bombardment ionization or bombardment excitation.

Accordingly, the present invention provides a first active layer comprising a light emitting center of a predetermined concentration for emitting light at a first wavelength; a first dielectric layer comprising a wide bandgap or dielectric material adjacent the first active layer; and an electrode set for applying an electric field to the first active layer and the first dielectric layer; wherein the first dielectric layer allows electrons to pass through the first dielectric layer to excite the light emitting centers in the first active layer through impact ionization or impact excitation at an excitation energy for emitting light at the first wavelength. It relates to a light emitting structure having a thickness to obtain sufficient energy from the electric field when passed through.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention will be described in more detail in conjunction with the accompanying drawings which provide preferred embodiments.

1 is a cross-sectional view of an electroluminescent solid state device according to an embodiment of the present invention.

Figure 2 is a cross-sectional view of a super-lattice semiconductor structure conforming to the device of Figure 1;

Fig. 3 is a cross-sectional view of another super-lattice semiconductor structure conforming to the device of Fig. 1;

Fig. 4 is a cross-sectional view of another super-lattice semiconductor structure conforming to the device of Fig. 1;

1 , an embodiment of the present invention provides an electroluminescent solid state device 1 comprising a conductive substrate 11 such as an N-type or P-type silicon wafer. A light emitting film structure 20 comprising one or more relatively thin active layers having light emitting centers such as semiconductor nanoparticles in a wide bandgap semiconductor dielectric matrix or other semiconductor material is deposited over the top of a conductive substrate 11 . Film structure 20 may be deposited by one of many suitable methods, such as plasma enhanced chemical vapor deposition (PECVD), molecular beam epitaxy, pulsed laser deposition, sputtering, and a sol-gel process. An upper optically transparent, current injection (electrode) layer 21 , for example iridium tin oxide (ITO), is mounted over the film structure 20 , which, along with the backside electrical contacts 25 , receive AC or DC power. allow it to be applied. Preferably, the transparent current injection layer 21 has a thickness of 150 nm to 500 nm. Preferably, the chemical composition and thickness of the transparent current injection layer 21 causes the light emitting structure 20 to have a resistance of less than 70 ohm-cm. A buffer electrical contact 22 , for example TiN, is positioned between the transparent current injection layer 21 and the overlying electrical contact 23 , for example a metal such as aluminum. The buffer electrical contact 22 provides an ohmic contact point between the front transparent current injection layer 21 and the top electrical contact 23, while the top electrical contact 23 is a suitable surface for wire bonding contacts. provides Other suitable materials for the transparent current injection layer 21 and the buffer electrical contact 22 may optionally be employed. A back reflector 24 is provided between the film structure 20 and the substrate 11 to reflect the light internally emitted towards the substrate 11 back towards the emitting surface, ie, the transparent current injection layer 21 . can

The substrate 11 on which the film structure 20 is formed is selected to withstand high temperatures of the order of 1000° C. or more. Examples of suitable substrates are silicon wafers or polysilicon layers, which are for example 1x10<sp>20 </sp>inside<sp></sp>5x10<sp>21</sp>/cm<sp>3</sp>It may be n-doped or p-doped with a dopant of , and includes fused silica, zinc oxide layers, quartz, sapphire silicon carbide, or a metal substrate. The substrate 11 may optionally have a thermally grown oxide layer, which may have a thickness of up to about 2000 nm, with a thickness of 1 nm to 20 nm being preferred. The substrate 11 may optionally have a deposited electrically conductive layer, which layer preferably has a thickness between 50 nm and 2000 nm, preferably between 100 nm and 500 nm. The thickness of the substrate is not critical as long as thermal and mechanical stability is maintained.

Film structure 20 may consist of a single active layer or multiple active layers, each layer having an independently selected composition and thickness, for example: with or without rare earth doping elements, and with or without carbon doping. Group IV (eg Si, Ge, Sn and Pb) wide bandgap semiconductors such as oxide or nitride matrices or semiconductor nanoparticles such as group IV semiconductors (eg Si, Ge, Sn and Pb) in a dielectric matrix. Optionally, the active layers may be composed of a rare earth oxide or other semiconductor material having a light emitting center activated by bombardment ionization or bombardment excitation. By using active layers having different compositions, a multicolor structure can be provided. For example, semiconductor nanoparticle layers doped with erbium, thulium and europium in a single structure are green (aterbium), blue (cerium) and red (Europium) or a color combination thereof, for example, provides a structure capable of emitting light in white. The layers may be stacked or configured side-by-side as individually controllable circuit elements.

