Ingaas photodetector pn junction region profile laser beam induced current detection method, substrate and system

Abstract

This invention discloses a method, substrate, and system for detecting laser beam-induced current in the PN junction region profile of an InGaAs photodetector, belonging to the field of semiconductor optoelectronic device characterization technology. The method includes: cleaving a chip with completed front-side processing along the natural cleavage surface of InP to obtain a smooth profile; fixing the chip with the profile facing upwards onto a composite sapphire bonding substrate, and extracting electrodes through conductive silver paste and ultrasonic bonding; performing a two-dimensional step-scan of the profile using a 900-1100nm laser, simultaneously acquiring photocurrent signals to obtain a two-dimensional photocurrent distribution map of the profile; and quantitatively extracting junction depth, lateral diffusion width, and uniformity parameters based on the photocurrent peak position. This invention can directly obtain the longitudinal two-dimensional electrical activity distribution of the PN junction region, effectively avoiding interference between the surface passivation layer and the InP cap layer, and achieving a visual assessment of junction depth, lateral diffusion, and defects, providing a realistic and efficient detection method for device process optimization.

Description

Technical Field

[0001] This invention belongs to the field of semiconductor optoelectronic device characterization technology, specifically relating to a method, substrate, and system for detecting laser beam-induced current in the PN junction region profile of an InGaAs photodetector. Background Technology

[0002] The core performance of InGaAs photodetectors is determined by the vertical and horizontal doping distribution of the PN junction. As the core functional unit of the photodetector, the structural characteristics of the PN junction (such as junction depth and doping distribution) directly determine key performance parameters such as quantum efficiency, response speed, and spectral characteristics. Therefore, accurately characterizing the impurity doping distribution of the PN junction and the uniformity of the photoresponse of the photosensitive element is a crucial foundation for optimizing device fabrication and improving detection performance.

[0003] In existing technologies, scanning capacitance microscopy (SCM) can only obtain static carrier concentration and cannot simulate the working state under illumination; secondary ion mass spectrometry (SIMS) only provides a one-dimensional distribution of chemical composition and cannot distinguish electrically active impurities; traditional front-incident laser beam-induced current (LBIC) is limited by the passivation layer, InP cap layer and light injection depth, and can only detect near-surface information, and is insensitive to deep absorption layers and interface defects. All of the above methods struggle to simultaneously provide a two-dimensional electrical activity distribution in the PN junction region, resulting in low efficiency and large errors in process-performance correlation analysis. Therefore, there is an urgent need for a new detection method that can directly observe the PN junction profile, take into account both depth and lateral information, and is immune to surface process fluctuations. Summary of the Invention

[0004] Purpose of the invention: In order to solve the above technical problems, the present invention provides a method, substrate and system for detecting cross-sectional laser beam induced current that can directly and quickly obtain the two-dimensional electrical activity distribution of the PN junction region of an InGaAs photodetector, thereby realizing quantitative characterization of junction depth, lateral diffusion width and uniformity, and providing a real basis for process optimization.

[0005] Technical solution: The present invention provides a method for detecting laser beam-induced current in the PN junction region profile of an InGaAs photodetector, comprising: A photodetector chip with an InGaAs absorption layer that has completed the front-side process is diced and cleaved to obtain a smooth profile perpendicular to the PN junction interface. The cleaved chip is fixed to the soldering substrate with the cross-section facing upwards, and the electrical interconnection between the chip's N electrode, P electrode and the corresponding conductive area of ​​the substrate is achieved by conductive silver paste and ultrasonic bonding, respectively. The cross-section was scanned in two dimensions using an LBIC testing system. The laser beam was incident vertically onto the exposed cross-section, and the photocurrent signal of each pixel was collected simultaneously to obtain a two-dimensional distribution map of the photocurrent of the cross-section. Based on the peak position of photocurrent, the precise boundaries of the PN junction in the depth and lateral directions are determined, and the junction depth, lateral diffusion width and uniformity parameters are quantitatively extracted to achieve a profile characterization of the doping distribution and carrier collection efficiency in the PN junction region.

[0006] Furthermore, the cleavage occurs along the natural cleavage plane of the InP-based material, with a cleavage surface roughness Ra≤5nm, eliminating the need for subsequent mechanical polishing.

