RU2606200C1 - Diagnostic technique for electric microheterogeneitiesin semiconductor heterostructures, based on ingan/gan - Google Patents

Abstract

FIELD: electronics.

SUBSTANCE: invention relates to microelectronics and can be used for imaging electric microheterogeneities of technological origin: dislocations, pores, precipitates, etc. in semiconductor heterostructures with arbitrary design of active region, grown on Al₂O₃ substrates. Method of diagnostics of electric microheterogeneities in semiconductor heterostructures, based on InGaN/GaN substrate Al₂O₃ together with heterostructure and surface of earthed aluminium foil is irradiated by electron beam with energy density from 0.1 to 0.8 J/cm² and rate of increase of electric field intensity in heterostructure is no lower than 5⋅10¹³ V/cm⋅s. Limit power density is determined, above which electric discharges and related microdestruction arise in heterostructure. Microdestructions are recorded using optical microscope after multi-pulse radiation dose of not less than 6⋅10^-6 Kl/cm². Size and spatial distribution of electric microheterogeneities are visually indicated in heterostructure.

EFFECT: method enables diagnosing in atmospheric air without using complex and expensive equipment on substrates with arbitrary design of active region.

1 cl, 4 dwg

Description

The invention relates to semiconductor microelectronics and can be used to visualize the spatial distribution of electrical microinhomogeneities of technological origin: dislocations, pores, precipitates, etc. in semiconductor epitaxial heterostructures based on InGaN / GaN grown on Al ₂ O ₃ substrates.

По отклонению положения интерференционного пика тормозного излучения на шкале детектора от угла падения рентгеновского луча определяют погрешность положения образца. С учетом полученной погрешности независимым перемещением устанавливают трубку в положение

при котором ось симметрии между трубкой и детектором перпендикулярна к выбранной системе кристаллографических плоскостей. При таком положении трубки проводят пошаговое сканирование в диапазоне углов, характеризующих выбранную систему кристаллографических плоскостей. Независимым перемещением устанавливают трубку на угол

выводя максимум тормозного пика за границы характеристического пика. Затем проводят пошаговое сканирование всех слоев гетероструктуры, оставляя неизменным угловое положение характеристического пика от системы кристаллографических плоскостей путем перемещения шкалы детектора, и определяют угловые положения пиков от всех слоев гетероструктуры.A known method for the diagnosis of semiconductor epitaxial heterostructures [1. RU 2498277 C1, IPC G01N 23/02 (2006.01), H01L 21/66 (2006.01), publ. November 10, 2013], including scanning a sample under conditions of Bragg reflection in a step-by-step mode produced by changing the angle of incidence of the X-ray beam, using X-ray single-crystal diffractometry with a non-monochromatic, quasi-parallel X-ray beam and a position-sensitive detector. The x-ray tube and detector are set relative to the angular position of the characteristic peak from one of the systems of crystallographic planes of the heterostructure at an angle

The deviation of the position of the interference peak of bremsstrahlung on the detector scale from the angle of incidence of the x-ray beam determines the error in the position of the sample. Taking into account the obtained error, the tube is set to position by independent movement

in which the axis of symmetry between the tube and the detector is perpendicular to the selected system of crystallographic planes. With this position of the tube, step-by-step scanning is carried out in the range of angles characterizing the selected system of crystallographic planes. Independently move the tube at an angle

bringing the maximum of the braking peak beyond the boundaries of the characteristic peak. Then, step-by-step scanning of all layers of the heterostructure is carried out, leaving the angular position of the characteristic peak from the system of crystallographic planes unchanged by moving the detector scale, and determining the angular positions of the peaks from all layers of the heterostructure.

The disadvantage of this method is the inability to determine in which defective areas the greatest localization of the electric current will occur. This is especially important when analyzing heterostructures intended for the manufacture of semiconductor devices.

