RU203297U1 - TWO WAVE PHOTODIODE FOR MEDIUM WAVE INFRARED RADIATION - Google Patents

Abstract

The utility model relates to photonics, namely, to medium-wave infrared (IR) optical devices based on A3B5 semiconductors, operating as part of pyrometers. A two-wave photodiode for medium-wave infrared radiation based on a monolithic heterostructure containing an InAs semiconductor substrate with an electronic type of conductivity and epitaxial layers grown on one of its surfaces from narrow-gap semiconductors A3B5 with a hole and electronic type of conductivity with a long-wavelength photosensitivity increasing along the boundary wavelength in which the layer on the surface of the heterostructure and the layer adjacent to the substrate are made with an electronic type of conductivity. 9 ill.

Description

The utility model relates to photonics, namely, to medium-wave infrared (IR) optical devices based on semiconductors A ³ B ⁵ , operating as part of pyrometers [Sotnikova G. Yu. et al, "Mid-infrared radiation technique for direct pyroelectric and electrocaloric measurements", 2020, Rev. Sci. Instrum., V.91, 1 ArtNo: # 015119, DOI: http://dx.doi.org/10.1063/1.5108639] and / or gas analyzers [B.A. Matveev, G.Yu. Sotnikova, “Medium-wave IR LEDs based on A ³ B ⁵ heterostructures in gas analytical instrumentation. Possibilities and Applications 2014-2018 ", Optics and Spectroscopy, 2019, Volume 127, Issue. 2, 300-305]. The first time representatives of such devices were sources and receivers of radiation from indium arsenide, which evolved from simple "diffusion" diodes based on homo pn junctions [N.P. Esina et al., "Investigation of the electroluminescence of pn junctions in indium arsenide," FTT, vol. 9, c. 5, pp. 1324-1328 (1967)] to light and photodiodes (PDs), including avalanche PDs based on pin-homo structures [Ian C. Sandall et al, "Linear array of InAs APDs operating at 2 pm" , OPTICS EXPRESS 25780, Vol. 21, No. 22 | DOI: 10.1364 / OE.21.025780 | (2013)] and heterostructures.

The development of methods for detecting and classifying the object of measurement, as well as methods for masking the target, makes it necessary to switch from the above-mentioned single-wave photodetectors to multi-wavelength ones, which make it possible to use the ratio method to determine the temperature of the object T and reduce the influence of interference. Indeed, applying the Planck formula

to the measurement data of the photocurrent in two narrow bands of radiation with wavelengths λ1 and λ2, it is possible to form a (photo) signal that weakly depends on the emissivity of the object by setting ε1 = ε2:

with the desired result [A. Rogalski, "lnfrared Detectors", second ed., CRC Press, Tailor and Francis Group, 2012, International Standard Book Number: 978-1-4200-7671-4] as:

The most popular are T measurements in two ranges of the electromagnetic radiation spectrum: 3-5 μm (medium-wave infrared region - MIR-IR or MW) and 8-14 μm (long-wave infrared region - LWIR or LW), corresponding to the first and second transparency windows of the atmosphere.

From the above it becomes clear that the single-wave matrix photodetectors produced by the domestic industry (for example, the Sapfir plant) [P. Gindin et al., "Matrix and submatrix photodetector modules", PHOTONICS, Issue No. 6/42, pp. 62-72 (2013)] have a limited range of applications, therefore the creation of two-wave photodetectors [E. Dmitriev, “Photodetectors for work in multispectral optoelectronic systems. Problems of creation ", ELECTRONICS: Science, Technology, Business, 8, 2005, p. 2, p. 7 4-79] is an urgent task.

The problem of creating multiwave photodetectors was initially solved by creating photodetector modules containing discrete semiconductor photodetectors with an increasing boundary wavelength of photosensitivity along the incoming beams [V.P. Ponomarenko, A.M. Filachev, "lnfrared Techniques and Electro-optics in Russia: A History 1946-2006", SPIE vol. PM165, ISBN 0-9194-6355-8 (see page 96)]. This well-known technical solution (according to the NIIPF nomenclature - in a 28NF blade type module) uses a Si photodiode, a PbS photoresistor and an InSb photodiode in series, covering the total spectral range from 1 to 5 μm. In this case, all three photodetectors are separated from each other by an air (or vacuum) layer located between the sapphire substrate, on which PbS is deposited, and the silicon photodetector, as well as between the immersion lens of the PbS photodetector and the InSb photodiode.

