Patent Application
20210210646
The disclosure relates to broadband ultraviolet (UV) to shortwave infrared (SWIR) optoelectronic devices operating within the wavelength range of 0.2 μm to 1.8 μm including photodetectors and photodetector arrays, and sensors and systems employing the same.
Broadband photodetectors, detector arrays, sensors and systems operating in the wavelength range between about 0.2 μm and 1.8 μm range have a wide range of applications, including fiber optic communications, and sensing and imaging including for military, biomedical, agricultural, industrial, environmental and scientific applications. The devices may be used for spectral analysis of a variety of materials, including food, pharmaceuticals and chemicals.
To cover such a broad wavelength range, devices made using different group-IV and compound III-V semiconductor devices must be integrated together, since detectors made from different materials only operate efficiently over narrower wavelength ranges than the desired range in many sensing applications. For example, for wavelengths in the near infrared (NIR) to SWIR, between wavelengths of about 0.9 μm and 1.8 μm, indium gallium arsenide (InGaAs) materials are usually grown on indium phosphide (InP) substrates. The composition and thickness of the InGaAs layers are chosen to provide the required functionality, such as light emission or absorption at desired wavelengths of light and are also lattice-matched or very closely lattice-matched to the InP substrate, in order to produce high quality materials that have low levels of crystalline defects, and high levels of performance. Visible and NIR wavelengths (from about 0.35 μm up to about 1.1 μm) may be detected by silicon devices on Si substrates, or by GaAs-based detectors on GaAs substrates. Gallium nitride (GaN) based devices may be used to detect UV and visible wavelengths from about 0.2 μm to 0.45 μm. However, each of these semiconductor materials has a different lattice constant, preventing monolithic integration of the materials without undertaking difficult and complex growth and processing steps. Typically, multiple different sensors or imagers must be used to provide broad spectral coverage for practical systems. Silicon detectors may be produced that can absorb UV light, either through substrate thinning, or by surface treatment using a fluorescent or phosphorescent layer that can absorb light between about 0.2 and 0.35 μm. However, the maximum absorption wavelength is 1.1 μm. Some attempts to reduce the minimum wavelength absorption for InGaAs detectors on InP substrates have been made. Detectors with absorption at wavelengths as short as 0.5 μm have been made, but the spectral responsivity in the visible range is low, and the device processing is complex, requiring careful substrate thinning Although InGaAs on InP materials currently dominates the short wavelength infrared (SWIR) photodetector market, the material system has several limitations, including the high cost of InP substrates, low yields due to fragility of the InP substrates, and limited InP wafer diameter (and associated quality issues at larger diameters). From a manufacturing perspective and an economic perspective, gallium arsenide (GaAs) represents a better substrate choice. However, the large lattice mismatch between GaAs and the InGaAs alloys required for infrared devices produces poor quality materials that compromise electrical and optical performance. Attempts have been made to produce long-wavelength (greater than 1.2 μm) materials for photodetectors on GaAs based on dilute nitride materials such as GaInNAs and GaInNAsSb. However, where device performance is reported, it has been much poorer than for InGaAs/InP devices. For example, the dilute nitride-based devices have very low spectral responsivity, which make the devices unsuited for practical sensing and photodetection applications. Furthermore, although GaAs can absorb visible wavelengths of light, when designing SWIR detectors using dilute nitride materials, absorption at wavelengths outside of the dilute nitride layer causes the short wavelength absorption of the detectors to be limited to about 0.9 μm. Other considerations for photodetectors include dark current and specific responsivity.
For example, Cheah et al., “GaAs-Based Heterojunction p-i-n Photodetectors Using Pentenary InGaAsNSb as the Intrinsic Layer”, IEEE Photon. Technol. Letts., 17(9), pp. 1932-1934 (2005), and Loke et al., “Improvement of GaInNAs p-i-n photodetector responsivity by antimony incorporation”, J. Appl. Phys. 101, 033122 (2007) report photodetectors having a responsivity of only 0.097 A/W at a wavelength of 1300 nm.
In U.S. Application Publication No. 2016/0372624, Yanka et al. disclose optoelectronic detectors having dilute nitride layers (InGaNAsSb). Although certain parameters that relate to semiconductor material quality are described, no working detectors having practical efficiencies are taught within the broad compositional range disclosed.
To take advantage of the manufacturing scalability and cost advantages of GaAs substrates, there is continued interest in developing long-wavelength materials on GaAs that have improved optoelectronic performance. There is also interest in developing devices based on these materials that are capable of operating at visible and UV wavelengths, so that one device may be able to provide a broad wavelength range of operation that is usually covered by two or more separate devices based on different material systems.
According to the present invention, semiconductor optoelectronic devices comprise: a substrate; a first doped III-V layer overlying the substrate; an active region overlying the first doped III-V region, wherein, the active region comprises a lattice matched dilute nitride layer or a pseudomorphic dilute nitride layer; the dilute nitride layer has a bandgap within a range from 0.7 eV and 1.0 eV; and the dilute nitride layer has a minority carrier lifetime of 1 ns or greater, wherein the minority carrier lifetime is determined using time-resolved photoluminescence at 25° C.; a second doped III-V layer overlying the active region; and a luminescent layer overlying the second doped III-V layer, wherein the semiconductor optoelectronic device is configured to have a spectral responsivity within a range from 0.2 μm 1.8 μm.
According to the present invention, photodetector arrays comprise a plurality of the semiconductor optoelectronic devices of any one of claims 1 to 10.
According to the present invention, sensors comprise at least one semiconductor optoelectronic device of any one of claims 1 to 10, and at least one optical filter overlying the at least one semiconductor optoelectronic device.
According to the present invention, methods of forming semiconductor optoelectronic devices comprise: forming a substrate; forming a first doped III-V layer overlying the substrate; forming an active region overlying the first doped III-V layer, wherein, the active region comprises a lattice matched dilute nitride layer or a pseudomorphic dilute nitride layer; the dilute nitride layer has a bandgap within a range from 0.7 eV and 1.0 eV; and the dilute nitride layer has a minority carrier lifetime of 1 ns or greater, wherein the minority carrier lifetime using time-resolved photoluminescence at 25° C.; forming a second doped III-V layer overlying the active region; and forming a luminescent layer overlying the second doped III-V layer.