A known method for detecting microdefects that cause channels of current leakage in the emitting crystals InGaN / GaN high-power LEDs [2. Zakheim A.L., Kuryshev G.L., Mizerov M.N. et al. Investigation of thermal processes in high-power InGaN / GaN flip-chip LEDs using infrared thermal imaging microscopy // FTP. 2010. V. 44, No. 3. P. 390-396], including the transmission of electric current through the LEDs and the measurement of temperature fields arising in the LEDs as a result of self-heating, using a UTK-1 infrared microscope with a resolution of a unit of micrometers, in which thermal radiation is recorded by an InAs device with a charge-injection photodetector array with the number of elements 128 × 128 (element pitch 50 μm) and a spectral sensitivity range of 2.5-3.1 μm. The field of view of the microscope is 400 × 400 μm (~ 3 μm per element). The resulting distribution map of infrared radiation corresponds to the distribution of current density, because the temperature at a certain point in the heterostructure is directly proportional to the current density with the dominance of nonradiative recombination during heat release. Current leakage channels (areas of electrical microinhomogeneities) in infrared digital photographs appear as dark spots.

The main disadvantage of this method is the impossibility of studying heterostructures before cutting into chips and assembling them into an LED. In addition, the need to produce an LED for several hundred hours leads to a long duration of the diagnostic process.

A known method for identifying channels of increased transport of minority charge carriers in the active region of semiconductor heterostructures based on InGaN / GaN [3. Yakimov E.B. Characterization of GaN and structures based on it by scanning electron microscopy // Scientific notes of the Faculty of Physics. 2014. T. 2, 142501. S. 1-5], selected as a prototype, including the deposition of Schottky barriers Ni / Au on the surface of the sample, scanning the heterostructure by an electron beam in vacuum at room temperature with a current of ~ 10 ^-10 A in a raster electron a Jeol-840A microscope with a Keithley 428 current amplifier, and measuring the spatial distribution of induced electric currents in a heterostructure. Based on the obtained distribution of the induced current collected at each scanning point, the local electrical characteristics of the heterostructure are judged, and the areas of the increased transport channels of minority charge carriers and extended defects in the heterostructures are identified.

In the induced current mode, structural defects in heterostructure layers with a small amount of InGaN / GaN quantum wells (<5) appear as dark dots due to a local increase in the recombination rate in defective regions.

For structures with a large number of quantum wells (≥5), dark points associated with defects are practically not detected, and against the background of a large-scale inhomogeneity of the recombination rate, two types of defects appear, giving a bright contrast in the induced current mode.

Small (0.1-0.2 μm) light points are associated with dislocations, large light areas in the images are associated with clusters of dislocations and / or with micropipes.

The main advantages of this method include:

1. The ability to diagnose heterostructures to the stage of cutting the plate into chips and assembling them into a light-emitting device.

2. The ability to visualize channels of increased transport of charge carriers in the active region of InGaN / GaN-based heterostructures.

The main disadvantages of this method include:

1. A significant effect of the design of the active region of the heterostructure on the dependence of the induced current on the primary electron energy and on the image of extended defects due to the effective capture of nonequilibrium charge carriers by quantum wells.

2. The inability to identify defective areas in which the greatest localization of electric current occurs.

3. The need to use a vacuum system and other complex and expensive equipment.

4. The need for preliminary preparation of the studied samples (spraying Schottky barriers), the complexity and duration of the diagnostic process.

The objective of the invention is the diagnosis of electrical microinhomogeneities in semiconductor heterostructures based on InGaN / GaN grown on Al ₂ O ₃ substrates with an arbitrary design of the active region.

A method for diagnosing electrical microinhomogeneities in semiconductor heterostructures based on InGaN / GaN grown on Al ₂ O ₃ substrates involves irradiating the heterostructure with an electron beam, which leads to visualization of the spatial distribution of electrical microinhomogeneities. The Al ₂ O ₃ substrate, together with the heterostructure and the grounded aluminum foil deposited on its surface, is irradiated in atmospheric air by an electron beam with an energy density of 0.1 to 0.8 J / cm ² and an increase in the electric field strength in the heterostructure of at least 5⋅10 ¹³ V / cm⋅s. The threshold energy density is determined above which electric discharges and related micro-fractures occur in the heterostructure. Microfracture recorded with an optical microscope after multi-pulse irradiation dose of not less 6⋅10 ^-3 C / cm ^2. Visually judge the size and spatial distribution of electrical microinhomogeneities in the heterostructure.

