Patent Application
20240222937
This invention is in the field of multilayer structures, including indium phosphide or gallium arsenide structures, the structures containing porous or etched layers therein, which can be used in electronic applications, such as photonic devices.
Semiconductor laser diodes have found many applications in modern society. In the world of laser diodes, vertical cavity surface emitting lasers (VCSELs) are known to be superior to the edge-emitting lasers (EEL) in terms of cost, manufacturability, flexibility, beam quality, and potential integration. So far EEL is commercially available from the wavelength of approximately 400 nm (violet) to about 2,000 nm (near-infrared).
VCSELs are only commercially available from ˜700 nm (red) to 1 μm. Nevertheless, the wavelength range from 1,200 nm to 1,600 nm is an important one since this is traditionally the range used in silica fibers for long-haul single-mode telecommunication. 1,550 nm is also important as a window for atmospheric wireless transmission of signals and energies.
VCSELs emitting at 1,200 to 1,600 nm (or 1.2 to 1.6 μm) are typically prepared epitaxially on indium phosphide (InP) substrates. For at least the past 20 years, the pursuit of the long-wavelength VCSELs has followed one of three approaches, all with the active (light emitting) regions prepared on InP substrates. However, the methods for forming the vertical cavity, and specifically the formation of the n-side reflective mirror, differs dramatically. None of the 3 approaches have achieved mainstream, large-scale production in spite of the compelling market demand. The three representative approaches are as follows:
Notwithstanding the above methods, fabrication of VCSELs on InP substrates remains very challenging. Therefore, to date, technology space using long-wavelength VCSELs for defense and commercial applications, amongst others, remains essentially un-addressed.
Thus, there is a need for novel semiconductor structures, which can be used as mirrors, and can be fabricated via simplified methods, and can be used to fabricate VCSELs of desired wavelengths.
Therefore, it is an object of the invention to provide such structures which address and overcome the issues known to-date in the manufacture of devices, such as VCSELs.
It is yet another object of the invention to provide novel methods for preparing such structures.
It is still a further object of the invention to provide methods of using the structures described, such as for use in VCSELS.
Multilayer structures containing indium phosphide (InP) or gallium arsenide (GaAs) layers which are porous or electropolished within the structure are described herein.
In one instance, a non-limiting exemplary of the multilayer structure contains:
Selective incorporation of a low index material, such as air, into selected regions of the multilayer structures by electrochemically etching has the effect of lowering the refractive index, as compared to the bulk InP or GaAs. Thus, it is possible to tune the refractive index of the porosified regions within the multilayer structure selectively.
Selective incorporation of air by porosification or electropolishing into selected regions of doped layers in a multilayer structure by electrochemical etching can affect electrical properties, as compared to the bulk (non-porous) equivalent InP or GaAs. Thus, it is possible to tune the electrical properties of the porosified regions within the multilayer structure selectively.
Selective incorporation of air by porosification or electropolishing into selected regions of doped layers in a multilayer structure by electrochemical etching can affect thermal properties, as compared to the bulk (non-porous) equivalent InP or GaAs.
In one non-limiting example, a method of forming a multilayer structure, the method includes the steps of:
The multilayer structures can be used in various applications including electronic, photonic, and optoelectronic applications. Applications for such multilayer structures include, more specifically, fiber-based communications, free-space communications, LiDAR, sensing and range-finding, night vision, and chemical sensing, among others. In particular, the multilayer structures can be used to provide high performance VCSELs with excellent optical and electrical performance, as compared to previously reported VCSELs. VCSELs have many advantages compared to more commonly used edge emitting laser diodes (EELDs), such as superior beam quality, compact form factor, low operating power, cost-effective wafer-level testing, higher yield and lower cost in manufacturing. VCSELs, in general, find important applications in various fields including information processing, micro-display, pico-projection, laser headlamps, high-resolution printing, biophotonics, spectroscopic probing, and atomic clocks.
Multilayer structures made of indium phosphide (InP) or gallium arsenide (GaAs) which contain layers which are porous or etched (i.e., electropolished) within the structure are described herein. Methods of manufacturing and using the multilayer structures are also described. For example, the structures can be used as distributed Bragg reflector bottom mirrors for high-performance VCSELs.
“Porosity,” as used herein refers to the volumetric ratio of air present in a porosified medium, such as a III-nitride layer(s), which is expressed as a percentage.
“Electropolishing,” as used herein, refers to an n-doped indium phosphide or gallium arsenide being etched away completely or substantially etched away (where “substantially etched away” refers to etching greater than 95%, 96%, 97%, 98%, or 99%) leaving a void where n-doped material originally existed. The void represents the low-index medium (i.e., air). The air typically has a refractive index of about 1.