According to the present invention, methods of forming semiconductor optoelectronic devices comprise: forming a substrate; forming an etch-stop/release layer overlying the substrate; forming a first doped III-V layer overlying the etch-stop/release layer; forming an active region overlying the first doped III-V layer, wherein, the active region comprises a lattice matched dilute nitride layer or pseudomorphic dilute nitride layer; the dilute nitride layer has a bandgap within a range from 0.7 eV and 1.0 eV; and the dilute nitride layer has a minority carrier lifetime of 1 ns or greater; forming a second doped III-V layer overlying the active region; removing the substrate and the etch-stop/release layer; and forming a luminescent layer underlying the first doped III-V layer.
According to the present invention, semiconductor optoelectronic devices are made according to the present invention.
The drawings described herein are for illustration purposes only. The drawings are not intended to limit the scope of the present disclosure.
The following detailed description refers to the accompanying drawings that show, by way of illustration, specific details and embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. Other embodiments may be utilized, and structural, logical, and electrical changes may be made without departing from the scope of the invention. The various embodiments disclosed herein are not necessarily mutually exclusive, as some disclosed embodiments may be combined with one or more other disclosed embodiments to form new embodiments. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the embodiments of the present invention is defined only by the appended claims, along with the full scope of equivalents to which such claims are entitled.
The term “lattice matched” as used herein means that the two referenced materials have the same lattice constant or a lattice constant differing by less than +/−0.2%. For example, GaAs and AlAs are lattice matched, having lattice constants differing by 0.12%.
The term “pseudomorphically strained” as used herein means that layers made of different materials with a lattice constant difference up to +/−2% can be grown on top of a lattice matched or strained layer without generating misfit dislocations. The lattice parameters can differ, for example, by up to +/−1%, by up to +/−0.5%, or by up to +/−0.2%.
The term “layer” as used herein, means a continuous region of a material (e.g., a semiconductor alloy) that can be uniformly or non-uniformly doped and that can have a uniform or a non-uniform composition across the region.
“Region” refers to one or more semiconductor layers. The region is identified based on the function of the region in the semiconductor device.
The term “bandgap” as used herein is the energy difference between the conduction and valence bands of a material.
The term responsivity of a material as used herein refers to the ratio of the generated photocurrent to the incident power of radiation.
The term “spectral sensitivity” or “spectral responsivity” as used herein refers to the relative efficiency of detection, of a light, signal as a function of the frequency or wavelength of the light signal.
“Active region” refers to a layer (or layers) within a device capable of processing light over a desired wavelength range. Processing is defined to be a light emission, a light receiving, a light sensing and light modulation. For example, light absorbed by an active region produces photogenerated carriers (electrons and holes).
“Overlying” is used to refer to the position of a semiconductor layer with respect to another semiconductor. A first semiconductor layer that overlies a second semiconductor layer can be adjacent and in contact with the second semiconductor layer or there can be one or more semiconductor layers between the first semiconductor layer and the semiconductor layer.
“Adjacent” refers to the position of a first semiconductor layer with respect to a second semiconductor layer such that the first and second semiconductor layers are in physical contact.
Substrate 102 can have a lattice constant that matches or nearly matches the lattice constant of the substrates such as GaAs or Ge. The lattice constants of GaAs and Ge are 5.65 Å and 5.66 Å, respectively, and growth of III-V materials with similar compositions without defects can be grown on either substrate. The close matching of the lattice constants of Ge and GaAs allows, for example, high-quality GaAs to be epitaxially grown on a Ge surface. In some embodiments, the substrate can be GaAs. Substrate 102 may be doped p-type, or n-type, or may be a semi-insulating (SI) substrate. The thickness of substrate 102 can be chosen to be any suitable thickness, such as between about 150 μm and 750 μm. Substrate 102 can include one or more layers, for example, the substrate can include a buffered substrate, such as a buffered Si substrate that is engineered to have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. A material such as a substrate having a lattice constant that nearly matches the lattice constant of GaAs or Ge means that the material such as the substrate has a lattice constant different than that of GaAs or Ge by less than or equal to 3%, less than 1%, or less than 0.5% of the lattice constant of GaAs or Ge. Examples of buffered silicon substrates that can provide a lattice constant approximately equal to that of GaAs or Ge include SiGe buffered Si, SiGeSn buffered Si, and rare-earth (RE) buffered Si, such as a rare-earth oxide (REO) buffered Si. A layer such as SiGe, SiGeSn, or a RE-containing layer can form a buffer layer (or lattice engineered layer) grown on a substrate such as Si having a low number of defects and/or dislocations in the buffer layer. The buffer layer can provide a lattice constant at the top of the buffer layer approximately equal to that of a GaAs or Ge substrate, facilitating the ability to form high quality III-V layers on top of the buffer layer, with a low number of defects and/or dislocations in the overlying III-V semiconductor layers and/or dilute nitride layers. A low number of defects can include comparable or fewer defects than would occur in an In0.53Ga0.47As layer grown on an InP substrate.
First doped region 104 can have a doping of one type and the second doped region 108 can have a doping of the opposite type. If first doped region 104 is doped n-type, second doped region 108 is doped p-type. Conversely, if first doped region 104 is doped p-type, second doped region 108 is doped n-type. Examples of p-type dopants include C and Be. Examples of n-type dopants include Si and Te. Doped region 104 and 108 cab be chosen to have a composition that is lattice matched or pseudomorphically strained with respect to the substrate. The doped region can comprise any suitable III-V material, such as GaAs, AlGaAs, GalnAs, (Al)GaInP, AlInP, (Al)GaInPAs, GaInNAs, or GaInNAsSb. The bandgap of the doped region can be selected to be larger than the bandgap of active region 106. In some embodiments, the bandgap of the doped regions, or at least a portion of the doped regions can be selected to be larger than the bandgap of GaAs such that optical absorption by the doped regions in the visible wavelength range is reduced. Doping levels can be within a range, for example, from 1×1015 cm−3 to 2×1019 cm−3.