The basis of the proposed method for the diagnosis of electrical microinhomogeneities in semiconductor heterostructures is the phenomenon of electrical breakdown of solids under the influence of high-current electron beams of nanosecond duration, the effect of the accumulation of microdestructions during multipulse irradiation [4. Oleshko V.I., Shtanko V.F. The mechanism of destruction of high-resistance materials under the influence of powerful electron beams of nanosecond duration // FTT. 1987.Vol. 29, No. 2. P. 320-324] and the well-known property of solids is a decrease in electric strength in the presence of structural defects in them [5. Grekhov IV, Serezhkin Yu.N. Avalanche breakdown of the pn junction in semiconductors. L .: Energy. Leningra. Department, 1980. S. 84-90]. An increase in the electron beam energy density above the threshold value P _n (or current density j _p at constant energy and current pulse duration) leads to an increase in the internal electric field strength in the irradiated object and the initiation of electric breakdown in the region of electron beam deceleration. In contrast to low-current electron beams having a current density j from 10 ^-10 to 10 ^-5 A / cm ² , high-current (J≥10 A / cm ² ) allow initiating an electrical breakdown not only in materials that accumulate the injected negative space charge, in the so-called cathodoelectrets, but also in high-resistance materials of various classes — ion crystals, semiconductors of group A ₂ B ₆ and polymers. This is due to the fact that the radiation-pulse conductivity of solids does not depend linearly, but more weakly, on the power of the ionizing radiation source. In this case, the energy of high-current electron beams, in contrast to low-current ones, is released in high-resistance materials through two main channels. Part of the energy is released homogeneously as a result of ionization of the medium by high-energy electrons (ionization losses), and part is converted into the energy of the electric field of a negative space charge and is released locally in the regions of electrical microinhomogeneities, which leads to heating and destruction of the irradiated sample [6. IN AND. Oleshko. Abstract of dissertation for the degree of Doctor of Philosophy "Threshold processes in solids in the interaction with high-current electron beams." Specialty 04/01/07. 2009, S. 11]. In thin-film semiconductor heterostructures grown on Al ₂ O ₃ substrates, the thickness of which exceeds the mean free path of the electron beam, the same processes develop, as evidenced by experimental data: close thresholds of electron-beam destruction of macrocrystals and thin-film semiconductor heterostructures in the multi-pulse irradiation mode, local the nature of microdestruction and the effect of the accumulation of microdestruction with increasing radiation dose.

The minimum threshold for the electron beam energy density of 0.1 J / cm ² is determined by the minimum rate of increase of the electric field strength in the InGaN / GaN semiconductor heterostructure, at which an electrical breakdown develops and local micro-fractures associated with it are formed.

The maximum threshold in the energy density of the electron beam of 0.8 J / cm ² due to the destruction of the sample by dynamic stresses (thermal shock), manifests itself in the form of cracks and is not associated with electrical breakdown of the sample.

The exposure dose threshold of 6⋅10 ^-5 C / cm ² is associated with the effect of the accumulation of microdestructions and is due to the minimum radiation dose at which all electrical microinhomogeneities present in the irradiation zone of the semiconductor heterostructure are visualized. With an increase in the radiation dose above the threshold, a slight increase in the size of the existing microdestruction is observed.

List of illustrations:

FIG. 1. The implementation scheme of the diagnostic method.

FIG. 2. Photographs of the spatial distribution of microdestructions corresponding to localization sites of electrical microinhomogeneities in the heterostructure of GS-1 sample.

FIG. 3. Photographs of the spatial distribution of microdestructions corresponding to localization sites of electrical microinhomogeneities in the heterostructure of the GS-2 sample.

FIG. 4. Photographs of the spatial distribution of microdestructions corresponding to localization sites of electrical microinhomogeneities in the heterostructure of the GS-3 sample.

Example

In FIG. 1 shows a diagram of the implementation of the proposed diagnostic method, including an electron accelerator 1, aperture 2 with a variable diameter of the hole, aluminum foil 3, an electron beam 4, aperture 5 mounted in the groove of the metal table 6, aluminum foil 7 deposited on a heterostructure 8 grown on a substrate Al ₂ O ₃ 9, optical microscope 10 with micron resolution. In FIG. 1 also shows the electron mean free path R _e in the sample and the region in which a negative space charge 11 is formed.