“Refractive Index” or “Index of Refraction,” are used interchangeably and refer to the ratio of the velocity of light in a vacuum to its velocity in a specified medium, such as a layer of a III-nitride, according to the formula n=c/v, where c is the speed of light in vacuum and v is the phase velocity of light in the medium.
“Refractive Index Contrast,” as used herein refers to the relative difference in refractive index between two mediums having different indices of refraction and which are in contact and form an interface.
Numerical ranges include ranges of thicknesses, ranges of doping concentrations, ranges of integers, ranges of times, ranges of voltages, ranges of length, ranges of diameters, ranges of concentrations, etc. The ranges disclose individually each possible number that such a range could reasonably encompass, as well as any sub-ranges and combinations of sub-ranges encompassed therein. For example, a layer may have a thickness in the range of about 1 nm to 10 nm, where the range also discloses thicknesses that can be selected independently from about 2, 3, 4, 5, 6, 7, 8, and 9 nm, as well as any range between these numbers (for example, 3 nm to 8 nm), and any possible combination of ranges between these values.
Use of the term “about” is intended to describe values either above or below the stated value, which the term “about” modifies, in a range of approx. +/−10%; in other instances the values may range in value either above or below the stated value in a range of approx. +/−5%. When the term “about” is used before a range of numbers (i.e., about 1-5) or before a series of numbers (i.e., about 1, 2, 3, 4, etc.) it is intended to modify both ends of the range of numbers and/or each of the numbers recited in the entire series, unless specified otherwise.
Multilayer structures made of indium phosphide (InP) or gallium arsenide (GaAs) which contain layers which are porous or etched (i.e., electropolished) within the structure are described in detail below.
In one instance, a non-limiting exemplary of the multilayer structure contains:
In preferred embodiments, the multilayer structures are made of a single type of doped and undoped (or lowly doped) material. For example, a multilayer structure made of all indium phosphide layers or all gallium arsenide layers. However, in less preferred instances, a mixture of different types of materials is possible.
As shown in
In instances where the substrate is present it can be made of indium phosphide, gallium arsenide, sapphire, silicon, or silicon carbide of any suitable thickness. Preferably an undoped or low doped layer is the first layer deposited on the substrate. In most instances, the multilayer structure is formed of only one type of semiconductor material, InP or GaAs. The layers of alternating layers of n-doped and undoped (or low doped) InP or GaAs are homoepitaxially formed and controllably n-doped using art known techniques. In some instances, the alternating layers can be grown, for example, on a suitable substrate (i.e., c-plane of a sapphire substrate, a silicon substrate, or a silicon carbide substrate) by metal organic chemical vapor deposition (MOCVD). The doped and undoped layers are preferably planar layers. The dimensions of layer(s), whether doped or undoped, can be of any size, area, or shape suitable for a specific application. In some instances, the area is in the range of between about 0.1 to 100 cm2, 0.1 to 90 cm2, 0.1 to 80 cm2, 0.1 to 70 cm2, 0.1 to 60 cm2, 0.1 to 50 cm2, 0.1 to 40 cm2, 0.1 to 30 cm2, 0.1 to 20 cm2, 0.1 to 10 cm2, 0.1 to 5 cm2, or 0.1 to 1 cm2. Electrochemical etching requires that InP or GaAs be doped with an n-type dopant. Accordingly, doped layers, as present, are formed during deposition/formation. Exemplary dopants can include, but are not limited to, n-type Ge and Si dopants. Such dopant sources can include, for example, silane (SiH4), germane (GeH4), and isobutylgermane (IBGe). For n-type doped layers of InP or GaAs formed, the n-type doping concentration can be uniform across the entirety of the layer or the doping concentration may form a gradient (i.e., a layer having a graded dopant concentration across an axis of the layer, such width). The doping concentration is considered high at doping concentration levels of at least about 1×1019 cm−3 or higher; or is the range of between about 0.1×1019 cm−3 to 10×1020 cm−3. In some cases, the high doping concentration level may be about 1×1019 cm−3, 2×1019 cm−3, 3×1019 cm−3, 4×1019 cm−3, 5×1019 cm−3, 6×1019 cm−3, 7×1019 cm−3, 8×1019 cm−3, 9×1019 cm−3, or 10×1019 cm−3. The doping concentration is considered to be moderate at doping concentration levels of greater than about 1×1018 cm−3 to less than 1×1020 cm−3, 2×1018 cm−3 to less than 1×1020 cm−3, 3×1018 cm−3 to less than 1×1020 cm−3, 4×1018 cm−3 to less than 1×1020 cm−3, or 5×1018 cm−3 to less than 1×1020 cm−3. In some instances, the moderately doped concentration level is in the range of 1×1019 cm−3 to less than 1×1020 cm−3 or in the range of about 0.5×1019 cm−3 to 10×1019 cm−3. In some cases, the moderate doping concentration level may be about 1×1018 cm−3, 2×1018 cm−3, 3×1018 cm−3, 4×1018 cm−3, 5×1018 cm−3, 6×1018 cm−3, 7×1018 cm−3, 8×1018 cm−3, 9×1018 cm−3, or 10×1018 cm−3. Moderate to high n-type doping is subject to the electrochemical etching process and results in controlled porosification and/or electropolishing of the doped layers, depending on the conditions used during the electrochemical etching process.
