RU2721717C1 - Device for producing polarized light based on oriented array of nanoplates gase/gaas - Google Patents

Abstract

FIELD: optoelectronics.SUBSTANCE: device for producing polarized light with degree of linear polarization of light ~(50–60):1 includes source (1) of pulsed or permanent unpolarised light, in form of, for example, commercial LED based on III-nitrides of blue-green, blue or ultraviolet range (with emission wavelength in range of 380–550 nm), and external polarizing element (2), made in the form of substrate (3) of GaAs with orientation (001), on which there is a layer with thickness of ~150–250 nm, consisting of nanoplates (4) of two-dimensional GaSe crystal, oriented along selected directions <111> GaAs substrates. External polarizing element (2), in form of an array of oriented nanoplates (4) of GaSe, is spontaneously formed when growing a layer of GaSe by molecular beam epitaxy on substrate (3) of GaAs with orientation (001) in the epitaxial growth temperature range T=500–520 °C.EFFECT: device has small dimensions, mainly determined by dimensions of non-polarized light source, high efficiency of converting unpolarised radiation into polarized (up to 90 % by intensity) and high degree of linear polarization of light (50:1 and more).6 cl, 4 dwg

Description

The present invention relates to the field of optoelectronics, and more particularly to the field of optical integrated circuits using polarized light sources (IPS) of the visible spectral range for encoding and transmitting information.

Polarized light sources are currently required for many commercial applications. In particular, polarized light is used in non-linear frequency converters, where phase matching in a non-linear crystal is usually obtained for only one direction of polarization. Also, IPSs are widely used in liquid crystal displays (LCDs) due to a significant reduction in energy consumption compared to conventional light sources. The use of polarized light is also very effective for indoor lighting, as it can significantly reduce glare from horizontal surfaces, reduce eye strain and reduce radiation power requirements by eliminating unnecessary additional polarizing components. Polarized light is needed to create systems for three-dimensional display of information, can be used in various medical applications, etc. Another important application is the use of IPS in optical integrated circuits. All this huge range of applications determines the relevance of the search for new ideas and the development of designs of effective compact IPS. When creating such sources, it is necessary to take into account both their energy efficiency (including with the aim of solving the urgent problem of reducing CO ₂ emissions into the atmosphere), and the economic component, that is, the cost of their production.

The main requirements for IPS to ensure the possibility of its integration into optical integrated circuits are: small dimensions of the device; a small number of components and the simplicity of their coordination among themselves; low cost of the device; a high degree of linear polarization of light (50: 1 or more); high efficiency of conversion of unpolarized radiation to polarized.

Methods for directly producing polarized light are known in the art. In particular, coherent polarized light is emitted by lasers and laser diodes (LD). The degree of light polarization depends on the LD current, and its minimum value is observed near the laser generation threshold. When the LD is operating in modes near the maximum rated output power, the degree of radiation polarization can reach 100: 1 (~ 0.98) and higher. For example, the radiation polarization degree of commercial blue-green GaN-based laser diodes manufactured by OSRAM according to the specification reaches 100: 1 (https://www.worldstartech.com/products/laser-diodes/green-laser-diode-osram/).

The main disadvantage of IP-based LDs is associated with the use of high pump currents necessary to create a laser generation regime, which, accordingly, requires the use of powerful heat sinks in packaged LDs. It is the latter that lead to an increase in the dimensions of the devices, which prevents the use of such IPSs in optical integrated circuits.

The advantages of using solid-state light emitting diodes (LEDs) as light sources are determined by their small size and high brightness. However, the degree of polarization of LED radiation is much less than in LD. In particular, for commercial standard III-nitride blue and ultraviolet LEDs grown on standard c-sapphire (0001) substrates, the degree of radiation polarization is ~ 0.3-0.4 (see, J. Shakya et al., Polarization of III-nitride blue and ultraviolet light-emitting diodes, Applied Physics Letters 86, 091107, 2005).