The electron accelerator 1 is a high-current pulse accelerator design G.A. Mesyatsa and B.M. Kovalchuk [7. Kovalchuk B.M., Mesyats G.A., Semin B.M., Shpak V.G. A high-current nanosecond accelerator for studying fast processes // Instruments and experimental technique. 1981. No. 4. P. 15-18] with a current pulse duration of 15 ns, an average electron energy in the beam of 250 keV and a maximum pulse current of 3 kA.

The diaphragm 2 with a variable diameter of the hole, installed behind the anode of the vacuum diode of the electron accelerator 1, is used to discretely vary the energy density of the electron beam 4 on the surface of the heterostructure 8. The holes of the diaphragm 2 are designated from D1 to D15, they correspond to the energy densities of the electron beam 4 from 0.1 up to 0.8 J / cm ² in increments of 0.05 J / cm ² . The energy density of the electron beam 4 at the potential location of the heterostructure 8 is measured by the radiation-chemical method [8. Serikov L.V., Yurmazova T.A., Shiyan L.N. The method of dosimetry of ionizing radiation / a.s.No 1544030. 1989].

As aluminum foil 3 and aluminum foil 7 use aluminum tape with a thickness of 15-20 microns. Aluminum foil 3 prevents the penetration of air into the vacuum diode of electron accelerator 1. Aluminum foil 7 ensures uniformity of the electric field in the braking zone of the electron beam in the sample. Aluminum foil 7 is grounded, because in contact with the metal table 6, placed on the grounded casing of the electron accelerator 1.

The diaphragm 5 limits the irradiation zone of the sample and cut out from the electron beam 4 a region with a uniform energy density over the cross section. A metal table 6 with a hole is used to fix the diaphragm 5 and as a support for placing the diagnosed sample.

An optical microscope 10 is used to detect microdestructions generated at localization sites of electrical microinhomogeneities after multi-pulse irradiation of the heterostructure 8. As the optical microscope 10, a transmitted light microvisor μVizo-101 is used.

Diagnostic samples are three InGaN / GaN-based LED heterostructures with different active-region designs grown by metal-organic gas-phase epitaxy on 400 μm thick sapphire substrates [0001].

The heterostructure of the GS-1 sample consists of a 4-μm thick n-GaN layer, an active region containing ten In _0.14 Ga _0.86 N quantum wells 2 nm thick, and 15 nm GaN barriers, a p-Al _0.1 Ga ₀ layer _{, 9} N with a thickness of 30 nm and a p-GaN layer with a thickness of 200 nm.

The heterostructure of the GS-2 sample consists of a 3-μm thick n-GaN layer, an active region containing five In _0.12 Ga _0.88 N quantum wells 2.5 nm thick, and 10 nm GaN barriers, a p-Al layer of _0.1 Ga _0.9 N with a thickness of 20 nm and a p-GaN layer with a thickness of 140 nm.

The heterostructure of the GS-3 sample consists of a 3-μm thick n-GaN layer, an active region containing one In _0.12 Ga _0.88 N quantum well 2.5 nm thick, and a 10-nm GaN barrier, p-Al _0.1 Ga _0.9 N with a thickness of 20 nm and a p-GaN layer with a thickness of 140 nm.

The proposed method for the diagnosis of electrical microinhomogeneities is implemented as follows. Set the diaphragm 2 with the hole D1 corresponding to the energy density of the electron beam P = 0.1 J / cm ² . The vacuum diode, which is part of the electron accelerator 1, is pumped out to a pressure of 10 ^-3 Torr.

An aluminum foil is glued onto the surface of the heterostructure 8 of the GS-1 sample 7. The diagnosed sample is placed on a metal table 6 with a diaphragm 5 whose diameter is 2 mm and irradiated at room temperature in atmospheric air in a multipulse mode with a dose of D _{p of} at least 6 менее10 ^{- 5} C / cm ² . After irradiation, the aluminum foil 7 is removed from the GS-1 sample surface, placed on a μVizo-101 transmitted light microvisor stage, and the irradiated surface of the semiconductor heterostructure is photographed with a spatial resolution of ~ 1 μm. Visually, photographs judge the presence of microdestructions in the irradiated zone of the heterostructure. In the absence of microdestruction, the energy density of the electron beam 4 is discretely increased through an interval of 0.05 J / cm ² by sequential use of the diaphragm 5 with holes D2, D3, etc. and repeat photographing the irradiated zones of the heterostructure of the GS-1 sample for each energy density after multipulse irradiation with a dose of at least 6⋅10 ^-5 C / cm ² until the threshold energy density P _{p of the} electron beam 4 is reached, which is characterized by the formation of micro damage in the irradiated zones (Fig. 2A), initiated by electrical breakdown. For the GS-1 sample, the threshold energy density of the electron beam thus determined is P _n = 0.15 J / cm ² .