As noted above, the multilayer structure contains layers of undoped InP or GaAs, which are not affected (porosified or etched) when the multilayer structure is electrochemically etched. In general, the multilayer structure contains undoped layers of InP or GaAs. In some instances, however, the multilayer structure may contain layers of low doped InP or GaAs where the doping concentration is considered to be low at doping concentration levels of less than about 20×1017 cm−3 or in the range of between about 0.5×1017 cm−3 to 10×1017 cm−3. In some cases, the moderate doping concentration level may be about 1×1017 cm−3, 2×1017 cm−3, 3×1017 cm−3, 4×1017 cm−3, 5×1017 cm−3, 6×1017 cm−3, 7×1017 cm−3, 8×1017 cm−3, 9×1017 cm−3, or 10×1017 cm−3.
The number of alternating layers of n-doped and undoped (or low doped) InP or GaAs forming the multilayer structure prior to electrochemical etching is not particularly limited. In some instances, the alternating layers are formed such that there is an n-doped layer present between every undoped (or low doped) layer of InP or GaAs. In some cases, there may be 3 to 10 alternating layers (formed of a pair of n-doped and undoped (or low doped) layers of InP or GaAs). For example,
The thicknesses of any one of the n-doped or undoped (low doped) layers, prior to electrochemical etching, may each independently range in between about 50 to 500 nm (and subranges therein). In some instances, the total thickness of the multilayer structure, before or after electrochemical etching, may range from between about 600 nm to about 8,000 nm or 600 nm to about 6,000 nm, and sub-ranges within. The dimensions and/or shape of the layers or substrate may be of any suitable shape/dimension required for an application.
Following electrochemical etching, in the multilayer structures the undoped or low doped InP or GaAs layers are typically unaffected (i.e., non-porosified or substantially non-porosified (where “substantially non-porosified” refers to having a degree of porosity of less than 25%, 20, 15%, 10%, 10%, 5%, 4%, 3%, 2%, or 1% in the undoped (or low doped) layers). In some instances of the method, unintentional porosification of undoped (or lowly doped) layers occur, where even lowly n-doped layers can be porosified during EC etching.
Following electrochemical etching, in the multilayer structures the n-doped InP or GaAs layers can be porosified, as compared to prior to electrochemical etching. Porosification can be high where the layer contains at least one portion which has between about 30% and 90%, or greater porosity. In some instances, the porosity is at least about 30%, 40%, 50%, 60%, 70%, 80%, or 90% or greater. Incorporation of a low index material, such as air, into layer (or portion thereof) by porosification has the effect of lowering the refractive index, as compared to the bulk InP or GaAs before porosification.
For the n-doped layers, electrochemical etching can result in different degrees of porosities and pore morphologies by changing the type and concentration of electrolyte, n-doping concentration of the layers, and applied bias voltage, as discussed in greater detail in Section III below.
Electrochemical etching can be used to selectively create lateral or horizontal pores in the multilayer structure. These selectively form from side surfaces of the multilayer structure, as shown in
In a given multilayer structure, all or a portion of the doped InP or GaAs layer may be porosified during electrochemical etching. In some instances, electrochemical etching proceeds from a side wall and the extent of porosification of a layer is at least about 10, 20, 30, 40, 50, 60, 80, or 90% of the longest planar dimension of the doped layers. In some other instances, where electropolishing occurs, the extent of electropolishing of a layer is at least about 10, 20, 30, 40, 50, 60, 80, or 90% of the longest planar dimension of the doped layers. Porosification may occur uniformly or non-uniformly within each doped layer during the electrochemical etching process. Electropolishing may occur uniformly or non-uniformly within each doped layer during the electrochemical etching process.
As noted above, in some instances the doped layer is electropolished away (completely removed) which leaves little or no material between the undoped (or low doped) layers, where the doped InP or GaAs material used to be. The dimensions of the void space formed due to electropolishing depends on the dimensions of the doped InP or GaAs layer and the extent of material that was electropolished away. As shown in
Selective incorporation of a low index material, such as air, into selected regions of the multilayer structures by electrochemically etching has the effect of lowering the refractive index, as compared to the bulk InP or GaAs. Thus, it is possible to tune the refractive index of the porosified regions within the multilayer structure selectively.