One type of preferred multilayer structure 20 provided by one embodiment of the present invention is the super-lattice structure shown by the example of FIG. 2 . The structure comprises multiple active layers 12 and 14 with a wide bandgap semiconductor or dielectric buffer layer 13 over a substrate, for example semiconductor nanoparticles. Each of the active layers 12 and 14 has a thickness of 1 nm to 10 nm and is deposited over the substrate 11 . To generate light of the same or different wavelengths, for example, all active layers 12 emit one wavelength and all active layers 14 emit light of a different wavelength. Silver contains the same or different materials, for example rare earth doping elements. The two wavelengths of light produced by the two sets of active layers 12 and 14 are combined together or with other layers (not shown) to produce the desired color, for example white. The active layers 12 and 14 are separated by buffer layers 13 such as silicon dioxide layers. A transparent current injection layer 21 is deposited on top of the multilayer structure 20 of the superlattice structure. A thickness of 50 nm to 2000 nm is preferred, depending on the amount of voltage possible, more preferably a thickness of 150 nm to 750 nm, but there is no maximum thickness for the superlattice structure.

The structures shown in Figure 2 and subsequent figures show adjacent layers in contact with each other without intervening layers; However, additional layers may be used so long as they do not interfere with the disclosed layers. Therefore, the terms coating and contacting do not exclude the possibility of additional intervening but non-interfering layers.

By embedding small semiconductor nanoparticles in a semiconductor nitride matrix, for example group IV semiconductor nanoparticles such as silicon in a group IV semiconductor nitride matrix such as silicon, the surface passivation effect of the nanoparticles by nitrogen atoms and exciton ( Exciton), due to the influence of the strong coupling of the wave function of electrons and holes, the emission lifetime of semiconductor nanoparticles can approach the nanosecond and/or sub-nanosecond region. However, uniformly deposited SiN with semiconductor nanoparticles formed therein<sb>x</sb> Films generally have a relatively wide range of sizes and isolation distances between nanoparticles of random spatial distribution. In addition, the semiconductor nanoparticles formed in the semiconductor nitride film can form interconnected small clusters at high temperatures, which can affect the luminous efficiency, severely limiting the flexibility of the device process after film deposition. The combination of different nanoparticle sizes and separation distances can have a significant impact on the electroluminescent efficiency of semiconductor nanoparticle structures formed in such films.

In films in which semiconductor nanoparticles are embedded in a semiconductor nitride matrix, the current conduction in the films can be significantly affected by the high trap density of the semiconductor nitride host and therefore in order to obtain energy from the electric field to generate excitons in the semiconductor nanoparticles. It can adversely affect the efficiency of the injected charge carriers. However, the design structure according to the present invention eliminates all the aforementioned problems by providing buffer layers between active layers of semiconductor nitride and thereby securing an appropriate distance between nanoparticles. Moreover, by providing thin active layers, the size of nanoparticles can be more precisely controlled.

Referring to FIG. 3 , a design film structure 31 according to another embodiment of the present invention is formed by a plurality of different stacks 32 , 33 and 34 of organized layers, wherein active layers 35 , 36 and 37 are separated by buffer layers 38, 39 and 40, respectively, comprised of a pure wide bandgap semiconductor or dielectric material.

For the design film structure 31 driven by AC voltage, a pair of electrodes 52 and 53 is located on opposite sides of the stacked layers 35-40. Since current flows in both directions when the voltage oscillates, the buffer layers 38 to 40 are disposed after the electrodes 52 and 53, respectively. Ideally one of the electrodes, eg electrode 52, is transparent, eg ITO, a reflective layer or coating 50 to reflect light back through the transparent electrode 52, one of the electrodes; For example, it is added between the electrode 53 and the remaining stack layers 35 to 40 .

The size of nanoparticles, eg, nanocrystals, is approximately equal (+10%) to the thickness of the active layers 35 , 36 and 37 (or 12 and 14 ) in which they are present. The size of the nanoparticles in each of the active layers 35 , 36 and 37 , ie the thickness of the active layers 35 , 36 and 37 , is designed for a specific excitation energy to produce light emission of a desired color. For silicon nanoparticles in a silicon dioxide matrix host doped with rare earth, the theoretical relationship between the nanoparticle diameter d(nm) and the excitation energy E(eV) is given by:

<i>E</i> = 1.143 + 5.845/(<i>d</i><sp>2</sp> +1.274<i>d</i> + 0.905)-6.234/(<i>d</i><sp>2</sp> + 3.391<i>d</i> + 1.412);

For example, ~1.9 eV for red photons (<i>d</i> = 2.9 nm), ~2.3 eV for green photons (<i>d</i> = 2.1 nm), ~2.8 eV for blue photons (<i>d</i> = 1.6 nm). Rare earth ionic species placed in or after the layer of nanoparticles are selected to emit at a wavelength matching the excitation energy of the nanoparticles in the layer (or vice versa).