The entire surface of the heterostructure 8 is diagnosed by moving the GS-1 sample relative to the diaphragm 5 and irradiating the new regions of the heterostructure 8 with an electron beam 4 with the previously established parameters P _p and D _p . On the basis of a visual inspection of the irradiated surface of the sample using a transmitted light microvisor, μVizo-101 judges the size and spatial distribution of microdestructions, which are localization sites of electrical microinhomogeneities in a semiconductor heterostructure. The place of the greatest accumulation of microdestructions in the diagnosed heterostructure of the GS-1 sample is shown in FIG. 2b.

Multipulse irradiation of the heterostructure with a dose equal to the threshold D _p allows us to visualize all types of defects of technological origin located in the irradiation zone and responsible for the electrical microinhomogeneity of the diagnosed heterostructure.

Knowing the parameters of the electron beam 4 (average electron energy U = 250 keV, duration of the electron beam current pulse t = 15 ns) and the experimentally determined threshold energy density of the electron beam for initiating electric breakdown in the GS-1 sample (P _n ≈0.15 J / cm ² ), we estimate the magnitude of the electric field strength and its growth rate in the heterostructure grown on Al ₂ O ₃ . When the grounded irradiated surface of the heterostructure 8 and a thickness of Al ₂ O ₃ 9 exceeding the electron mean free path R _e (Fig. 1):

where E is the electric field strength, V / cm;

ε ₀ is the dielectric constant, f / cm;

ε is the dielectric constant of the sample;

ρ _e is the bulk density of the injected charge, C / cm ³ ;

j _e is the amplitude of the electron current density, A / cm ² ;

Q _e is the surface density of the injected charge, C / cm ² ;

R _e is the effective range of electrons in the sample, cm;

R _p - threshold energy density of the electron beam, J / cm ² ;

j _p - threshold amplitude of the electron current density, A / cm ² ;

U is the average electron energy, eV;

t is the duration of the current pulse of the electron beam, s

Since the thermalization of the negative space charge 11 of the electron beam 4 occurs in the Al ₂ O ₃ 9 substrate, then at R _p = 0.15 J / cm ² , j _p = 40 A / cm ² , t = 15 ns, Q _e ⁼ j _p ⋅t = 6⋅10 ^-7 C / cm ² and ε = 9.3, the field strength in Al ₂ O ₃ reaches E≈0.7⋅10 ⁶ V / cm, and dE / dt≈4.7⋅10 ¹³ B / (cm⋅s). In the InGaN / GaN-based heterostructure 8 located on the Al ₂ O ₃ 9 surface, the amplitude values of the electric field strength and its growth rate will be, respectively, E≈1.2⋅10 ⁶ V / cm and dE / dt≈0.8 ⋅10 ¹⁴ V / (cm⋅s). In this case, a fast (electric-discharge) component of the current induced by the electron beam neutralizes the electric field associated with the negative space charge 11 of the electron beam 4 injected into Al ₂ O ₃ .

An electrical breakdown initiated in semiconductors and dielectrics upon irradiation with a high-current electron beam with a rise rate of at least dE / dt ~ 10 ¹⁴ V / (cm⋅ s) develops in the form of streamer discharges (filamentous current channels) [4. Oleshko V.I., Shtanko V.F. The mechanism of destruction of high-resistance materials under the influence of powerful electron beams of nanosecond duration // FTT. 1987.Vol. 29, No. 2. S. 320-324]. Non-equilibrium charge carriers in this case are formed by impact ionization or tunneling effect near the head of the streamer, wherein the electric field E reaches a value of 10 ⁶ to 10 ⁷ V / cm, the carriers concentration n _e - from October ¹⁹ to 10 ²⁰ cm ^-3 , current density j _e ~ 10 ⁶ A / cm ² , which leads to local destruction of the samples due to Joule heating.