Prior to electrochemical etching, in multilayer structures formed of InP each of the layers has an index of refraction of about 3.2. Electrochemical etching, selectively porosifies or may completely electropolish doped InP layers which lower the index of refraction below 3.2. In some instances, the index of refraction of porosified InP layers is about 1.5 to 2.7. When the InP layers are electropolished away the index of refraction is about 1. Consequently, the refractive index contrast (Δn) between the InP layers, after electrochemical etching, may be in the range of about 0.5 to about 2. In some instances, the refractive index contrast (Δn) is at least about 1.1, 1.2, 1.3, 1.4, or 1.5. In still other instances, the refractive index contrast ratio (Δn) is at least about 1.5.
Prior to electrochemical etching, in multilayer structures formed of GaAs each of the layers has an index of refraction of about 3.95. Electrochemical etching, selectively porosifies or may completely electropolish doped InP layers which lower the index of refraction below 3.95. In some instances, the index of refraction of porosified GaAs layers is about 1.5 to 3.4. When the GaAs layers are electropolished away the index of refraction is about 1. Consequently, the refractive index contrast (Δn) between the GaAs layers, after electrochemical etching, may be in the range of about 0.5 to about 2.5. In some instances, the refractive index contrast (Δn) is at least about 1.1, 1.2, 1.3, 1.4, or 1.5. In still other instances, the refractive index contrast ratio (Δn) is at least about 1.5.
As discussed in the Examples below, forming layers of alternating indices of refraction can create successive constructive or destructive interferences. When the thicknesses of each layer corresponds to, respectively, ¼ of the optical wavelength, the stack of alternating layers of the multilayer structure together act as a reflective mirror that can be used to support the long wavelengths needed for infrared emitting VCSELs. In certain instances, the multilayer structure acts as a mirror and demonstrates a reflectance of at least about 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, or 99.9%.
Selective incorporation of air by porosification or electropolishing into selected regions of doped layers in a multilayer structure by electrochemical etching can affect electrical properties, as compared to the bulk (non-porous) equivalent InP or GaAs. For electrically injected devices, especially those requiring high current densities, good electrical transport is essential for high device performance.
In some instances, porosification of regions of the doped InP or GaAs layers or electropolishing of regions of the doped InP or GaAs layers results in a multilayer structure where it is possible to retain a carrier (electron) concentration of above about 5×1018 cm−3 and electrical mobilities of at least about 50, 60, 70, 80, 90, 95 cm2/V s, or greater, compared to the bulk (non-porous) equivalent bulk InP or GaAs.
Selective incorporation of air by porosification or electropolishing into selected regions of doped layers in a multilayer structure by electrochemical etching can affect thermal properties, as compared to the bulk (non-porous) equivalent InP or GaAs.
In some instances, porosification of regions of the doped InP or GaAs layers or electropolishing of regions of the doped InP or GaAs layers results in a multilayer structure where the thermal conductivity of the multilayer structure, as a whole, in the range of between about 1 to 25, 2 to 20, 2 to 15, or 2 to 10 W/m·K. In still some other instances, the average thermal conductivity is at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 W/m·K.
Unlike the photoelectrochemical (PEC) methods previously used, the conductivity selective electrochemical (EC) etching methods rely on electrically injected holes, rather than photogenerated holes, to oxidize doped indium phosphide or gallium arsenide permitting their selective porosification or electropolishing. The methods do not require exposure to ultraviolet (UV) illumination.
In one non-limiting example of a method of forming a multilayer structure, the method includes the steps of:
In preferred embodiments of the method, the multilayer structures are fabricated from a single type of doped and undoped (or lowly doped) material. For example, the multilayer structure is made of all indium phosphide layers or all gallium arsenide layers. However, in less preferred instances of the method, it is possible to use different types of materials (i.e., InP and GaAs together).
For the method described above, the substrate, when present, may be a sapphire, silicon, or silicon carbide substrate, or may be made of a preferably undoped indium phosphide or gallium arsenide layer. The single-crystalline substrate may have any suitable thickness.
The undoped or low doped indium phosphide or gallium arsenide layers, as well as the n-doped indium phosphide or gallium arsenide layers, can each be epitaxially or homoepitaxially grown according to art known methods, such as metal organic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE).
In the methods, undoped or low doped indium phosphide or gallium arsenide layers, as well as the n-doped indium phosphide or gallium arsenide layers, as present, may each independently range between about 50 to 500 nm (and subranges therein). The thicknesses of the substrate or substrate layer, as may be present, may each independently have any suitable size but may range from about 50 to 500 nm (and subranges therein). The dimensions and/or shape of the above layers or substrate may be of any suitable shape/dimension required for an application. Lastly, the total thickness of the multilayer structure formed by the method preferably ranges from between about 600 nm to about 8,000 nm or about 600 nm to about 6,000 nm.