More advanced is the layered architecture for multi-wavelength IR matrices, which can be created by "gluing" several matrices into a single whole [M.A. Kinch, "HDVIP FPA technology at DRS Infrared Technologies", Proc. SPIE 4369, 566 (2001), Peter D. Dreiske, "Development of Two-Color Focal-Plane Arrays Based on HDVIP®", Proc. of SPIE Vol. 5783, (2005), doi: 10.1117 / 12.609127], installation of a certain sequence of rulers in scanning devices [K.V. Kozlov et al., "Modern infrared photodetectors for scanning equipment for remote sensing of the Earth (review)", Uspekhi Applied Physics, 2017, volume 5, No. 1, p. 63] or when growing layers with different photosensitivity boundaries on a single substrate [E .R. Blazejewski et al, "Bias-switchable dual-band HgCdTe infrared photodetector", J. Vac. Sci. Technol. B10,1626 (1992)]. In the latter case, the most common is the creation of monolithic structures of the n-p-n or p-n-p type by growing from organometallic gases (MOVPE) or from molecular beams (MBE).

In such monolithic photodetectors, the short wavelength region of photosensitivity (Short wave - SW) is located first along the course of the rays incident on the element, behind it is a wide-gap barrier layer in which there is no absorption; adjacent to it is a second photosensitive area (Long wave - LW). In this version of the photodiode array, there are only 2 contacts per pixel, therefore the choice of the spectral range is carried out by choosing the polarity of the applied voltage. When the bias is applied, one of the diodes is turned on in the reverse direction and the other in the forward direction. A reverse-shifted PD (for example, a SW PD) contributes to the photocurrent, while a forward-shifted (LW PD) does not. When the polarity of the bias is changed, the situation changes, and the previously biased PD (LW PD) begins to generate a photocurrent, while the other (SW PD) is considered "short-circuited". The disadvantage of this design is that it is difficult to select the optimal displacement for each of the p-n junctions, among the advantages is the ease of implementation. Due to the large difference in the resistance of the PD, in practice, additional switching is used, which allows you to connect two different amplifiers to optimize their operation [W. A. Radford et al., "Third Generation FPA Development Status at Raytheon Vision Systems," Proceedings of SPIE 5783, 331-39, 2005].

The choice of the spectral range for measurements in the above-described "two-contact" design can also be carried out using "optical switching" using an external source [OO. Cellek, Y.-H. Zhang, "Optically addressed multiband photodetector for infrared imaging applications," Proc. SPIE 8268, Quantum Sensing and Nanophotonic Devices IX, 82682N (20 January 2012); doi: 10.1117 / 12.909063]. In the absence of illumination from the side of the additional source, the device generates a signal corresponding to the change in the photocurrent in the short-wave diode due to the fact that the current in the circuit is limited by the dark current of the diode, which is many orders of magnitude lower than the dark current of the long-wavelength receiver.

The above principle of constructing radiation detectors can be extended to a larger number of spectral ranges, for example, for 3. In this case, the ratio of the resistance of each part of the device must satisfy simple criteria - the resistance of the short-wave photodetector must be at least an order of magnitude higher than the sum of the resistances for the other two parts of the structure.

The separation of spectral ranges is also possible through the use of "vertical" Bragg microcavities tuned to two different wavelengths, for example, in a structure with photoconductive PbTe layers on a Si substrate [Jianfei Wang et al, "Monolithically integrated, resonant cavity-enhanced dual -band mid-infrared photodetector on silicon, Appl. Phys. Lett. 100, 211106 (2012); https://doi.org/10.1063/1-4722917]. According to the authors, such an approach for the creation of two-wave radiation detectors, for example, at wavelengths of 1.6 and 3.7 mm, is less laborious than the creation of structures with active layers of different chemical compositions.