Group IV without rare earth doping, for example group IV in a nitride matrix host of silicon, for example nanoparticles of silicon or group IV in a silicon dioxide matrix host without rare earth doping, for example nanoparticles of silicon For , the excitation energy equation for generating a specific excitation energy to produce light emission of a desired color from the nanoparticles is shown as follows:

E = E<sb>0</sb><sb></sb>+ C/d<sp>2</sp><sp></sp>(E<sb>0</sb> = 1.16 eV and C = 11.8 eV-nm<sp>2</sp>)

Thus, the thickness of the red light emitting layer, that is, the diameter of the nanoparticles in the active layer with silicon nanoparticles in the silicon nitride matrix is 4 nm, 3.25 nm for the green light emitting layer, and 2.6 nm for the blue light emitting layer am.

The thickness of active layers without nanoparticles is usually determined experimentally based on a compromise between energy requirements and light brightness. On the other hand, if the active layer is extremely thin, the energy can be accurately known for the entire layer and thus the energy matching can be optimized; However, if the active layer is extremely thin, there is no light emitting center and no light. Since there can be more light emitting centers per square mm, the thicker the active layer, the brighter the layer can be; However, there may be a loss of efficiency as the energy cannot be optimized through the entire thickness.

The thickness of the buffer layers 38 , 39 and 40 (or 13 ) is closely matched to the size of the nanoparticles in the neighboring nanoparticle active layers 35 , 36 and 37 (or 12 and 14 ). In order to excite the nanoparticles with the correct energy for an electric field perpendicular to the plane of the layers 35-40, the electrons must obtain sufficient energy from the applied electric field - obtained in the buffer layers 38, 39 and 40 (in eV). The energy (measured) is equal to the electric field multiplied by the thickness of the buffer layers 38 , 39 and 40 . For example, for an applied electric field of 5 MV/cm, the thickness of the buffer layer should be 3.8 nm or thicker to excite the nanoparticles to 1.9 eV (1.9 eV / 0.5 eV/nm = 3.8 nm) and 2.3 eV. For excitation it must be 4.6 nm or thicker, and for excitation at 2.8 eV it must be 5.6 nm or thicker. For a design film structure 31 powered by AC power, where neighboring nanoparticle layers (e.g. 35 and 36) emit light at different wavelengths, an interposed buffer layer (e.g. 38) must be thick enough to excite the nanoparticles in the higher energy layer.

Based on silicon nanoparticles in a silicon oxide matrix and doped with rare earth ions or other impurities such as carbon, the film structure 31 has a luminous flux (optical output), efficiency (internal power conversion efficiency and external luminous efficiency), It provides very large improvements in color rendering index (CRI), device reliability and lifetime, and device manufacturability/cost/yield of solid state light emitting devices.

Rare earth ions may be introduced into active layers 35 , 36 and 37 , buffer layers 38 , 39 and 40 , or both. The preferred structure introduces the rare earth only into the active layers 35, 36 and 37 such that the energy transfer efficiency from the nanoparticles to the rare earth ions is maximized and the radiative emission efficiency of the excited rare earth ions is at a concentration that is maximized. Because of the complexity of the physical process involved, optimization is usually an empirical process. Rare earth ionic species placed in or after the nanoparticle active layer are selected to emit at a wavelength that matches the excitation energy of the nanoparticles in the active layer (or vice versa). Preferably, the rare earth elements are cerium, praeseodymium, neodynium, promethium, gadolinium, erbium, thulium, Lanthanide elements such as ytterbium, samarium, dysprosium, terbium, europium, holmium or lutetium: , they may be selected as actinide elements such as thorium.

Although other impurities may be placed anywhere in structure 31, they will usually only be introduced into active layers 35, 36 and 37, if necessary. For example, excitation of nanoparticles delivered to rare earth ions in a broad bandgap semiconductor or dielectric, e.g., silicon oxide, matrix, because the measured excitation energy of nanoparticles is not observed to be as high as theoretically expected. Carbon atoms may be needed to boost energy.

Buffer layers 38, 39 and 40 should be of the highest quality so that device lifetime and reliability under high applied electric fields can be maximized. That is, it must be compact with few defects, and such a material must be obtainable within the possibilities of a particular process technology.

Silicon-rich silicon oxide with or without carbon and rare earth doping for active layers 35 , 36 and 37 and silicon dioxide for buffer layers 38 , 39 and 40 are preferred materials in the design film structure. Other material systems such as silicon-rich silicon nitride with or without rare earth doping for active layers 35, 36 and 37 and silicon nitride for buffer layers 38, 39 and 40 may also be used in this design structure. have. Rare earth oxides containing light emitting centers may also be used for active layers 35 , 36 and 37 .