A higher degree of intrinsic linearly polarized light emission can be obtained in LED structures grown on non-polar / semi-polar substrates (see YJ Zhao, QM Yan, D. Feezell, K. Fujito, CG Van de Walle, JS Speck, SP DenBaars , S. Nakamura, Optical Polarization Characteristics of Semipolar Opt. Express 21, A53-A59, 2013). However, in the case of GaN LEDs grown on semi-polar substrates, the degree of radiation polarization also does not exceed ~ 0.8 (10: 1).

Radiation with a high degree of polarization can be obtained by integrating nanostructures into the LED design, such as, for example, a metal-based polarization grating (see M. Ma, DS Meyaard, QF Shan, J. Cho, EF Schubert, GB Kim, M. -H. Kim, C. Sone, Polarized Light Emission from GaInN Light-Emitting Diodes Embedded with Subwavelength Aluminum Wire-Grid Polarizers, Appl. Phys. Lett. 101, 061103, 2012). However, the degree of polarization of the radiation is still insufficient for commercial applications.

Due to the unique combination of optical and physical characteristics and properties of anisotropic liquids, luminescent liquid crystals (LC) are promising for the implementation of polarized LEDs based on organic materials (OLED - organic LEDs). A device is known for producing polarized light based on OLED (see patent US7037599, IPC Н05В 33/14, published May 2, 2006), including a device containing a substrate, anode and cathode electrodes, a uniaxial orienting layer located between the anode and cathode electrode, and light emitting a layer located above the uniaxial orienting layer and capable of generating polarized light, the molecules in the light emitting layer must be aligned uniaxially along a certain direction, and a carrier restriction layer located above the light emitting layer, the carrier restriction layer blocking holes, electrons or excitons, and in this case, the recombination of holes from the anode and electrons from the cathode is limited by this layer. An oligomeric fluorene or a mixture of oligomeric fluorenes is used as the light emitting layer.

One of the main disadvantages of OLED-based IPSs is associated with a shorter OSID lifetime compared to IPSs using standard LEDs based on group 3 nitrides as a source of non-polarized light. In addition, the maximum degree of polarization (dichroic ratio) demonstrated in highly efficient polarizing OLEDs using oligofluorene-based luminescent LC materials does not exceed 30: 1, which is still below the requirements for commercial use (more than 40: 1).

A device for producing polarized light is also known (see patent US6710541, IPC G02B 5/30; H01L 51/52, published March 23, 2004), including a non-polarized light source consisting of an organic electroluminescent or photoluminescent device and a polarizer based on a cholesteric liquid crystal. The main disadvantage of the known IPA is also associated with an insufficient degree of linear polarization of light.

The main direction in the development of IPS designs is currently associated with the task of increasing their maximum efficiency. Adding an external polarizer to the IPA design leads to significant light loss (at least 50%), therefore, when developing IPA designs, the main attention is paid to increasing the efficiency of collection and subsequent reuse of part of the light flux with “undesirable” polarization. This, however, leads to a complication of the design of devices and, to a large extent, to an increase in the size and cost of IPS.

A device for producing polarized light is known (see patent RU 2479071, IPC H01L 31/44, published April 10, 2013), including an LED crystal having a first surface, a second surface and at least one side face connecting the first and second surfaces , a light-polarizing layer, a light-blocking layer, a light-reflecting layer, wherein the light-polarizing layer is located on the first surface, the light-blocking layer is located at least on one side face, and the light-reflecting layer is located on the second surface of the LED crystal. It is proposed to use a reflective polarizer on a cholesteric liquid crystal or a polarizer with a wire mesh as a light-polarizing layer. The light-polarizing layer can be formed directly on the LED crystal in the scale of the whole plate, and then sequentially sliced, or, alternatively, made on a carrier substrate, sliced and sequentially attached to the LED crystal using the corresponding bonding contact.