Known two-wave photodiode for the mid-infrared region of the spectrum based on a monolithic heterostructure containing a semiconductor substrate with an electronic type of conductivity, transparent for radiation in the region of the working spectrum of the photodiode, epitaxial layers grown on it from narrow-gap semiconductors with hole and electronic type of conductivity, at least two of which have a thickness comparable to the reciprocal of the absorption coefficient of radiation in the operating wavelength range and increasing along the rays of the long-wavelength cutoff wavelength of photosensitivity, ohmic contacts, at least one of which is conjugated with one of the layers with a hole-type conductivity [Yanka Robert W., Noble Milton L, "A-dual band IR sensor having two monolithically integrated staring detector arrays for simultaneous, coincident image readout," US Pat. No. 5,512,750, priority dated June 3, 1994].

In the known solution, on opposite sides of the CdZnTe substrate oriented in the (111) plane, i.e. On the (111) A and ( 111 ) B sides, pn structures based on a CdHgTe solid solution were grown by liquid-phase epitaxy or from molecular beams in an ultrahigh vacuum, with a narrow-gap absorbing layer and an adjacent pn junction located on the back side of the heterostructure. Thanks to this approach, photodiodes and arrays have been created that are sensitive in the mid-wave and long-wave regions of the IR spectrum, with the possibility of independent registration of signals in the two above-mentioned spectral regions.

The disadvantages of the known solution are the high cost of CdZnTe substrates used for epitaxial growth, the high degree of spatial inhomogeneity of the composition of the CdHgTe solid solution (and, accordingly, the characteristics of the PD) due to the high partial pressure of Hg during the growth of structures, as well as the instability of metallurgical boundaries due to weak bond Hg-Te. All of the above leads to a decrease in the percentage of yield of suitable products due to the spread of parameters obtained in one technological process.

The disadvantage of the known solution is also the difficulty of its use in designs with immersion optics due to the lack of a free surface of the substrate for its mounting. At the same time, all currently existing large-scale applications of medium-wave PDs on InAs substrates are associated with the use of immersion lenses.

The main disadvantage of the known solution is the forced use of two separated in time technological processes for growing structures: the existing methods and equipment assume the growth of layers only on one of the substrate surfaces in one process. This increases the time required to create the heterostructure, increases the cost of production, and often leads to a deterioration in the parameters of the manufactured photodetectors.

The task of the utility model is to improve the operational characteristics by using materials А ³ В ⁵ and providing the possibility of using immersion conjugation with lenses.

The problem posed is solved by the fact that in a two-wave photodiode for the mid-infrared region of the spectrum based on a monolithic heterostructure containing a semiconductor substrate with an electronic type of conductivity, transparent for radiation in the region of the working spectrum of the photodiode, epitaxial layers grown on it from narrow-gap semiconductors with a hole and electronic type of conductivity , at least two of which have a thickness comparable to the reciprocal of the absorption coefficient of radiation in the operating wavelength range and the long-wavelength cutoff wavelength of photosensitivity increasing along the rays, ohmic contacts, at least one of which is conjugated with one of the layers with hole type of conductivity, the substrate is made of indium arsenide, epitaxial layers are made of semiconductors А ³ В ⁵ and are located on one of its sides, while the layer on the surface of the heterostructure and the layer adjacent to the substrate are made with electronic type of conductivity.

In the proposed solution, two independent p-n heterostructures are sequentially grown on one side of the InAs substrate, in which layers based on narrow-gap semiconductors A³IN^five, such as, InAs, lnAs_1-xSb_x, InAs_1-xySb_xP_y, In_1-xGa_xAs_{1- y}Sb_y, have a thickness comparable to the reciprocal of the absorption coefficient of radiation in the operating wavelength range. Such layers are usually called "absorbing" and they are located with increasing along the rays of the long wavelength cutoff wavelength of the photosensitivity. In addition to the semiconductors listed above, other solid solutions can be used as the material of the absorbing layers, for example, GaInAsSbP (and GaInAsSbP / InAs heterostructures), AlInPSbAs (and AlInPSbAs / InAs heterostructures), as well as layers with quantum dots and nanowires.