Depositing n-doped indium phosphide or gallium arsenide layers, as present, requires use of a dopant during deposition/formation. Exemplary dopants can include, but are not limited to n-type Ge and Si dopants. Such dopant sources can include, for example, silane (SiH4), germane (GeH4), and isobutylgermane (IBGe). The doping concentration can be uniform across the entirety of a doped III-nitride layer or the doping concentration may form a gradient (i.e., a graded dopant concentration across an axis of the layer, such width). The doping concentration is considered high at doping concentration levels of at least about 1×1019 cm−3 or higher; or is the range of between about 0.1×1019 cm−3 to 10×1020 cm−3. The doping concentration is considered to be moderate at doping concentration levels of greater than about 1×1018 cm−3 to less than 1×1020 cm−3, 2×1018 cm−3 to less than 1×1020 cm−3, 3×1018 cm−3 to less than 1×1020 cm−3, 4×1018 cm−3 to less than 1×1020 cm−3, or 5×1018 cm−3 to less than 1×1020 cm−3. In some instances, the moderately doped concentration level is in the range of 1×1019 cm−3 to less than 1×1020 cm−3 or in the range of about 0.5×1019 cm−3 to ×1019 cm−3. Moderate to high n-type doping is subject to the electrochemical etching process and results in controlled porosification and/or electropolishing of the doped layers, depending on the conditions used during the electrochemical etching process. As noted previously, the multilayer structure contains layers of undoped InP or GaAs, which are not affected (porosified or etched) when the multilayer structure is electrochemically etched. In general, the multilayer structure contains undoped layers of InP or GaAs. In some instances of the method described, the multilayer structure may contain layers of low doped InP or GaAs where the doping concentration is considered to be low at doping concentration levels of less than about 20×1017 cm−3 or in the range of between about 0.5×1017 cm−3 to 10×1017 cm−3.
In step (e) a capping layer is deposited over the entire multilayer structure, where the capping layer may be made of a silicon oxide (i.e., SiO2) or other suitable material, such as silicon nitride (SiNx), hafnium oxide (HfO2), and photoresist material. Suitable photoresist materials are known in the art. The capping layer can have any suitable thickness needed and can be in a range of between 10 to 3000 nm. The capping layer can be epitaxially or homoepitaxially grown according to art known methods, such as metal organic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE), plasma-enhanced chemical vapor deposition (PECVD), atomic layer deposition (ALD), physical vapor deposition (PVD), and sputter.
In step (f) removal of at least a portion of the capping layer to selectively exposes at least one sidewall of the multilayer structure to enable the EC process of step (g) to proceed at the exposed sidewall(s). Suitable techniques for removal of the capping layer, such as silicon dioxide layer, can include, for example, inductively coupled plasma reactive-ion etching (ICP-RIE), which can be used to etch the capping layer in a selective manner to expose the doped layer(s) at the sidewall(s) of the multilayer structure. Alternatively, a capped structure may be physically cleaved in order to expose the doped layer(s) at the sidewall(s) of the multilayer structure.
Porosification and/or electropolishing occurs during the electrochemical (EC) etching process of step (g) and can be controlled based on the concentration of electrolyte, doping concentration, and applied bias voltage (as discussed below). The applied bias voltage is typically a positive voltage in the range of about 0.1 to 10 V, 1.0 to 5V, or 1.0 to 2.5V. In some instances, based on the original doping concentration and the type of etchant used, the applied bias ranges from less than about 1V to at least about 10V, or greater. In some instances, porosity can be selectively minimized when lower relative doping concentration(s) are used, where in one non-limiting instance, a doping concentration of 5×1018 cm−3 in a sample produces a lower porosity as compared to a doping concentration of 2×1019 cm−3 when both are etched under the same conditions. This can be generally expected for all relative concentration differences, where the higher doping concentration will be subject to greater porosification when compared to a lower relative doping concentration, all other electrochemical etching parameters being constant. In some instances, depending on concentration of electrolyte, doping concentration, and selection of applied bias voltage applied voltage(s), electrochemical etching conditions may selectively and controllably result in only porosification (having between about 30% and 90%, or greater porosity introduced) or complete electropolishing (i.e., total or near total removal (i.e., greater than 95%, 96%, 97%, 98%, or 99% removal of doped material). The electric field direction during the EC etching process can be used to control the direction of the etching direction and thereby control the direction of the pores etched into the doped InP or GaAs layer. For example, during step (g) of the methods, the EC etching direction may be a function and determined by the electric field direction. The EC etching preferably produces a lateral etching direction. The rate of lateral etching during step (g) can be about 0.1 μm/min, 0.2 μm/min, 0.3 μm/min, 0.4 μm/min, 0.5 μm/min, 0.6 μm/min, 0.7 μm/min, 0.8 μm/min, 0.9 μm/min, 1 μm/min, 2 μm/min, 3 μm/min, 4 μm/min, 5 μm/min, 6 μm/min, 7 μm/min, 8 μm/min, 9 μm/min, 10 μm/min, 20 μm/min, 30 μm/min, 40 μm/min, or 50 μm/min.