[0007] Furthermore, the conductive substrate is a composite sapphire structure, including a main sapphire sheet and a secondary sapphire sheet fixed to the left side of the main sapphire sheet by photoresist. The upper surface of the secondary sapphire sheet is provided with an Au layer and is connected to the N electrode of the chip by conductive silver paste. The upper surface of the right side of the main sapphire sheet is provided with an Au layer and is connected to the P electrode of the chip by ultrasonic bonding metal leads. There is an isolation gap between the two Au layers to prevent silver paste overflow and short circuit.

[0008] Furthermore, the Au layer has a thickness of 300nm–500nm and an isolation spacing width of 100μm–300μm.

[0009] Furthermore, the LBIC scanning parameters are: laser wavelength 900nm–1100nm, beam spot diameter ≤5μm, step interval 0.1μm–1μm, incident power 1μW–100μW, and scanning range covering the entire profile depth and at least one complete photosensitive element period.

[0010] Furthermore, this method also includes: The two-dimensional photocurrent distribution map obtained by scanning is integrated along the depth direction to obtain the lateral diffusion curve, and the lateral diffusion width is extracted by the full width at half maximum (FWHM) method. Connect the peak photocurrent positions of multiple pixels in the same photosensitive element array direction to evaluate junction depth consistency, and use the 3σ value as the evaluation index of process uniformity.

[0011] Furthermore, after the test is completed, the entire conductive substrate is immersed in acetone to dissolve the photoresist, so that the chip can be separated from the substrate without damage. The substrate can be reused after cleaning.

[0012] Furthermore, the InGaAs photodetector is a planar PIN structure, which, from bottom to top, includes an N-type InP substrate, an N-type InP buffer layer, an intrinsic InGaAs absorption layer, an N-type InP cap layer, a SiNx or SiO2 passivation layer, and a P-type region and a front P electrode formed by Zn diffusion.

[0013] This invention provides a solderable substrate for use in the method of claim 1, comprising: The main sapphire wafer has an Au layer on its upper right surface. The secondary sapphire wafer is cured onto the left side of the primary sapphire wafer using photoresist, and its upper surface has an Au layer. An insulating trench is provided between the Au layer on the upper right side of the main sapphire sheet and the Au layer on the upper surface of the secondary sapphire sheet to prevent the conductive adhesive from overflowing. The Au layer on the upper surface of the secondary sapphire wafer is used to interconnect with the N electrode of the chip using conductive silver paste, and the Au layer on the upper right side of the primary sapphire wafer is used to bond with the P electrode of the chip via wires to achieve electrical lead-out during LBIC testing.

[0014] Furthermore, the thickness of both the main sapphire wafer and the secondary sapphire wafer is 250μm–350μm, the shear strength after photoresist curing is ≥20MPa, and the overall surface flatness of the substrate is ≤5μm.

[0015] This invention provides a LBIC detection system for the PN junction region profile of an InGaAs photodetector, comprising: The solderable substrate as described above; The LBIC scanning unit is used to output a focusable laser with a wavelength of 900nm–1100nm and perform two-dimensional scanning of the cross-section of a chip fixed to a soldering substrate. The current detection unit is connected to the positive and negative terminals of the conductive soldering substrate and is used to synchronously acquire photocurrent signals. The data processing unit is used to convert the photocurrent signal into a two-dimensional distribution map and extract parameters such as junction depth, lateral diffusion width, and uniformity.

[0016] Furthermore, the current detection unit adopts a phase-locked amplification mode with an integration time of 1ms–100ms and a noise current ≤10pA.

[0017] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: (1) It works directly under illumination to obtain carrier collection information consistent with the actual operation of the device, thus solving the defect that SCM cannot simulate illumination; (2) Provides a complete two-dimensional electrical activity distribution in the PN junction region, making up for the deficiency of SIMS which only has one-dimensional chemical information; (3) The laser acts directly on the exposed profile, eliminating the attenuation and distortion of the signal by the surface passivation layer, InP cap layer and interface state, and the detection depth can reach the entire absorption layer; (4) Utilizing the natural cleavage surface of InP-based materials, with a roughness Ra≤5nm, mechanical polishing is not required, simplifying sample preparation and avoiding the introduction of damage; (5) The composite sapphire conductive substrate enables rapid chip assembly and non-destructive disassembly. The substrate can be reused, reducing testing costs. Attached Figure Description