Doping levels can be constant within a doped region, and/or the doping profile may be graded, for example, the doping level can increase from a minimum value to a maximum value as a function of the distance from the interface between the first doped region 104 and the active region 106. Doped regions 104 and 108 can have a thickness within a range, for example, from 50 nm to 3 μm, from 100 nm to 2 μm, or from 200 nm to 1 μm.
Active region 106 can include an active layer. Active region comprises at least one layer capable of processing light over a desired wavelength range. Processing is defined to be a light emission, a light receiving, a light sensing and light modulation.
The active layer can be lattice matched or pseudomorphically strained with respect to the substrate and/or to the doped regions. The bandgap of the active layer can be lower than that of the doped regions 104 and 108.
The active layer can include a dilute nitride material. A dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where x, y and z can be 0≤x≤0.4, 0<y≤0.07 and 0<z≤0.04, respectively. X, y and z can be 0.01≤x≤0.4, 0.02≤y≤0.07 and 0.001≤z≤0.04, respectively. In other embodiments, dilute nitride materials can have compositions as disclosed in U.S. Pat. No. 8,962,993, where x, y and z can be 0≤x≤0.24, 0.02≤y≤0.05 and 0.001≤z≤0.2, respectively. A dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where, for example, 0.12≤x≤0.24, 0.03≤y≤0.07 and 0.005≤z≤0.04; 0.13≤x≤0.2, 0.03≤y≤0.045 and 0.001≤z≤0.02; 0.13≤x≤0.18, 0.03≤y≤0.04 and 0.001≤z≤0.02; or 0.18≤x≤0.24, 0.04≤y≤0.07 and 0.01≤z≤0.024.
The active layer can have a bandgap within a range from 0.7 eV and 1.0 eV such that the active layer can absorb light at wavelengths up to about 1.8 μm such as, for example from 0.2 μm to 1.24 μm, or from 0.2 μm to 1.8 μm. Bismuth (Bi) may be added as a surfactant during growth of the dilute nitride material, improving material quality (such as defect density), and the device performance. The thickness of the active layer can be within a range, for example, from 0.2 μm to 10 μm. The thickness of the active layer can be within a range, for example, from 0.5 μm to 5 μm. The thickness of the active layer can be within a range, for example, from 1 μm to 4 μm, from 1 μm to 3 μm, or from 1 μm to 2 μm. The active layer can be compressively strained with respect to the substrate 102. Strain can improve device performance. For a photodetector, the parameters most relevant to device performance include the dark current, operating speed, noise, and responsivity.
Active region 106 is shown as a single layer, but it will be understood that active region 106 can include more than one dilute nitride layer, with at least two bandgaps between 0.7 eV and 1.0 eV. Examples of multi-bandgap and graded bandgap active layers are described in U.S. Application No. 62/816,718, filed on Mar. 11, 2019, which is incorporated by reference in its entirety. In some examples, active region 106 can include layers having different doping profiles. Examples of doping profiles for dilute nitride optical absorber materials are described in U.S. Application Publication No. 2016/0118526, which is incorporated by reference in its entirety.
Active region 106 and doped regions 104 and 108 form a p-i-n or an n-i-p junction. This junction provides the basic structure for operation of a device such as a photodetector or a light-emitting diode. For photodetectors, p-i-n epitaxial structures can have low background doping in the intrinsic region (active layer) of the devices which are typically operated at 0 V or at very low bias. Therefore, the active region 106 may not be deliberately doped. The active region can comprise an intrinsic layer or an unintentionally doped layer. Unintentionally doped semiconductors do not have dopants intentionally added but can include a non-zero concentration of impurities that act as dopants. The background carrier concentration of an intrinsic or unintentionally doped active layer, which is equivalent to the background dopant concentration, can be, for example, less than 1×1016 cm−3 (measured at room temperature, 25° C.), less than 5×1015 cm−3, or less than 1×1015 cm−3. The minority carrier lifetime (measured at 25° C.) within the active layer can be, for example, greater than 1 ns, greater than 1.5 ns, or greater 2 ns. The minority carrier lifetime can be affected by defects within the semiconductor that contribute to the background carrier concentration, as well as other defect types that can act as recombination centers but do not contribute carriers.
Substrate 202 can have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. The substrate can be GaAs. Substrate 202 may be doped p-type, or n-type, or may be a semi-insulating (SI) substrate. The thickness of substrate 202 can be chosen to be any suitable thickness such as, for example, between about 150 μm and 750 μm. Substrate 202 can include one or more layers, for example, a Si layer having an overlying SiGeSn buffer layer that is engineered to have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. This can mean the substrate has a lattice constant different than that of GaAs or Ge by less than or equal to 3%, less than 1%, or less than 0.5% that of GaAs or Ge.
First doped region 204 can have a doping of one type and the second doped region 210 can have a doping of the opposite type. If first doped region 204 is doped n-type, second doped region 210 is doped p-type. Conversely, if first doped region 204 is doped p-type, second doped region 210 is doped n-type. Examples of p-type dopants include C and Be. Examples of n-type dopants include Si and Te. Doped regions 204 and 210 can be chosen to have a composition that is lattice matched or pseudomorphically strained to the substrate. The doped regions can comprise any suitable III-V material, such as GaAs, AlGaAs, GalnAs, AlInP, (Al)GaInP, (Al) GaInPAs, GaInNAs, and GaInNAsSb. The bandgap of the doped regions can be selected to be larger than the bandgap of active region 208. In some embodiments, the bandgap of the doped regions, or at least a portion of the doped regions can be selected to be larger than the bandgap of GaAs such that optical absorption by the regions in the visible wavelength range is reduced. Doping levels can be within a range, for example, from 1×1015 cm−3 to 2×1019 cm−3. Doping levels may be constant within a region and/or the doping profile may be graded, for example, the doping level can increase from a minimum value to a maximum value as a function of the distance from the interface between the second doped region 210 and the active region 208. Doped layers 204 and 210 can have a thickness, for example, within a range from 50 nm and 3 μm, from 100 nm to 2 μm, or from 200 nm to 1 μm.
Active region 208 can be lattice matched or pseudomorphically strained to the substrate and/or to the doped regions. The bandgap of active region 208 can be lower than that of the doped regions 204 and 210. Active region 208 can comprise a layer capable of processing light over a desired wavelength range. Processing is defined to be a light emission, a light receiving, a light sensing and light modulation.