When using a polarizing layer based on a cholesteric liquid crystal, the achieved degree of linear polarization of light does not exceed 30: 1 (~ 0.93) (laboratory values). In the case of using a polarizer with a wire mesh, the main disadvantages are serious technological difficulties in the formation of a polarizing layer and, as a result, the high cost of the final device. In particular, it was shown (see M. Ma et al., Appl. Phys. Lett. 101, 061103, 2012) that when forming a polarizer with an Al wire mesh on the back of the InGaN LED sapphire substrate using a combination of electron beam lithography and inductively coupled plasma reactive ion etching, it is possible to achieve a degree of linear polarization of radiation of more than 0.9. However, the required period of the metal lattice of the order of 100-150 nm turns out to be very close to the resolution of the electron beam lithography method.

A known liquid crystal display projector including a light source, a collimator lens, a planar polarization converter, a liquid crystal display panel and projection lenses (see US patent 5940149, IPC G02B 27/28; G02B 5/04; G02B 5/124; G02B 5 / 30, published on August 17, 1999). The polarization converter in this device is made in the form of a set of films and includes plenary components: a prismatic film reflecting a polarizing film and a quarter-wave moderator film located between the prismatic and polarizing films. The prismatic surface has alternating transmitting and reflecting prismatic faces located at certain basic angles.

Fundamentally, polarization beam splitters can provide polarization contrast up to 3000: 1 for specific wavelengths. However, the disadvantage of this technical solution is the need for a clear coordination of all elements of the polarization converter. Moreover, the well-known plenary light polarization converter is optimized for use in full-color projection systems with a large LCD diagonal. The presence of many components requiring coordination, including a collimator lens, makes it difficult (or even impossible) to use this polarization converter in optical integrated circuits.

The closest set of essential features to this device is a device for producing polarized light (see patent US7037599, IPC Н05В 33/14, published 02.05.2006), adopted as a prototype. The prototype device consists of a source of unpolarized light (LED), optically connected to an external polarizing element, providing a high degree of polarization of light. In the external polarizer, an oligomeric fluorene or a mixture of oligomeric fluorenes is used as a light-emitting layer, which allows one to obtain polarized light at the output with a degree of linear polarization up to 30: 1.

The advantage of the known solution is the compactness and simplicity of the device (essentially, one element). The disadvantages of the known IPS are the insufficient degree of linear polarization of light and a shorter lifetime compared with IPS using standard LEDs based on group III nitrides as a source of unpolarized light.

The objective of this technical solution was to develop a device for producing polarized light, which would satisfy the basic requirements for IPS for integration into optical integrated circuits, namely, it had small dimensions, determined mainly by the size of the source of unpolarized light, and high conversion efficiency of unpolarized radiation to polarized (up to 90 %) and a high degree of linear polarization of light (50: 1 or more).

The problem is solved in that the device for producing polarized light includes a source of non-polarized light, optically connected to an external polarizing element. New in the device is that the external polarizing element is made in the form of a GaAs substrate with the (001) orientation, on which a layer of nanoplates of a two-dimensional GaSe crystal, oriented along the selected directions <111> of the GaAs substrate, is formed.

The unpolarized light source may be pulsed or constant.

The unpolarized light source can be made in the form of a light emitting diode.

The light emitting diode can be made on the basis of III-N compounds with a radiation wavelength in the range of 380-550 nm (i.e., green, blue or ultraviolet range).

A layer of nanoplates of a two-dimensional GaSe crystal oriented along the selected directions <111> of a GaAs substrate can be made with a thickness of 150-250 nm.

The use of an oriented array of GaSe nanoplates to create polarized light is determined by the fundamental selection rules for optical transitions, which allow optical transitions only along the c axis of the GaSe crystal. The band gap of GaSe lies in the visible spectral range (Eg ~ 2 eV, T = 300 K), which leads to the emission of linearly polarized light in the red-orange spectral range (λ ~ 620 nm) with a degree of polarization close to 100% (~ 0 , 96-0.97 or (50-60): 1) and with the conversion efficiency of non-polarized radiation to polarized at a level of up to 90% in intensity.