[0018] Figure 1 This is a cross-sectional view of a composite sapphire solder substrate; Figure 2 Top view of a composite sapphire solder substrate; Figure 3 This is a schematic diagram of the laser beam-induced current detection method of the present invention; Figure 4 This is a schematic diagram of the chip's cross-sectional structure; Figure 5 This is a cross-sectional photograph of the chip after cleavage according to the present invention. Figure 5 (a) is a cross-sectional frontal photograph. Figure 5 (b) is a cross-sectional side view photograph; Figure 6 This is the original image of the laser beam-induced current test of the cross-section of the chip after cleavage in Embodiment 1 of the present invention; Figure 7 for Figure 6 Quantitative response curve, Figure 7 (a) shows the photocurrent response curve in the y-direction of the photosensitive element arrangement. Figure 7 (b) shows the photocurrent response curve of a single photosensitive element along the z-direction of the material. Figure 7 (c) is a combination of the response curves of the 4th, 5th and 6th photosensitive elements along the z-direction; Figure 8 This is a diagram showing the optical signal response of a traditional front-incident LBIC detector. Figure 9 This is an SCM detection image. Figure 9 (a) is the actual test diagram of SCM. Figure 9 (b) is the SCM quantization curve. Detailed Implementation

[0019] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings.

[0020] In the figure: 1. Passivation layer, 2. Diffusion mask layer, 3. N-type InP cap layer, 4. Intrinsic InGaAs absorber layer, 5. N-type InP buffer layer, 6. N-type InP substrate, 7. N electrode, 8. PN junction region, 9. P electrode, 10. Conductive silver paste, 11. Wire, 12. Au layer, 13. Main sapphire wafer, 14. Secondary sapphire wafer.

[0021] The present invention provides a method for detecting the profile laser beam-induced current in the PN junction region of an InGaAs photodetector, comprising: A photodetector chip with an InGaAs absorption layer that has completed the front-side process is diced and cleaved to obtain a smooth profile perpendicular to the PN junction interface. The cleaved chip is fixed to the soldering substrate with the cross-section facing upwards, and the electrical interconnection between the chip's N electrode, P electrode and the corresponding conductive area of ​​the substrate is achieved by conductive silver paste and ultrasonic bonding, respectively. The cross-section was scanned in two dimensions using an LBIC testing system. The laser beam was incident vertically onto the exposed cross-section, and the photocurrent signal of each pixel was collected simultaneously to obtain a two-dimensional distribution map of the photocurrent of the cross-section. Based on the peak position of photocurrent, the precise boundaries of the PN junction in the depth and lateral directions are determined, and the junction depth, lateral diffusion width and uniformity parameters are quantitatively extracted to achieve a profile characterization of the doping distribution and carrier collection efficiency in the PN junction region.

[0022] like Figure 1 , Figure 2 The soldering substrate used in the above method includes: Main sapphire plate (13); The secondary sapphire wafer (14) is cured on the left side of the main sapphire wafer by photoresist, and an Au layer (12) is provided on its upper surface. The upper right surface of the main sapphire wafer (13) is also provided with an Au layer (12); An insulating trench is provided between the Au layer (12) on the upper surface of the secondary sapphire sheet (14) and the Au layer (12) on the upper right side of the main sapphire sheet (13) to prevent the conductive adhesive from overflowing. The Au layer (12) on the upper surface of the secondary sapphire wafer (14) is interconnected with the chip N electrode (7) by conductive silver paste (10). The Au layer (12) on the upper right side of the main sapphire wafer (13) is bonded to the chip P electrode (9) by wire (11) to realize electrical lead-out during LBIC testing.

[0023] like Figure 1 , Figure 2 ,like Figure 5 (a) Figure 5 As shown in (b), after dicing and cleaving the chip, a smooth and flat cross-section is obtained. The chip cross-sectional structure is as follows: Figure 4 As shown, the chip includes: an N electrode (7), an N-type InP substrate (6), an N-type InP buffer layer (5), an InGaAs intrinsic absorption layer (4), an N-type InP cap layer (3), a diffusion mask layer (2), a passivation layer (1), a PN junction region (8), and a P electrode (9).

[0024] Specific usage examples are as follows: Example 1: Profile LBIC Detection Method (1) Cleavage Sample Preparation: Take a 2-inch InGaAs PIN chip, the front side has been completed with Zn diffusion, passivation and electrode processes. Use a scalpel to scribing along the natural cleavage surface of InP to form a regular cleavage surface, and obtain a 5mm×5mm chip. The roughness of the cleavage surface Ra=2.3 nm was measured by atomic force microscopy (AFM), which meets the requirement that polishing is not required.