Active region 208 can include at least one active layer.
An active layer can include a dilute nitride material. The dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where x, y and z can be 0≤x≤0.4, 0<y≤0.07 and 0<z≤0.04, respectively. X, y and z can be 0.01≤x≤0.4, 0.02≤y≤0.07 and 0.001≤z≤0.04, respectively. In other embodiments, dilute nitride materials can have compositions as disclosed in U.S. Pat. No. 8,962,993, where x, y and z can be 0≤x≤0.24, 0.02≤y≤0.05 and 0.001≤z≤0.2, respectively. A dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where, for example, 0.12≤x≤0.24, 0.03≤y≤0.07 and 0.005≤z≤0.04; 0.13≤x≤0.20, 0.03≤y≤0.045 and 0.001≤z≤0.02; 0.13≤x≤0.18, 0.03≤y≤0.04 and 0.001≤z≤0.02; or 0.18≤x≤0.24, 0.04≤y≤0.07 and 0.01≤z≤0.04. An active layer can have a bandgap within a range from 0.7 eV to 1.0 eV such that the active layer can absorb light at wavelengths up to 1.8 μm. Bismuth (Bi) may be added as a surfactant during growth of the dilute nitride material, improving material quality (such as defect density), and the device performance. The thickness of an active layer can be within a range, for example, from 0.2 μm to 10 μm, from 0.5 μm to 5 μm, or from 1 μm to 4 μm. An active layer can be compressively strained with respect to the substrate 202. Strain can also improve device performance. For a photodetector, the device performance of most relevance includes the dark current, operating speed, noise and responsivity.
Active region 208 is shown as a single layer, but it will be understood that active region 208 can include more than one active layer such as one or more dilute nitride layers, with at least two bandgaps between 0.7 eV and 1.0 eV. Examples of multi-bandgap and graded bandgap active layers are described in U.S. Application No. 62/816,718. In some examples, active layer 208 can include active layers with different doping profiles. Examples of doping profiles for dilute nitride optical absorber materials are described in U.S. Application Publication No. 2016/0118526.
The multiplication region 206 can be a p-type III-V region configured to amplify the current generated by the active region 208 through avalanche multiplication. Thus, for each free carrier (electron or hole) generated by the active region 208, the multiplication region 206 generates one or more carriers via the avalanche effect. Thus, the multiplication region 206 increases the total current generated by the semiconductor device 200. Multiplication region 206 can comprise a III-V material, such as GaAs or AlGaAs. In some embodiments, multiplication region 206 can include a dilute nitride layer such as GaInNAs, GaInNAsSb or GaNAsSb. Examples of semiconductor materials and structures for multiplication region 206 are described in co-pending PCT International Application No. PCT/US2019/036857 filed on Jul. 18, 2018, which is incorporated by reference in its entirety.
Substrate 302 can have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. The lattice constants of GaAs and Ge are 5.65 Å and 5.66 Å, respectively, and growth of III-V materials with similar compositions without defects can be grown on either substrate. The close matching of the lattice constants of Ge and GaAs allows, for example, high-quality GaAs to be epitaxially grown on a Ge surface. In some embodiments, the substrate can be GaAs. Substrate 302 may be doped p-type, or n-type, or may be a semi-insulating (SI) substrate. The thickness of substrate 302 can be chosen to be any suitable thickness, typically between about 150 μm and 750 μm. Substrate 302 can include one or more layers, for example, the substrate can include a buffered substrate, such as a buffered Si substrate that is engineered to have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. A material such as a substrate having a lattice constant that nearly matches the lattice constant of GaAs or Ge means that the material such as the substrate has a lattice constant different than that of GaAs or Ge by less than or equal to 3%, less than 1%, or less than 0.5% of the lattice constant of GaAs or Ge. Examples of buffered silicon substrates that can provide a lattice constant approximately equal to that of GaAs or Ge include SiGe buffered Si, SiGeSn buffered Si, and rare-earth (RE) buffered Si, such as a rare-earth oxide (REO) buffered Si. A layer such as SiGe, SiGeSn, or a RE-containing layer can form a buffer layer (or lattice engineered layer) grown on a substrate such as Si having a low number of defects and/or dislocations in the buffer layer. The buffer layer can provide a lattice constant at the top of the buffer layer approximately equal to that of a GaAs or Ge substrate, facilitating the ability to form high quality III-V layers on top of the buffer layer, with a low number of defects and/or dislocations in the III-V semiconductor layers and/or dilute nitride layers. A low number of defects can include comparable or fewer defects than would occur in an In0.53Ga0.47As layer grown on an InP substrate.
First contact layer 304a and first barrier layer 304b provide a first doped region 305, having a doping of one type, and second barrier/window layer 308a and second contact layer 308b provide a second doped region 307, having a doping of the opposite type. If first doped layer 305 is doped n-type, second doped layer 307 is doped p-type. Conversely, if first doped region 305 is doped p-type, second doped region 307 is doped n-type. Examples of p-type dopants include C and Be. Examples of n-type dopants include Si and Te. Doped region 305 and 307 can be chosen to have a composition that is lattice matched or pseudomorphically strained with respect to the substrate. The doped regions can comprise any suitable III-V material, such as GaAs, AlGaAs, GalnAs, AlInP, (Al)GaInP, (Al)GaInPAs, GaInNAs, and GaInNAsSb. The contact and barrier region and doped layers can have different compositions and different thicknesses. The bandgap of the doped regions and doped layers can be selected to be larger than the bandgap of active region 306. In some embodiments, the bandgap of the doped region and doped layers, or at least a portion of the doped layers can be selected to be larger than the bandgap of GaAs such that optical absorption by the doped layers in the visible wavelength range is reduced. In particular, for a device intended to be a photodetector illuminated through the top surface, second barrier/window layer 308a can include a material such as AlInP, AlGaAs, (Al)GaInP, or (Al)GaInPAs. The larger bandgap of layer 308a reduces optical absorption in this layer for visible wavelengths of light, allowing visible light to be absorbed within active region 306. This can reduce the short wavelength cutoff for a photodetector from about 0.9 μm to about 0.4 μm, thereby allowing the photodetector to have a responsivity over a broader wavelength range. The use of window/barrier layer 308a allows a reduced thickness for second contact layer 308b, further reducing the optical losses for layer 308b, through which incident light passes into the active region 306 of device 300. The doping level of first contact layer 304a can be chosen to be higher than the doping level of first barrier layer 304b. A higher doping facilitates electrical connection with a metal contact. Similarly, the doping level of second contact layer 304b can be chosen to be higher than the doping level of second barrier layer 304a. Higher doping levels facilitate electrical connection with a metal contact. Doping levels can be within a range, for example, from 1×1015 cm−3 to 2×1019 cm−3. Doping levels may be constant within a layer and/or the doping profile may be graded. For example, the doping level can increase from a minimum value to a maximum value as a function of the distance from the interface between the doped layer 308a and the active region 306. Each of barrier and contact layers 304a, 304b, 308a and 308b can independently have a thickness, for example, within a range from 50 nm to 3 μm, from 100 μm to 2 μm, or from 200 nm to 1 μm.