The claimed device is illustrated by a drawing, where FIG. 1 schematically shows the first embodiment of the inventive photodiode in longitudinal section, and the vertical arrow shows the direction of the photon flux incident on the PD. In this device, 3 layers (2, 3, 4) are grown on an n-type substrate (1), two of which (2, 4) have n-type conductivity, and one (3) has p-type conductivity. The first along the ray path (2) has a smaller red cutoff wavelength of photosensitivity than the third layer (4) (λ2 _cut-off <λ4 _cut-off ). The layers are grown on one side of a substrate that is transparent in the working range (1). To each of the layers (2, 3, 4), individual ohmic contacts are connected, connected to the conductors: (5) - to layer (3), (6) - to layer (2), (7) - to layer (4) ...

The inventive device is also illustrated by a drawing, where in FIG. 2 schematically shows the second embodiment of the photodiode, where the numbers (2, 4) mean absorbing layers with an electronic type of conductivity, surrounded by wider-gap (non-absorbing) layers with an electronic type of conductivity, for example, representing double heterostructures (DHS), and the number ( 3) a set of layers with p-type conductivity is understood. In this case, a layer can also be understood as a system of layers with a given type of conductivity, for example, a superlattice with the required effective band gap. The electrical connection of the layers (2) to the external circuit in this embodiment is via a conductive substrate (1) to which one of the contacts (6) is connected.

The PD works as follows: the flux of external excess / nonequilibrium photons, shown by arrows in Figs. 1, 2, enters the PD and some of the photons with the highest energy are absorbed in layer (2), creating electron-hole pairs in it, which diffuse to the boundary p- and n-type semiconductors in the direction perpendicular to them. The pairs fall into the space charge region at the interface of p- and n-type semiconductors, where they are separated by the field of the pn junction and change the potential drop across it. With a short-circuited circuit, a photocurrent flows through contacts 5 and 6, which will form a useful signal in the first "short-wave" part of the spectrum. The photocurrent, usually proportional to the number of absorbed photons, is the useful signal used to measure the characteristics of the radiation incident on the PD. For practical purposes, the useful signal is amplified using amplifiers, for example, using transimpedance amplifiers [GA Gavrilov, et al., "The limiting sensitivity of a photodetector based on photodiodes A ³ B ^{5 of the} mid-IR spectrum", Letters ZhTF, 37 (18), 50-57 (2011)]

Simultaneously with the above processes, the "long-wave" radiation not absorbed in the layers (2) penetrates into the layer (4) with a greater cutoff wavelength of photosensitivity or into the region near the interface between the layers (3) and (4). In this case, processes similar to those described above occur, and a photocurrent corresponding to the intensity of the photon flux in the long-wavelength part of the photosensitivity spectrum of the two-wave PD begins to flow in the circuit formed with the participation of contacts (5), (7).

Example.

Epitaxial 6-layer structures were grown at IoffeLED LLC on n-InAs (100) substrates by LPE in a single technological process and contained two double heterostructures (DHS, English - DH) with the following layers (starting from the substrate): 1. wide-gap "Window" N-InAs _1-xy Sb _x P _y (х = 0.12 ÷ 0.13; у = 0.2 ÷ 0.24), t = 1.5 ÷ 2.5 μm, Eg (77 K) = 470 ÷ 490 meV; 2. short-wavelength photosensitive region of n-InAs, 1 = 3 ÷ 4 μm, Eg (77 K) = 410 meV; 3.wide-gap "window" / p-type contact layer Р-InAs _1-xy Sb _x P _y (x = 0.12 ÷ 0.13; y = 0.2 ÷ 0.24), t = 1.5 ÷ 2.5 μm, E _g (77K) = 470 ÷ 490 meV; 4. wide-gap "window" (contact / buffer layer) InAs _1-xy Sb _x P _y (x = 0.11 ÷ 0.12; y = 0.07 ÷ 0.08), t = 1.5 ÷ 2.5 μm, Ε _g (77 K) = 380 ÷ 390 meV; long-wavelength photosensitive region InAs _1-x Sb _x (x = 0.08 ÷ 0.09), t = 3 ÷ 4 μm, E _g (77 K) = 330 ÷ 340 meV; contact layer InAs _1-xy Sb _x P _y , (x = 0.11 ÷ 0.12; y = 0.07 ÷ 0.08), t = 2 ÷ 3 μm, E _g (77 K) = 380 ÷ 390 meV. The growth of the second double heterostructure was not carried out on a small portion of the substrate; therefore, in this “reference” region, there was only the first three-layer InAsSbP / InAs / InAsSbP heterostructure.