The advantages of a device for producing polarized light based on an oriented array of GaSe / GaAs nanoplates are a high degree of linear polarization of light ((50-60): 1) in the visible range of the spectrum, compact dimensions of the device, determined mainly by the size of an unpolarized light source ( LEDs), as well as technological compatibility with relatively inexpensive GaAs substrates.

Monochalcogenide compounds, such as GaSe, have an important distinctive feature: they are direct-gap in volume and indirect-gap in the monolayer limit. More precisely, in monolayers, due to the formation of a valence band with lateral maxima (such as the “Mexican hat”), the structure becomes clearly indirect, while the two types of transitions have close energy in the bulk material [D.V. Rybkovskiy, A.V. Osadchy, E.D. Obraztsova, Transition from parabolic to ring-shaped valence band maximum in few-layer GaS, GaSe, and InSe. Phys. Rev. B 90, 235302 (2014)]. The direct gap in the case of multilayer nanoplates dramatically reduces the requirements for the thickness of structures for instrument applications. The fine exciton structure in GaSe is such that the strong resonance of the direct-gap exciton interacts only with light polarized in the plane perpendicular to the monolayers (or tetra-layers, since each monolayer in the case of GaSe consists of four Se-Ga-Ga-Se atomic planes) due to symmetry rules selection. These two circumstances were decisive when choosing GaSe as a material for a compact polarizer.

This technical solution is illustrated in the drawing, where:

in FIG. 1 schematically shows a device for producing polarized light;

in FIG. Figure 2 shows the image obtained by scanning electron microscopy of the cross section of a layer of GaSe nanoplates grown on a GaAs (001) substrate at a temperature of T _S = 500 ° C;

in FIG. Figure 3 schematically shows the appearance of the polarized component of photoluminescence from an array of GaSe nanoplates.

in FIG. Figure 4 shows the spectra of linearly polarized photoluminescence at room temperature of an array of GaSe nanoplates formed on a GaAs (001) substrate, in the region of the fundamental absorption edge (Е || с and Е ⊥ с - parallel and perpendicular to the crystal axis components of the electric field, respectively).

A device for producing polarized light (Fig. 1, Fig. 2) contains a source 1 of pulsed or constant non-polarized light, for example, in the form of a light emitting diode based on compounds III-N of the blue-green, blue or ultraviolet range (with a radiation wavelength in the range 380-550 nm) and a polarizing element 2. The source 1 is optically connected to the polarizing element 2. The polarizing element 2 is made in the form of a GaAs substrate 3 with the (001) orientation, on which GaSe nanoplates 4 are formed, oriented along the selected directions <111> GaAs substrates. The effective thickness of the layer of 4 GaSe nanoplates is 150-250 nm.