Active region 306 can be lattice matched or pseudomorphically strained to the substrate and/or to the barrier layers 304a and 308a. The bandgap of active region 306 can be lower than that of barrier and contact layers 304a, 304b, 308a and 308b. Active region 306 can comprise a single active layer or multiple active layers capable of processing light over a desired wavelength range. Processing is defined to be a light emission, a light receiving, a light sensing and light modulation.
An active layer can include a dilute nitride material. The dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where x, y and z can be 0≤x≤0.4, 0<y≤0.07 and 0<z≤0.04, respectively. X, y and z can be 0.01≤x≤0.4, 0.02≤y≤0.07 and 0.001≤z≤0.04, respectively. In other embodiments, dilute nitride materials can have compositions as disclosed in U.S. Pat. No. 8,962,993, where x, y and z can be 0≤x≤0.24, 0.02≤y≤0.05 and 0.001≤z≤0.2, respectively. A dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where, for example, 0.12≤x≤0.24, 0.03≤y≤0.07 and 0.005≤z≤0.04; 0.13≤x≤0.2, 0.03≤y≤0.045 and 0.001≤z≤0.02; 0.13≤x≤0.18, 0.03≤y≤0.04 and 0.001≤z≤0.02; or 0.18≤x≤0.24, 0.04≤y≤0.07 and 0.01≤z≤0.04. An active layer can have a bandgap within a range from 0.7 eV to 1.0 eV such that the active layer can absorb light at wavelengths up to 1.8 μm. Bismuth (Bi) may be added as a surfactant during growth of the dilute nitride, improving material quality (such as defect density), and the device performance. The thickness of an active layer can be, for example, within a range from 0.2 μm to 10 μm or from 1 μm to 4 μm. The minority carrier concentration of an active layer can be, for example, less than 1×1016 cm−3 (measured at room temperature, 25° C.), less than 5×1015 cm−3, or less than 1×1015 cm−3. Active layer 306 can be compressively strained with respect to the substrate 302. Strain can also improve device performance. For a photodetector, the parameters most relevant to device performance include the dark current, operating speed, noise and responsivity. In
The semiconductor layers 402, 404a, 404b, 406, 408a and 408b correspond to layers 302, 304a, 304b, 306, 308a and 308b, respectively, of device 300. Multiple lithography and materials deposition steps may be used to form the metal contacts, passivation layer, and luminescent layer. The device has a mesa structure, produced by etching. This exposes the underlying layers. A passivation layer 414 is provided that covers the side-walls of the device and the exposed surfaces of layers so as to reduce surface defects and dangling bonds that may otherwise affect device performance. The passivation layer 414 can be formed using a dielectric material such as, for example, silicon nitride, silicon oxide, or titanium oxide.
Luminescent layer 416 is configured to absorb at ultraviolet wavelengths and to emit light at longer wavelengths such as at wavelengths that can be absorbed by active region 406 of device 400. Luminescent layer 416 can be an organic material and may be a fluorescent or a phosphorescent material that is able to absorb at UV wavelengths of light, and re-emit, either though fluorescence or phosphorescence, at visible wavelengths of light and that can be absorbed by active region 406 of device 400. Luminescent layer 416 can absorb light at wavelengths of light, for example, between about 150 nm and about 450 nm and can emit light at wavelengths between about 450 nm and 650 nm. Luminescent layer 416 can have a thickness, for example, of about 1 μm or can have a thickness between about 0.1 μm and about 2 μm.
Examples of luminescent materials include Lumigen® chemiluminescent reagent available from Beckman Coulter Company, Unichrome® phosphors described in U.S. Pat. No. 5,795,617 and available from Acton Optics and coatings, other organic materials such as those described in U.S. Pat. No. 5,986,268, and inorganic coatings such as those described by Franks in “Inorganic Phosphor Coatings for Ultraviolet Responsive Image Detectors”, MSc thesis, University of Waterloo, 2000.
Optionally, an anti-reflection or encapsulant layer (not shown) can overlie luminescent layer 416. The antireflection or encapsulant layer can include dielectric materials that are transparent at ultraviolet wavelengths as low as about 0.2 μm such as Al2O3, and MgF2, and the thickness can be, for example, from about 10 nm and 400 nm.
As shown in
Substrate 602 can have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. The lattice constants of GaAs and Ge are 5.65 Å and 5.66 Å, respectively, and growth of III-V materials with similar compositions without defects can be grown on either substrate. The close matching of the lattice constants of Ge and GaAs allows, for example, high-quality GaAs to be epitaxially grown on a Ge surface. In some embodiments, the substrate can be GaAs. Substrate 602 may be doped p-type, or n-type, or may be a semi-insulating (SI) substrate. The thickness of substrate 602 can be chosen to be any suitable thickness. Substrate 602 can include one or more layers, for example, the substrate can include a buffered substrate, such as a buffered Si substrate that is engineered to have a lattice constant that matches or nearly matches the lattice constant of GaAs or Ge. A material such as a substrate having a lattice constant that nearly matches the lattice constant of GaAs or Ge means that the material such as the substrate has a lattice constant different than that is less than or equal to 3%, less than 1%, or less than 0.5% of the lattice constant of GaAs or Ge. Examples of buffered silicon substrates that can provide a lattice constant approximately equal to that of GaAs or Ge include SiGe buffered Si, SiGeSn buffered Si, and rare-earth (RE) buffered Si, such as a rare-earth oxide (REO) buffered Si. A layer such as SiGe, SiGeSn, or a RE-containing layer can form a buffer layer (or lattice engineered layer) grown on a substrate such as Si having a low number of defects and/or dislocations in the buffer layer. The buffer layer can provide a lattice constant at the top of the buffer layer approximately equal to that of a GaAs or Ge substrate, facilitating the ability to form high quality III-V layers on top of the buffer layer, with a low number of defects and/or dislocations in the III-V semiconductor layers and/or dilute nitride layers. A low number of defects can include comparable or fewer defects than would occur in an In0.53Ga0.47As layer grown on an InP substrate.