The EC etching of step (g) can be carried out under an applied bias voltage from about 1 min to 24 hours, 1 min to 12 hours, 1 min to 6 hours, 1 min to 4 hours, 1 min to 2 hours, 1 min to 1 hour, or 1 min to 30 minutes. In some instances, the EC etching of step (g) is carried out under an applied bias voltage for at least about 5 min, 10 min, 15 min, 20 min, 25 min, 30 min, 35 min, 40 min, 45 min, 50 min, 55 min, 60 min, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 10, hours, 15 hours, 20 hours, 24 hours, or greater. The EC etching of step (g) can be carried out under an applied bias voltage at room temperature or at a temperature in the range of about 10° C. to about 50° ° C. The EC etching of step (g) can be carried out under an applied bias voltage under ambient conditions or optionally under an inert atmosphere (such as of nitrogen or argon).
The EC etching carried out in step (g) can be carried out in different types and concentrations of a high conductivity electrolyte (either salt or acid). Exemplary high conductivity electrolytes can include, but are not limited to halide ions (fluoride, chloride, bromide, iodide), hydrochloric acid (HCl), sulfuric acid (H2SO4), hydrofluoric acid (HF), KOH, NaOH, Ba(OH)2, Ca(OH)2, Sr(OH)2, NH4OH, NaCl, NaF, nitric acid (HNO3), organic acids and their salts (such as oxalic acid and citric acid), and mixtures thereof. The concentration of the electrolyte in the high conductivity electrolyte solutions, typically aqueous, can be in the range of between about 0.1 to 10M. In some other instances, concentration of the electrolyte in the high conductivity electrolyte solutions, typically aqueous, can be defined as a percentage (volume/volume) of the electrolyte to solvent(s), such as water, in which it is dissolved in and can be in the range of between about 0.1 to 30% by volume. In still other instances, concentration of the electrolyte in the high conductivity electrolyte solutions, typically aqueous, can be defined as a percentage (weight/volume) of the electrolyte to solvent(s), such as water, in which it is dissolved in and can be in the range of between about 0.1 to 30% by weight. The electrolytes listed above normally do not etch InP or GaAs at room temperature, but can etch InP or GaAs under an electrochemical anodic conditions applied during step (g).
It is believed that electrochemical etching, as described above, proceeds from the edge of the exposed sidewall(s), such as in a lateral direction to preferentially form horizontal pores. Lateral etching causes porosification producing pores, typically nanopores, to be formed horizontally or predominantly horizontally within a doped layer during step (g). The multilayer structure has a vertical axis from the lowest to the top-most layer where planar layers of alternating doped and undoped (or lowly doped) InP or GaAs exist. When EC etching is induced, porosification of n-doped layers occurs perpendicular or predominantly perpendicularly to the vertical axis. Predominantly perpendicular, as used herein, refers to pores that, on average, are oriented within about 20, 15, 10, or 5 degrees of the perpendicular/horizontal plane to the vertical axis. In other words, porosification occurs along or predominantly along a horizontal direction that is parallel or near parallel to the planar direction of doped layers. See
Electrochemical etching generally consists of oxide formation and removal steps (Quill, N., et al. (2013). ECS transactions, 58(8), 25-38). It is believed that the presence of free holes at InP or GaAs/electrolyte interface is important for oxidation, and oxides formed can be easily dissolved in the various electrolytes. The free holes are supplied by electric-field assisted tunneling and their amount mainly depends on anodic bias and doping concentration. In some instances, electrochemical (EC) etching conditions result in no EC etching at low anodic bias and/or low doping concentrations (low doping is described above), whereas electropolishing (i.e., complete etching) is observed at large bias and/or high n-doping concentrations. Porosification is observed at intermediate bias and/or doping concentrations.
Indium phosphide can demonstrate complex electrochemical (EC) etching behavior, particularly when high doping concentrations are used in the InP layers.
It is also believed that the free holes created at pore tips are quickly consumed and participate in InP oxidation, which can result in a short hole diffusion length and can prevent EC etching of the walls between nanopores. In the case of heavily doped InP (n≥1×1019 cm−3), a small depletion width leads to ultra-thin pore walls, and nanopores could easily collapse, depending on electrolyte choice due to hole diffusion. The short hole diffusion length and/or passive layers formed on the pore walls are required for porosification of doped InP. The concentration of electrolyte can be increased to satisfy these conditions and can be used to achieve selective porosification of doped InP (see
The multilayer structures can be used in various applications including electronic, photonic, and optoelectronic applications. Applications for such multilayer structures include, more specifically, fiber-based communications, free-space communications, LiDAR, sensing and range-finding, night vision, and chemical sensing, among others.