Thus, the experiment and calculation indicate that a positive effect for an InGaN / GaN based semiconductor heterostructure grown on an Al ₂ O ₃ substrate is achieved at electron beam energy densities from 0.15 to 0.8 J / cm ² and velocity increase in electric field strength on its surface is not less than 0.8⋅10 ¹⁴ V / (cm⋅s).

At electron beam energy densities of more than 0.8 J / cm ² , fractures in the form of cracks associated with thermal shock begin to appear in heterostructures, which excludes the possibility of further use of the heterostructure. In FIG. 2 s photographs of cracks appearing in the heterostructure of the GS-1 sample after multipulse irradiation with an electron beam with an energy density of 1 J / cm ² and a dose of at least 6 менее10 ^-5 C / cm ^{2 are} presented.

GS-2 and GS-3 samples are used to demonstrate the possibility of diagnosing electrical microinhomogeneities by the proposed method in heterostructures with an arbitrary InGaN / GaN active region design. The diagnosis of electrical microinhomogeneities in the heterostructures of samples GS-2 and GS-3 is carried out as described above for sample GS-1. The threshold energy density of the electron beam determined from the experiment, which is necessary to initiate microdestruction in the heterostructures of samples GS-2 and GS-3, is P _n = 0.2 J / cm ² . Photographs of the spatial distribution of electrical microinhomogeneities in the heterostructures of samples GS-2 and GS-3 are shown in FIG. 3 (a, b) and FIG. 4 (a, b), respectively.

Based on a visual comparison of the spatial distribution patterns of electrical microinhomogeneities in the GS-1, GS-2 and GS-3 samples obtained using the μVizo-101 microvisor, it is concluded that there is no influence of the active region design on the visualization of electrical microinhomogeneities. The greatest number of electrical microinhomogeneities per unit surface area contains the heterostructure of the GS-1 sample, and the smallest is the heterostructure of the GS-3 sample.

Using the proposed method for the diagnosis of electrical microinhomogeneities in semiconductor heterostructures based on InGaN / GaN grown on Al ₂ O ₃ substrates provides the following advantages in comparison with the prototype:

1. The ability to carry out diagnostics in the air, eliminating the need to manufacture a special vacuum chamber and the use of sophisticated and expensive equipment, which reduces the time required for diagnostics by eliminating the preliminary preparation of the test samples (spraying Schottky barriers), and also eliminates the complex operation of measuring induced currents, which together leads to an acceleration and simplification of diagnosis.

2. Improving the reliability of diagnostic results, because the proposed method allows us to determine not all types of defects formed in the semiconductor heterostructure during the growing process, but only those defective regions in which localization of the electric current occurs, which leads to the appearance of local microdestructions, which will allow us to reject these parts of the heterostructure when cutting the plate into chips.

3. The ability to diagnose the spatial distribution of electrical microinhomogeneities in InGaN / GaN based heterostructures having an arbitrary design of the active region, because the effect of localization of electrical breakdown and related microdestruction is determined precisely by the presence of electrical microinhomogeneities, and not by the composition of the layers (including the active region) of the heterostructure.

Claims (1)

A method for diagnosing electrical microinhomogeneities in InGaN / GaN-based semiconductor heterostructures grown on Al ₂ O ₃ substrates, comprising irradiating the heterostructure with an electron beam, resulting in visualization of the spatial distribution of electrical microinhomogeneities, characterized in that the Al ₂ O ₃ substrate together with the heterostructure and deposited on its surface grounded aluminum foil is irradiated in the atmosphere by an electron beam with an energy density of 0.1 to 0.8 J / cm ² and slew rate Voltage nnosti electric field in the heterostructure 5⋅10 not lower than ¹³ V / sm⋅s determine the threshold energy density, above which a heterostructure having electric discharges and related microfracture which register with an optical microscope after multi-pulse irradiation dose of not less 6⋅10 ^-5 C / cm ² , and visually judge the size and spatial distribution of electrical microinhomogeneities in the heterostructure.

Images