Etch stop/release layer 603 can be provided to allow removal of substrate 602 through a combination of physical and chemical methods. Etch-stop/release layer 603 can be lattice matched or pseudomorphically strained with respect to substrate 602. The composition of layer 603 can be chosen to have a different etch chemistry than that of substrate 602 and first contact layer 604a. For example, using a GaAs substrate 602, layer 603 can include AlInP and GaInP. P-containing layers have a high etch selectivity with As-containing layers, allowing a layer of one type to be removed chemically, and therefore substrate 602 and etch-stop/release layer 603 may both be removed from device 600. Substrate removal is necessary for a bottom-illuminated device, because the substrate would otherwise prevent light with wavelengths less than about 0.9 μm being absorbed by active region 606. A comprehensive list of wet etchants, etch rates, and selectivity relationships is provided in Clawson, Materials Science and Engineering, 31 (2001) 1-438, Elsevier Science B.V.
First contact layer 604a and first barrier/window layer 604b provide a first doped region 605, having a doping of one type, and second barrier layer 608a and second contact layer 608b provide a second doped region 607, having a doping of the opposite type. If first doped region 605 is doped n-type, second doped region 607 is doped p-type. Conversely, if first doped region 605 is doped p-type, second doped region 607 is doped n-type. Examples of p-type dopants include C and Be. Examples of n-type dopants include Si and Te. Doped regions 605 and 607 can be chosen to have a composition that is lattice matched or pseudomorphically strained with respect to the substrate. The doped region can comprise any suitable III-V material, such as GaAs, AlGaAs, GalnAs, AlInP, (Al)GaInP, (Al)GaInPAs, GaInNAs, and GaInNAsSb. The contact and barrier layers can independently have different compositions and different thicknesses. The bandgap of the doped layers can be selected to be larger than the bandgap of active region 606. In some embodiments, the bandgap of the doped layers, or at least a portion of the doped layers can be selected to be larger than the bandgap of GaAs such that optical absorption by the layers in the visible wavelength range is reduced. In particular, for a device intended to be a photodetector illuminated through the bottom surface, first barrier/window layer 604b can include a material such as AlInP, AlGaAs, (Al)GaInP, or (Al)GaInPAs. The larger bandgap of first barrier/window layer 604b reduces optical absorption in this layer for visible wavelengths of light, allowing them to be absorbed within active region 606. This can reduce the short wavelength cutoff for a photodetector from about 0.9 μm to about 0.4 μm, thereby allowing the photodetector to have a responsivity over a broader wavelength range. The use of window/barrier layer 604b allows a reduced thickness for second contact layer 604a, further reducing the optical losses for layer 604a, through which incident light passes into the active region 606 of device 600. The doping level of first contact layer 604a can be chosen to be higher than the doping level of first barrier/window layer 604b. A higher doping facilitates electrical connection with a metal contact.
Similarly, the doping level of second contact layer 608b can be chosen to be higher than the doping level of second barrier layer 608a. Higher doping levels facilitate electrical connection with a metal contact. Doping levels can be within a range, for example, from 1×1015 cm−3 to 2×1019 cm−3. Doping levels may be constant within a region or layer and/or the doping profile may be graded, for example, the doping level can increase from a minimum value to a maximum value as a function of the distance from the interface between the doped layer 608a and the active region 606. Each of barrier and contact layers 604a, 604b, 608a and 608b can independently have a thickness, for example, within a range from 50 nm to 3 μm, from 100 nm to 2 μm, or from 200 nm to 1 μm.
Active layer 606 can be lattice matched or pseudomorphically strained with respect to the substrate and/or to the barrier layers. The bandgap of active μm 606 can be lower than that of barrier and contact layers 604a, 604b, 608a and 608b. Active μm 606 can comprise a layer capable of processing light over a desired wavelength range. Processing is defined to be a light emission, a light receiving, a light sensing and light modulation.
Active region 606 can include one or more active layers. An active layer can include a dilute nitride material. The dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where x, y and z can be 0≤x≤0.4, 0<y≤0.07 and 0<z≤0.04, respectively. X, y and z can be 0.01≤x≤0.4, 0.02≤y≤0.07 and 0.001≤z≤0.04, respectively. In other embodiments, dilute nitride materials can have compositions as disclosed in U.S. Pat. No. 8,962,993, where x, y and z can be 0≤x≤0.24, 0.02≤y≤0.05 and 0.001≤z≤0.2, respectively. A dilute nitride material can be Ga1-xInxNyAs1-y-zSbz, where, for example, 0.12≤x≤0.24, 0.03≤y≤0.07 and 0.005≤z≤0.04; 0.13≤x≤0.2, 0.03≤y≤0.045 and 0.001≤z≤0.02; 0.13≤x≤0.18, 0.03≤y≤0.04 and 0.001≤z≤0.02; or 0.18≤x≤0.24, 0.04≤y≤0.07 and 0.01≤z≤0.04. An active layer can have a bandgap within a range from 0.7 eV to 1.0 eV such that the active layer can absorb light at wavelengths up to 1.8 μm. Bismuth (Bi) may be added as a surfactant during growth of the dilute nitride, improving material quality (such as defect density), and the device performance. The thickness of an active layer can be, for example, within a range from 0.2 μm to 10 μm or from 1 μm to 4 μm. The carrier concentration of an active layer can be, for example, less than 1×1016 cm−3 (measured at room temperature, 25° C.), less than 5×1015 cm−3, or less than 1×1015 cm−3. An active layer can be compressively strained with respect to the substrate 602. Compressive strain can also improve device performance. For a photodetector, the parameters most relevant to device performance include the dark current, operating speed, noise and responsivity. Active region 606 is shown as a single layer, but it will be understood that active region 606 can include more than one active layer such as more than one dilute nitride region, with at least two bandgaps between 0.7 eV and 1.0 eV. Examples of multi-bandgap and graded bandgap active layers are described in U.S. Application No. 62/816,718, filed on Mar. 11, 2019. In some examples, active region 606 can include active layers with different doping profiles. Examples of doping profiles for dilute nitride optical active layers are described in U.S. Application Publication No. 2016/0118526.