In particular, the multilayer structures are useful in laser diodes, such as vertical-cavity surface-emitting lasers (VCSELs) where they can serve as distributed Bragg reflectors (DBRs). Multilayer structures used as a DBR made of indium phosphide should be able to provide long-wavelength VCSELs (i.e., emitting at infrared wavelengths from 900 to 2000 nm). Multilayer structures used as a DBR made of gallium arsenide should be able to provide VCSELs which can emit in the range of between about 800 to 1100 nm. The multilayer structures can be incorporated into different devices, such as VCSELs, using art known techniques.
The multilayer structures fabricated and etched in these examples possess a high index contrast, are lattice-matched, epitaxially compatible, and allow for manufacture-friendly fabrication of mirrors, DBRs, for VCSELs. When multilayer structures are formed of InP they can be used to produce long-wavelength VCSELs. The methods described above can allow for the mass production of DBRs based on the multilayer structures, such as made of indium phosphide, to produce infrared VCSELs emitting at 1200 to 2000 nm.
These multilayer structures, when EC etched, can demonstrate primarily or exclusively horizontal porosification in a layered, selective way. This is especially important in VCSEL applications, as horizontal porosification can be made without adversely affecting the quality of the active region that is typically grown on top of the DBR layers.
In certain instances, the multilayer structure acts as a DBR in a vertical cavity surface emitting laser (VCSEL) demonstrating a stopband at or around 1100 to 2000 nm for indium phosphide with a peak reflectance of at least about 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, or 99.9%. In certain instances, the multilayer structure acts as a DBR in a vertical cavity surface emitting laser (VCSEL) demonstrating a stopband at or around 1250 nm with a peak reflectance of at least about 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, or 99.9%. In some instances, the stopband has a wavelength width of between about 800 to 1100 nm for gallium arsenide with a peak reflectance of at least about 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, or 99.9%. Such properties can be tailored by tuning the properties of the multilayer structure, such as the number of layers, the thicknesses of the layers, the extent of porosification or electropolishing of doped layers by EC etching.
The multilayer structures can be used to provide high performance VCSELs with excellent optical and electrical performance, as compared to previously reported VCSELs. VCSELs have many advantages compared to more commonly used edge emitting laser diodes (EELDs), such as superior beam quality, compact form factor, low operating power, cost-effective wafer-level testing, higher yield and lower cost in manufacturing. VCSELs, in general, find important applications in various fields including information processing, micro-display, pico-projection, laser headlamps, high-resolution printing, biophotonics, spectroscopic probing, and atomic clocks.
The present invention will be further understood by reference to the following non-limiting examples.
Multilayer structures were formed from indium phosphide was homoepitaxially grown on single side polished n-InP substrate by MOCVD. Doped layers were doped with germanium dopant at the doping concentrations listed below. The multilayer structure contained alternating layers of n-doped (2×1019 cm−3) and undoped InP layers. A base layer was formed of an n-doped InP layer with a doping concentration of only 1×1017 to 1×1018 onto which a 300 nm layer of undoped InP was formed. A 140 nm n-doped InP layer was deposited onto the 300 nm layer. Then, a 110 nm layer of undoped InP was deposited onto the 140 nm layer. Additional layers of doped InP (140 nm; 2×1019 cm−3) and undoped InP (110 nm) were deposited. The final layer of InP was undoped. The structure is represented in
Subsequently, the multilayer structures were electrochemically (EC) etched in various concentrations of aqueous electrolytes (HCl, oxalic acid, and KOH) and at different anodic voltages, according to the methods described above. The anodic voltages were applied for 5 to 20 mins. For testing performed, HCl demonstrated a fast lateral etch rate (about 8-20 μm/min), whereas KOH showed a slower lateral etch rate (about 0.1-0.5 μm/min). The conditions tested and observed results are provided in Table 1 below.
Scanning electron microscopy images of the above are shown in
The results of electrochemically etching revealed that lateral etching produced horizontal and/or substantially horizontal pores, when porosification occurred. Layers that were porosified were found to be nanoporous after etching.
Dependent on the EC conditions used, including the electrolyte choice and concentration and bias voltage, the doped InP layers could be electropolished. When using an HCl electrolyte, for example, doped InP was electropolished at or below 2 molar concentration (M) of HCl, whereas at 2 and 3.3 M HCl nanopores were formed even at relatively higher anodic voltages (see
After the porosification of a multilayer structure of Example 1 at 1.5 V in 3.3M HCl, the etched structure's silicon dioxide layer was removed using a buffered oxide etch and then the reflectance of the nanoporosified InP multilayer structure was measured by a commercial Filmetrics F40 EXR capable of spot-measurement of thickness and optical properties of thin films within the spectral range of 400 and 1700 nm.