The formation of GaSe nanoplates 4 is carried out by their growth by molecular beam epitaxy (MBE) on a GaAs substrate 3 with the (001) orientation. At substrate temperatures 3 T _S = 500-520 ° C, the initial stage of growth proceeds with the formation of chemical bonds between GaSe and substrate 3 of GaAs, i.e. with the formation of transitional submonolayers. In this case, instead of the growth of the planar layer, the formation of an array of GaSe nanoplates 4 is observed with the optical axis c inclined relative to the normal to the surface of the GaAs (001) substrate 3, oriented mainly in the selected directions, i.e. along the <111> directions of the GaAs substrate 3. The operating temperature range Ts ~ 350-400 ° C, at which the formation of an oriented array of GaSe 4 nanoplates is observed, is determined by the following factors: at low substrate temperatures Ts ~ 350-400 ° C, GaSe growth proceeds according to the van der Waals epitaxy mechanism with the formation of relative to the planar layer, while the axis from the GaSe layer is directed normal to the surface of the GaAs substrate 3; at Ts> 520 ° C, GaSe is reevaporated from the surface of GaAs substrate 3 and no growth occurs; in the temperature range Ts ~ 400-500 ° C, an intermediate case is observed - a multidirectional array of GaSe nanoplates 4 is formed, while the axis from GaSe nanoplates 4 is directed at different angles to the surface of the GaAs substrate 3, i.e. there is no distinguished direction of the optical axis from GaSe nanoplates 4. Maintaining a ratio of fluxes of elements of the VI and III groups, close to stoichiometric, minimizes the inclusion of the parasitic phase of Ga ₂ Se ₃ in the grown GaSe layer. It is important to note that the formation of an oriented array of 4 GaSe nanoplates with the choice of the optimal MPE mode and the selected GaAs (001) substrate 3 occurs automatically. In FIG. Figure 2 shows the cross-sectional image obtained by scanning electron microscopy of a layer of GaSe nanoplates 4 grown on a GaAs (001) substrate 3 at a temperature T _S = 500 ° C. The nanoplates 4 are oriented mainly along the <111> directions of the GaAs substrate 3. This leads to the fact that in the image geometry of the cross section shown in FIG. 2 (crystal cleavage occurs along base planes of the type (011) and (0-11)), nanoplates 4 have an inclination relative to the normal to the surface of a GaAs substrate 3 at an angle of ~ 45 ° to the directions [1-10] and [-110] .

The appearance of the polarized photoluminescence component from an array of GaSe nanoplates 4 upon excitation by an unpolarized light source is schematically illustrated in FIG. 3. In the drive geometry shown in FIG. 3, light is incident on an array of inclined nanoplates 4 perpendicular to the GaAs substrate 3 and at an angle of ~ 45 ° to the optical axis from the nanoplates 4 from GaSe. Inside the nanoplate 4, the light contains both components of the electric field E || c and E ⊥ s due to oblique incidence of light relative to the axis c. According to Snell's law, the angle of refraction of the beam relative to c is 15 ° for a GaSe refractive index of n = 2.874. Under these conditions, the ratio of the amplitudes of the components can be estimated as E || s and E ⊥ s = 0.25 / 0.75. However, despite the dominance of the component E компоненты c, the main contribution to the radiation comes from the singlet exciton state polarized along the c axis, which, according to the selection rules, effectively interacts with the component E || with. The interaction of the triplet state with light is possible only with polarization E ⊥ s. It is allowed only due to spin-orbit interaction with the lower valence bands. This interaction is hundreds of times weaker than for E || c, since the energy distance between the upper and subsequent valence bands in GaSe is large, of the order of 500 meV (see E. Mooser, M. Schlüter. The Band-Gap Excitons in Gallium Selenide. Nuovo Cimento 18, 164-208, 1973). In the excitation geometry shown in FIG. 3, reflected light is actively involved in the excitation of polarized luminescence in a dense array of GaSe nanoplates 4. Repeated re-reflection in the array leads to almost complete utilization of the incident light flux. Provided that the luminous flux is completely utilized, the efficiency of converting unpolarized radiation to polarized radiation can be estimated through the internal quantum output