The semiconductor layers/regions 704a, 704b, 706, 708a and 708b correspond to layers/regions 604a, 604b, 606, 608a and 608b, respectively, of device 600. Multiple lithography and materials deposition steps may be used to form the metal contacts, passivation layer, and antireflection coating. The device has a mesa structure, produced by etching. This exposes the underlying layers. A passivation layer 714 is provided that covers the side-walls of the device and the exposed surfaces of layers and/or regions so as to reduce surface defects and dangling bonds that may otherwise affect device performance. The passivation layer 714 can be formed using one or more dielectric materials including, for example, aluminum oxide, silicon nitride, silicon oxide, and titanium oxide.
Luminescent layer 716 absorbs ultraviolet wavelengths and emits light at longer (visible) wavelengths that can be absorbed by active region 706 of device 700. Luminescent layer 716 can be an organic material and can be a fluorescent or a phosphorescent material that is able to absorb UV wavelengths of light, and re-emit, either though fluorescence or phosphorescence, visible wavelengths of light that can be absorbed by the active region 706 of device 700. For example, luminescent layer can absorb light at wavelengths of light between about 150 nm and about 450 nm and emit light at wavelengths between about 450 nm and 650 nm. Luminescent layer 716 can have a thickness, for example, of about 1 μm or can have a thickness between about 0.1 μm and about 2 μm.
Examples of luminescent materials include Lumigen® chemiluminescent reagent available from Beckman Coulter Company, Unichrome® phosphors described in U.S. Pat. No. 5,795,617 and available from Acton Optics and coatings, other organic materials such as those described in U.S. Pat. No. 5,986,268, and inorganic coatings such as those described by Franks in “Inorganic Phosphor Coatings for Ultraviolet Responsive Image Detectors”, MSc thesis, University of Waterloo, 2000.
Optionally, an anti-reflection or encapsulant layer (not shown) can underlie luminescent layer 716. The antireflection or encapsulant layer can include dielectric materials that are transparent at ultraviolet wavelengths as low as about 0.2 μm such as Al2O3, and MgF2.
A first metal contact 710 overlies a portion of the first contact layer 704a. A second metal contact 712 overlies a second portion of second contact layer 708b. Metallization schemes for contacting to n-doped and p-doped materials are known to those ordinarily skilled in the art. Photodetector 700 can be illuminated via the top surface of the device, i.e. through the interface between luminescent layer 716 (or an underlying antireflection layer) and air.
Responsivity was measured using a broad-band halogen lamp, with light monochromatized at 10 nm wavelength steps and calibrated using a NIST traceable InGaAs detector.
Arrays of photodetectors may also be formed using photodetectors provided by the present disclosure. An array of top-illuminated devices (such as device 400 or 500) may be surface-mounted to and wire-bonded to an underlying substrate and read-out circuitry. An array of bottom-illuminated devices (such as device 700) may be flipped vertically such that the bottom surface faces up and provides the illumination surface, and the top surface faces toward an underlying substrate and read-out circuitry. Devices may be electrically connected to the readout circuitry using an array of indium bumps on each detector (or pixel) of an array and the readout circuitry. For an array of detectors, the collected signals may be amplified by a readout integrated circuit (ROIC) comprising a transistor or a trans-impedance amplifier to form a Focal Plane Array (FPA).
Examples of photodetector arrays are shown in
To function as a spectral sensor, the incident light on a photodetector or photodetector array may be spectrally filtered. In an array of photodetectors, because luminescence from the luminescent material associated with UV light absorption is emitted at a visible wavelength, some photodetectors in an array may be coated with luminescent material, and thereby are UV-enhanced photodetectors, while other photodetectors in an array may not be coated with a luminescent material, such that only a selected number of pixels within the array are sensitive to UV light.
More generally, in some embodiments, a photodetector array may be divided into sub-regions, with a different spectral filter overlying each of the sub-regions such that each sub-region is sensitive to a selected range of incident wavelengths. At least one optical filter overlies the photodetector array. At least one sub-region of the photodetector array includes UV-enhanced photodetectors.
The number of pixels 1001/1051 underlying each of the optical filters may vary, for example, according to the sensitivity of the pixels at different wavelengths (or wavelength ranges) and/or the power of incident light at the wavelength or wavelength range. An electrical signal for each wavelength range, (and corresponding sub-region of device 1000 or device 1050) to be measured may be generated by a single pixel 1001, or a plurality of pixels underlying each filter region of device 1000 or device 1050. A larger number of pixels may be used for light detection at wavelengths where the responsivity (measured in A/W) is lower and a fewer number of pixels may be used for light detection at wavelengths where the responsivity is higher.
In some embodiments, an optical filter can have a fixed transmission wavelength for all pixels underlying the optical filter. An optical filter may include multiple different dielectric layers with different refractive indices and of desired thicknesses to allow a desired transmission wavelength range. In other embodiments, an optical filter may be a variable optical filter, having a lower and an upper wavelength cutoff defining a wavelength transmission range, where the transmission through the optical filter may vary spatially across the surface of the optical filter, with narrower and different sub-wavelength ranges within the broader wavelength range being transmitted to each underlying pixel. For example, a first pixel may receive light a first wavelength range, and a second pixel may receive light in a second wavelength range. This can increase the number of different wavelength ranges (spectral bands) that may be measured and resolved by device 1000 or device 1050. A variable optical filter may be achieved for example, by varying the thickness of one or more of the optical filter layers across the filter. The thickness change may be continuous, for example using a wedge filter, or it may be discrete, with different layer thicknesses used above each individual pixel underlying the filter. Examples of wedge-like filters are described in U.S. Pat. No. 7,575,860. Other optical filter designs having different thicknesses that are capable of providing variable transmission characteristics are described in U.S. Pat. No. 9,261,634, and in U.S. Pat. No. 10,170,509.