[0025] (2) Substrate assembly: The main sapphire wafer (13) is 300 μm thick. The sub-sapphire wafer (14) is cured on the left side by photoresist. A 400 nm Au layer (12) is deposited on the surface of the sub-sapphire wafer (14). A 200 μm isolation trench is reserved between the sub-sapphire wafer (13) and the Au layer (12) on the right side. The N electrode (7) on the back of the chip faces down and is aligned with the Au layer (12) of the sub-sapphire wafer (14). 5 nL of conductive silver paste (10) is applied and cured at 80 °C for 10 min. The P electrode (9) on the front side is ultrasonically bonded to the Au layer (12) on the right side of the main sapphire wafer (13) through the wire (11) to achieve electrical lead-out.

[0026] (3) LBIC Scanning: Place the assembled substrate on the LBIC system sample stage with the cross-section facing upwards. Laser beam-induced current detection is as follows: Figure 3 As shown, the laser wavelength is 980nm, the beam diameter is 5μm, the step interval is 1μm, the power is 10μW, and the scanning range is 80μm (depth) × 120μm (lateral), covering a single photosensitive element period. The lock-in amplifier integration time is 30ms, and the noise current is 6pA.

[0027] (4) Data processing: After obtaining the two-dimensional photocurrent matrix, the background dark current is first subtracted; the peak positions are extracted column by column along the depth direction, a histogram of junction depth distribution is plotted, and the lateral diffusion curve is obtained by integrating along the horizontal direction; such as Figure 6 The image shown is the original image of the laser beam-induced current test on the cross-section of the chip after cleavage. The bright areas in the image correspond to regions with strong photocurrent, i.e., locations near the PN junction where carrier collection efficiency is high. Through analysis of... Figure 6 By performing line-by-line scanning and data analysis of the photocurrent distribution, features such as... can be extracted. Figure 7 The quantitative response curve shown is shown. Figure 7 for Figure 6 Quantitative response curve, Figure 7 (a) shows the photocurrent response curve in the y-direction of the photosensitive element arrangement, reflecting the response consistency among different photosensitive elements; Figure 7 (b) shows the photocurrent response curve of a single photoelement along the z-direction of the material, which demonstrates the trend of photocurrent variation with depth and can be used to determine the longitudinal position of the PN junction and the carrier collection range. Figure 7 (c) is a combination diagram of the response curves of the 4th, 5th and 6th photosensitive elements along the z direction, which is convenient for comparing the response width and peak position of different photosensitive elements, thereby evaluating the uniformity of the device process.