Y = G _r / (G _r and G _nr ),

where Г _r and Г _nr are the rates of radiation and non-radiation attenuation, respectively. For the value of r _r, we can take the radiation velocity of the direct exciton transition at low temperature (T <10 K), when all states are localized, and transport to possible defective centers is practically absent. This value can be determined as Г _r = 1 / τ _r through the characteristic fast decay time of photoluminescence τ _r . Time-resolved photoluminescence studies in GaSe have shown that τ _r ≈200 ps. The GaSe band structure, in addition to singlet – triplet splitting (the triplet is ~ 2 meV lower in energy), is characterized by the fact that there is an indirect band gap, which is located at a lower energy level by ~ 20 meV than the direct band gap. In perfect structures grown by MBE, the presence of lower energy states of triplet nature is the main channel that determines the "non-radiation" attenuation, parasitic with respect to the desired effect. Radiation of indirect exciton with characteristic slow times of 5–6 does not dominate at low temperature, since it is lower in energy. However, at a temperature of (35–50) K, due to the small energy distance between the levels involved in the recombination process, the states are thermalized, and this time decreases. With a further rise in temperature, a single state forms with a characteristic long photoluminescence decay time of ~ 2.8 ns. Considering it as 1 / Г _nr , the maximum quantum yield of polarized radiation can be estimated as Y = 0.93. The external quantum yield depends on the efficiency of radiation output. This parameter is close to unity in the case of the branched-surface system used, consisting of inclined nanoplates 4, which significantly exceeds the planar surface of the substrate base 3 in terms of total area. Thus, we can conclude that the efficiency of converting non-polarized radiation to polarized radiation at room temperature is achievable level ~ 90% in intensity. To increase the efficiency of collecting polarized radiation, you can use microlenses or use an epoxy dome, as in LED designs.

Example. An experimental model of a device for producing polarized light was made. A blue LED with a radiation wavelength of 416 nm was used as a source of pulsed non-polarized light. By the MPE method, GaSe nanoplates were grown on GaAs (001) substrates. The sources of molecular beams were a standard elemental source Ga and a valve source Se with a decomposer (with a decomposition zone temperature T _Se (cr) = 500 ° C). The temperature of the substrate was T _S = 500 ° C. The growth was carried out on the surface of a 200 nm thick GaAs buffer layer grown on a GaAs (001) substrate in a separate MPE chamber. The growth of GaSe was initiated by the simultaneous opening of the Ga and Se dampers on the GaAs (001) surface. The nominal growth rate of GaSe was 1.3 nm / min. The ratio of the fluxes of elements of groups VI and III during growth was Se / Ga (BEP) ~ 22, which corresponds to stoichiometry on the growth surface under these conditions of MPE. The degree of linear polarization of light was ~ 50: 1 (0.96). A strong polarization dependence was confirmed in an experiment with quasi-resonant excitation (hv _exc = 2.33 eV), which fell along the normal to the plane of the substrate. The emitted light was recorded in the opposite direction. The linearly polarized PL spectra of the array of GaSe nanoplates in the region of the fundamental absorption edge are shown in FIG. 4. The spectral position (1.99 eV) of the radiation zone in FIG. 4 corresponds to the position of the direct exciton emission band at T = 300 K. The observed radiation of the free exciton is polarized mainly along E || with. When excited by linearly polarized light, a 90 ° rotation of the array of nanoplates around the OO 'axis led to the excitation and registration of the weak PL of the nanocrystal in the pure polarization E ⊥ s from the same point in the sample. From FIG. 4 it follows that the ratio of the integrated PL intensities in two different polarizations is ~ 50: 1.

Claims (6)

1. A device for producing polarized light, including a source of non-polarized light, optically connected to an external polarizing element, and an external polarizing element, characterized in that the external polarizing element is made in the form of a GaAs substrate with (001) orientation on which a layer of nanoplates is formed two-dimensional GaSe crystal oriented along the selected directions <111> of the GaAs substrate. 2. The device according to p. 1, characterized in that the unpolarized light source is pulsed. 3. The device according to claim 1, characterized in that the source of unpolarized light is made constant. 4. The device according to claim 1, characterized in that the source of unpolarized light is made in the form of a light emitting diode. 5. The device according to p. 4, characterized in that the light emitting diode is made on the basis of III-N compounds with a radiation wavelength in the range of 380-550 nm. 6. The device according to claim 1, characterized in that the layer of nanoplates of a two-dimensional GaSe crystal oriented along the selected directions <111> of the GaAs substrate is made of a thickness of 150-250 nm.

Images