Combinatorial etching and deposition techniques may be used to produce a multi-level wavelength filter across an array.
Spectral filtering may also be achieved using a diffraction grating to disperse light of different wavelength across an array, or to select a specific and tunable narrow wavelength band incident on a single photodetector. The grating may be a reflection grating or a transmission grating. Gratings are periodic structures that function to diffract different wavelengths of light from a common input path into different angular output paths. For an array of photodetectors, different wavelengths can be received by different pixels of the array, according the angular path between the grating and pixels. For a single photodetector, the grating may be rotated to steer different wavelengths of light onto a single photodetector. An example of a transmission grating is a surface relief transmission grating. Another example of a transmission grating is a volume phase holographic (VPH) grating. A VPH grating can be formed in a layer of transmissive material, such as a dichromated gelatin, which is sealed between two layers of clear glass or fused silica. The phase of incident light is modulated as it passes through the optically thick film that has a periodic differential hardness or refractive index. This is in contrast to a conventional grating in which the depth of a surface relief pattern modulates the phase of the incident light.
To fabricate optoelectronic devices provided by the present disclosure, a plurality of layers can be deposited on a substrate in a materials deposition chamber such as an MBE and/or MOCVD deposition chamber. The plurality of layers may include active layers, doped layers, contact layers, etch stop layers, release layers (i.e., layers designed to release the semiconductor layers from the substrate when a specific process sequence, such as chemical etching, is applied), buffer layers, or other semiconductor layers.
The plurality of layers can be deposited, for example, by molecular beam epitaxy (MBE) or by metal-organic chemical vapor deposition (MOCVD). Combinations of deposition methods may also be used.
A semiconductor optoelectronic device can be subjected to one or more thermal annealing treatments after growth. For example, a thermal annealing treatment can include the application of a temperature within a range from 400° C. to 1000° C. for from 10 seconds to 10 hours. Thermal annealing may be performed in an atmosphere that includes air, nitrogen, arsenic, arsine, phosphorus, phosphine, hydrogen, forming gas, oxygen, helium and any combination of the preceding materials.
Devices provided by the present disclosure can comprise a GaInNAsSb active layer overlying a GaAs substrate. The GaInNAsSb layer can be compressively strained with respect to the GaAs substrate. For example, the XRD peak slitting between the GaInNAsSb peak and the GaAs substrate peak can be, for example, from 600 arcsec to 1,000 arcsec, from 600 arcsec to 800 arc sec, or from 650 arcsec to 750 arcsec.
A dilute nitride layer such as a GaInNAsSb layer can have an intrinsic or unintentional doping equivalent to a doping concentration, for example, less than 1×1016 cm−3, less than 5×1015 cm−3, or less than 1×1015 cm−3, measured at room temperature (25° C.). A dilute nitride layer such as a GaInNAsSb layer can have an intrinsic or unintentional doping equivalent to a doping concentration, for example, from 0.5×1014 cm−3 to 1×1016 cm−3 or from 1×1015 cm−3 to 5×1015 cm−3, measured at room temperature (25° C.).
A dilute nitride layer such as a GaInNAsSb layer can have a minority carrier lifetime, for example, from 1.0 ns to 3.0 ns, from 1.5 ns to 2.5 ns, or from 1.5 ns to 2.0 ns. A dilute nitride layer such as a GaInNAsSb layer can have a minority carrier lifetime, for example, greater than 1.0 ns, greater than 1.5 ns, greater than 2.0 ns, or greater than 2.5 ns. To determine the minority carrier lifetime of the GaInNAsSb layer, time-resolved photoluminescence (TRPL) may be used. The TRPL kinetics are measured at room temperature (25° C.) at an excitation wavelength of 970 nm, with an average CW power of 0.250 mW, and a pulse duration of 200 fs generated by a Ti:Sapphire:OPA laser with a pulse repetition rate of 250 kHz and a laser beam diameter at the sample of 1 mm.
A dilute nitride layer such as a GaInNAsSb layer can have a bandgap, for example, from 0.7 eV to 1.0 eV, such as from 0.75 eV to 0.95 eV, or from 0.7 eV to 0.8 eV.
The absorption bandwidth of a dilute nitride layer such as a GaInNAsSb layer can have a full width half maximum, for example, from 50 nm to 150 nm, from 50 nm to 125 nm, from 50 nm to 70 nm, or from 75 nm to 125 nm, as determined by photoluminescence.
The dilute nitride layer such as a GaInNAsSb layer can have a thickness, for example, from 0.25 μm to 3.0 μm, from 0.5 μm to 2.0 μm, or from 0.5 μm to 1.0 μm.
A device such as a photodetector can have a diameter, for example, from 20 μm to 3 mm, from 0.5 mm to 2.5 mm, or from 1 mm to 2 mm. A device such as a photodetector can have a diameter, for example, greater than 20 μm, greater than 100 μm, greater than 500 μm, greater than 1 mm, or greater than 2 mm.
A UV-enhanced photodetector having a dilute nitride active layer can have the structure shown in
A UV-enhanced photodetector having a dilute nitride active layer can have the structure shown in
Spectral sensors can have an array of photodetectors as shown in
The present disclosure includes the Appendix entitled Dilute Nitride Photodetector Arrays for Sensing Applications, including pages 1-7. The Appendix is incorporated by reference in its entirety.
Finally, it should be noted that there are alternative ways of implementing the embodiments disclosed herein. Accordingly, the present embodiments are to be considered as illustrative and not restrictive. Furthermore, the claims are not to be limited to the details given herein and are entitled their full scope and equivalents thereof.
This application claims priority to U.S. Provisional Application Ser. No. 62/958,601, filed Jan. 8, 2020, which is hereby incorporated by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 62958601 | Jan 2020 | US |