FIG. 3 shows the distribution of P, Sb, as well as impurities (Zn, Sn) over the thickness in the 6-layer part of the structure, obtained by the method of dynamic secondary ion mass spectrometry, which confirms the above description of the structure. FIG. 4 shows the photoluminescence (PL) spectra measured "for reflection" in the 3- and 6-layer parts of the structure at 77 K. The long-wavelength maximum in each of the curves corresponds to the radiation of the active / photosensitive region. The difference in the band gap of the photosensitive region and the wide-gap layer was 50 and 80 meV. The obtained PL spectra are almost identical to the photoluminescence spectra of GVDs used to obtain single-wavelength photodetectors with photosensitivity maxima of 3.4 and 4.0 μm.

Methods multistage photolithography and "wet" chemical etching photodiodes samples were fabricated with contacts to separate the epitaxial layers and the absorbing circular region with a diameter D ₁ = 320 microns (the absorption layer InAs, A ₁ = 8⋅10 ^-4 cm ²⁾ and D ₂ = 180 microns (the absorption layer InAsSb, A ₂ = 2.5⋅10 ^-4 cm ²⁾ as shown in the diagrams in Fig. 5, Fig. 6 and in the photograph in FIG. 7, where the symbol "A" denotes the anode and the symbol "C" the cathodes. Then the chips were mounted on a sub-chip board and then on a TO-18 package, or they were docked with an immersion silicon lens with a diameter of 3.5 mm using chalcogenide glass and mounted in a Sr-type screw package, as shown in [www.ioffeled.com].

In fig. FIG. 8a and FIG. 8b shows the current-voltage characteristics (VAC) obtained with different connection options: individual / independent connection of the DHS (Fig. 8a: forward bias to contacts AC1 and AC2, and in Fig. 8b: forward and reverse I-V characteristics for connecting AC2) and " interdependent "VAC when connecting contacts C1-C2 (Fig. 8b). The I - V characteristic when connecting the contacts C1-C2 (Fig. 8b) most clearly illustrates the presence of two pn junctions, turned on towards each other. In this case, the values of the densities of the dark current in each of the pn junctions have values close to the values obtained with single-wave photodetectors.

In the left panel of FIG. 9 shows the spectra of Si photosensitivity in each of the channels of a PD mounted in an Sr package with an immersion lens at room temperature, and the right panel shows the spectra _of detectivity obtained in the standard way: D * = S 1 ⋅rsqrt (R ₀ A / 4kT). As seen from the data in FIG. 9, the obtained operational parameters are close to those for single-wave photodetectors with maximum photosensitivity of 3.4 and 4.0 μm [M.A. Remennyy et al, "InAs and InAs (Sb) (P) (3-5 pm) immersion lens photodiodes for portable optic sensors" SPIE Proceedings Vol. 6585 2007, 658504, DOI: 10.1117 / 12.722847], which confirms the usefulness of the utility model.

Claims (1)

A two-wavelength photodiode for the mid-infrared region of the spectrum based on a monolithic heterostructure containing a semiconductor substrate with an electronic type of conductivity, transparent for radiation in the region of the working spectrum of the photodiode, epitaxial layers grown on it from narrow-gap semiconductors with a hole and electronic type of conductivity, at least two of which have a thickness comparable to the reciprocal of the absorption coefficient of radiation in the operating wavelength range and the long-wavelength cutoff wavelength of photosensitivity increasing along the beams, ohmic contacts, at least one of which is conjugated with one of the layers with a hole-type conductivity, which differs in that that the substrate is made of indium arsenide, the epitaxial layers are made of A3B5 semiconductors and are located on one of its sides, while the layer on the surface of the heterostructure and the layer adjacent to the substrate are made with an electronic type of conductivity.

Images