The refractive indices of nanoporous InP layers depends on their porosities and they could be as small as 1 if the porosity is 100% (i.e., the doped InP is electropolished entirely). These layers of alternating indices of refraction (between the high index of the undoped InP layer and lower index porosified or electropolished layers) can create successive constructive or destructive interferences. When the thicknesses of each layer corresponds to, respectively, ¼ of the optical wavelength, these stacks of alternating layers can act as a reflective mirror which could be used to achieve long wavelength emitting VCSELs.
Multilayer structures were formed from indium phosphide was homoepitaxially grown on single side polished n-InP substrate by MOCVD. Doped layers were doped with germanium dopant at the doping concentrations listed below. The multilayer structure contained alternating layers of n-doped (5×1018 cm−3) and undoped InP layers. A base layer was formed of an n-doped InP layer with a doping concentration of only 1×1017 to 1×1018 onto which a 300 nm layer of undoped InP was formed. A 140 nm n-doped InP layer was deposited onto the 300 nm layer. Then, a 110 nm layer of undoped InP was deposited onto the 140 nm layer. Additional layers of doped InP (140 nm; 5×1018 cm−3) and undoped InP (110 nm) were deposited. The final layer of InP was undoped. The structure is represented in
Subsequently, the multilayer structures were electrochemically (EC) etched in various concentrations of aqueous electrolytes (HBr, H2SO4, KOH, and HCl) and at different anodic voltages, according to the methods described above. The anodic voltages were applied for 5 min to 1 hour. The conditions tested and observed results are provided in Table 2 below.
As shown in the observation in Table 2, doped layers that were porosified were found to be nanoporous after electrochemical etching. It was also found, that dependent on the EC conditions used, including the electrolyte choice, concentration thereof, and bias voltage applied, the doped InP layers could be either porosified or etched (electropolished) away.
The nanoporosified InP multilayer structure of Example 3 with a 5×1018 cm−3 doping concentration was porosified in 5% HCl electrolyte, at a 1.8V bias applied for 4 mins with Pt counterelectrode. Reflectance of the structure was measured by a commercial Filmetrics F40 EXR capable of spot-measurement of thickness and optical properties of thin films within the spectral range of 900 and 1800 nm. Additionally, the reflectance of a nanoporosified InP multilayer structure of Example 1, having a higher porosity, with a 2×1019 cm−3 doping concentration (porosified at 1.5 V for 15 minutes in 3.3M HCl) was also measured.
A vertical cavity using a nanoporous InP DBR structure, prepared according to Example 3 (8% HCl electrolyte, 1.8V bias applied for 4.5 mins with Pt counterelectrode), having twelve pairs ¼ λ layers as the bottom mirror on an undoped InP substrate was constructed. The vertical cavity portion of structure included a bottom layer of n-InP (about 930 nm), a layer of InAlGaAs multiple quantum wells (MQW) (81 nm), a layer of p-InP (71 nm), and a top layer InGaAs tunnel junction layer (20 nm). The complete vertical cavity structure including cavity and bottom DBR mirror is shown in
Reflectance measurements on the vertical cavity were performed using a Bruker Vertex 70+Hyperion 2000 within the spectral range of 900 and 1900 nm.
Lastly, an optical field simulation (%=1661 nm) was also performed using MATLAB software.
As shown in
A vertical cavity using a nanoporous InP DBR structure, prepared according to Example 3 (8% HCl electrolyte, 1.8V bias applied for 4.5 mins with Pt counterelectrode), having twelve pairs ¼ λ layers as the bottom mirror on an undoped InP substrate was constructed. The vertical cavity portion of structure included a bottom layer of n-InP (about 930 nm), a layer of InAlGaAs MQW (81 nm), a layer of p-InP (71 nm), a layer InGaAs tunnel junction layer (20 nm), and an a-Si-spacer (100 nm). Finally, a top SiO2 (252 nm)/a-Si (107 nm) DBR top mirror formed part of the structure. The complete vertical cavity structure including the cavity and top and bottom DBR mirrors described is shown in
Reflectance measurements on the vertical cavity were performed using a Bruker Vertex 70+Hyperion 2000 within the spectral range of 900 and 1900 nm.
Lastly, an optical field simulation (λ=1500 nm) was also performed using MATLAB software.
It is noted that the additional amorphous silicon (a-Si) layer was used in order to move the cavity mode to 1500 nm (see
Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of skill in the art to which the disclosed invention belongs. Publications cited herein and the materials for which they are cited are specifically incorporated by reference.
Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention. Such equivalents are intended to be encompassed by the following claims.
This application claims priority to and benefit of U.S. Provisional Application No. 63/183,337, filed May 3, 2021, which is hereby incorporated herein by reference in its entirety.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2022/027391 | 5/3/2022 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63183337 | May 2021 | US |