The concentration of nanoparticles in any layer can be altered by varying the excess silicon content in the layer during deposition and changing the annealing conditions (eg, annealing temperature and time). It is desirable that the nanoparticle density in the nanoparticle layers 35 , 36 and 37 be as high as possible to increase the intensity of the emitted light, but below the density at which interactions between nanoparticles or agglomeration of nanoparticles occur. .

The total number of repeated layers 35-40 in structure 31 is determined by the voltage to be applied to the entire film and the electric field required for efficient and reliable operation. In a simple approximation, there is a very small voltage drop across the nanoparticle active layers 35 , 36 and 37 , so the number of layers required is divided by the electric field and divided by the thickness of the buffer layers 38 , 39 and 40 . It will be equal to the applied voltage. For example, the applied voltage is 110V, the desired electric field in one dielectric layer 39 is 5MV/cm (ie, 0.5V/nm), whereby the nanoparticle active layer 36 is 2.1 nm thick and the dielectric layer is 4.6 nm If thickness, then the total number of layers 36/39 of the repeating pair is:

(110 V) / (0.5 V/nm) / (4.6 nm) = 48 pairs of layers.

A single color may be emitted by the design film structure by repeating the multilayer structure 20 with the same pair of active layers and dielectric layers, for example the same active layers 12 and 14 . A mixed color, eg white, may be emitted by the design structure 31 as the entire film will contain several pairs of layers for each constituent color. For example, the total N pairs of active/dielectric layer may include k pairs for blue (35/38), m pairs for green (36/39), and n pairs for yellow/red/orange (37/10). There is (k + m + n = N). The number of pairs of each color, for example 35/38, 36/39 and 37/40, can be varied so that any desired color representation index (CRI) can be obtained. For example, warm white requires more pairs of red (35/38) than blue, whereas cold white requires the opposite.

For all white or other multi-colored light emission, and for device 31 in which a back reflector 50 is included in the structure, the lowest energy (longest wavelength, eg, red) emitting layers are applied to the reflector 50 . The closest and highest energy (shortest wavelength, eg blue) layers are preferably located closest to the emitting surface. Layers emitting intermediate wavelengths, eg green, are placed between the layers emitting the longest and shortest wavelengths.

4 shows a design film structure 61 powered by DC power, i.e., an anode 62 and a cathode 63. Most of the active layers 35 , 36 and 37 and the buffer layers 38 , 39 and 40 are identical to those in the film structure 31 ; However, since electrons only semi-move in one direction, the buffer layers interposed between different types of active layers must have the correct thickness to excite the nanoparticles in the nanoparticle active layer closer to the anode 62 . Accordingly, the film structure 61 is preferably terminated by one of the first buffer layers 38 at the cathode 63 and a layer of nanoparticles 37 at the anode 62 . Moreover, since electrons move only in one direction, ie, from the cathode to the anode, one of the second buffer layers 39 is between the first stack 32 and the second stack 33 , and the third buffer layer 40 . is between the second stack 33 and the third stack 34 .

<u>detailed process</u>

Any process technique used to deposit multiple film structures 21 , 31 or 61 can change the film composition on the order of 1 nm scale. Preferred deposition techniques are plasma enhanced chemical vapor deposition (PECVD), preferably ECR-PECVD enhanced by electron cyclone resonance, or ICP-PECVD enhanced by inductively coupled plasma. Optionally, there is metal-organic chemical vapor deposition (MOCVD). Other deposition techniques with the required performance include molecular beam epitaxy (MBE); chemical beam epitaxy (CBE); atomic layer epitaxy (ALE); and pulsed laser deposition (PLD), also called pulsed laser epitaxy (PLE). There are many other thin film growth processes that are variations on the techniques described above. Any of these techniques may also be suitable for deposition of the structural films described above.

In our original homogeneous structure, the size of the nanoparticles is affected by the transient silicon concentration, annealing temperature and time. That is, an increase in any of these increases the nanoparticle size, and possibly is affected by other components of the film, such as carbon. In the case of a design structure with silicon-rich active layers, the size in the direction perpendicular to the plane is limited by the thickness of the silicon-rich layer, and can be approximately equal to it, as long as the excess silicon content is not very low. Annealing also has an effect, but this effect will be reduced if the nanoparticle size is approximately equal to the deposited layer thickness, ie it can only grow parallel to the plane and only very slowly. Impurity content can also have an effect.

All publications, patents, and patent applications cited in this specification are hereby incorporated by reference as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. Citation of any publication is for statements prior to the filing date, and is not to be construed as an admission that the present invention is not entitled to antedate an earlier date by virtue of prior invention.