[0028] SCM actual measurement as follows Figure 9 As shown in (a), the extracted data SCM quantization curve is as follows: Figure 9(b) shows the results. The test results are compared with those from SCM characterization. SCM focuses on static structural features, mapping the physical diffusion front of impurity ions in the longitudinal direction using high-resolution capacitance signals, such as accurately calibrating the diffusion depth of 1.2-1.4 μm within a 5.0 μm scale. In contrast, profiled LBIC focuses on dynamic functionalization verification. It not only reflects the spatial position of the junction but also directly confirms whether the junction possesses efficient carrier separation capabilities through the intensity of the photocurrent, thus transforming the "doping depth" at the process level into the "photosensitive depth" at the electrical performance level. Secondly, in terms of parameter extraction depth, SCM has an advantage in determining the abrupt changes in heterojunctions due to its nanometer-scale spatial resolution. The unique value of profiled LBIC lies in its ability to extract the effective minority carrier diffusion length through mathematical fitting of the signal decay curve. This quantitative assessment of carrier transport characteristics allows LBIC to diagnose electrical failures that SCM cannot detect, such as excessively fast interfacial recombination rates or shortened lifetimes caused by lattice defects within the absorption layer. To further illustrate the advantages of this invention, its test results can be compared and analyzed with those of traditional frontal incident LBIC characterization. For example... Figure 8 As shown, in the traditional front-illumination mode, the laser is incident from the surface of the device's passivation layer. The periodic peaks in the scan data clearly demonstrate that the signal reaches saturation values ​​in the ranges of 125–150 μm and 163–196 μm, proving that each photosensitive unit has extremely high and consistent photoelectric conversion efficiency. Simultaneously, the response signals between units significantly drop back to the reference level at positions such as 115 μm and 156 μm, forming distinct isolation regions. This directly reflects the device's excellent inter-pixel electrical isolation and low crosstalk characteristics, which can be used to evaluate the macroscopic uniformity of the detector's lateral mask and electrode fabrication processes. However, compared to front-illumination, which focuses on macroscopic screening of lateral parameters, the cross-sectional illumination LBIC used in this invention achieves a "tomographic scan" audit of the device's longitudinal layer structure. Its core advantage lies in its ability to accurately locate the InP / InGaAs heterointerface at positions such as 17.05 μm and intuitively calibrate the precise junction depth of each photosensitive element, thereby overcoming the inherent limitation of the front-illumination mode in resolving deep doping diffusion information. The combination of these two approaches constitutes a complete process diagnostic system, encompassing both horizontal and vertical dimensions, and from surface to bulk region. Laser beam-induced current detection optical signal response photosensitive element arrangement in the y-direction... Figure 7 As shown in (a), the minority carrier lifetime τ = 26 μs was calculated, indicating a low density of deep-level defects and non-radiative recombination centers in the material, suggesting good crystal quality in the epitaxial layer. The matching long diffusion length LD ≈ 17.7 μm indicates that photogenerated carriers can achieve long-distance diffusion transport before recombination, providing an effective basis for carrier collection in the device. The photosensitive elements are numbered, and analysis is performed on a single photosensitive element. The laser beam-induced current detection optical signal response along the material depth z direction is shown below. Figure 7As shown in (b), to estimate the diffusion depth of Zn within the absorption layer, this study performed multi-factor corrections on the LBIC measurement results. First, the total thickness of the surface window layer composed of the SiO2 passivation layer and the InP cap layer was deducted, limiting the junction depth analysis to the interior of the absorption layer. Second, the spatial displacement from the initial rise point of the signal to the peak position was corrected, considering the effects of the laser spot radius and the minority carrier diffusion length on response broadening, to reduce systematic biases introduced by optical and carrier transport processes. Furthermore, given that the doping gradient between the highly doped InP buffer layer and the lightly doped absorption layer may form a back surface field (BSF), the potential influence of its built-in electric field on carrier reflection and collection positions was also incorporated into the correction of the junction depth estimation. Taking into account surface structure, spot size, minority carrier diffusion characteristics, and BSF-related effects, the estimated physical junction depth of Zn within the absorption layer is approximately 1.24 μm. This result is consistent with that obtained by scanning capacitance microscopy (SCM). Figure 9 (b) The metallurgical junction depth of 1.235 μm in the SCM data diagram is consistent. Then, the required photosensitive data is spliced ​​together. Taking the 4th, 5th, and 6th photosensitive elements as an example, the following is obtained: Figure 7 (c) When the response width drops to 1 / e, it represents the response width, which is used to analyze the peak value and the response width of each photosensitive element. Table 1 shows a comparison of photosensitive element responses. The response characteristic data indicates that the peak responsivity of the three units exhibits good consistency, with a relative fluctuation deviation of only 1.6%. This result confirms that the device fabrication process has good stability and that the performance differences between units are small. Specifically, the response widths of the fourth and sixth units are 13.39 μm and 12.87 μm, respectively. These values ​​are similar, indicating that the carrier transport paths and effective response regions of these two units are basically the same. Although some parameters of the fifth unit show slight deviations, the signals of all units are highly similar in magnitude and variation pattern. This characteristic strongly suggests that the execution of each process step in the fabrication of this device is relatively standardized, and the overall process consistency is good.

[0029] Table 1 Unit Number Responsivity (peak current) Response width (Δx) Unit 4 9515.29 13.39μm Unit 5 9652.14 15.49μm Unit 6 9214.14 12.87μm

[0030] This invention provides a direct comparison of the consistency and differences in junction depth among different photosensitive elements, thereby evaluating the uniformity and stability of device fabrication. This method fully utilizes the excellent mechanical properties and highly controllable natural cleavage characteristics of InGaAs / InP crystals, achieving wafer profiles with extremely high surface flatness through simple cleavage processes. Due to this material characteristic, the sample preparation process is significantly simplified, eliminating the need for complex mechanical polishing and other treatments, thus providing a reliable foundation for rapidly and accurately characterizing junction depth and lateral diffusion behavior.

[0031] Example 2: Verification of the reusability of the solder conductive substrate After completing the above tests, the chip and substrate were immersed in acetone to dissolve the photoresist, and the chip and substrate were detached together. The wires (11) were manually cut off. After ultrasonic cleaning with isopropanol and drying with nitrogen, there was no residual adhesive on the substrate surface. The new chip was assembled 5 times, and the bonding pull force test was still greater than 5g, which met the requirements for repeated use.

[0032] Example 3: Comparison of different laser powers Keeping other parameters constant, the difference from Example 1 is that the laser power was set to 1 μW, 50 μW, and 100 μW respectively. The results showed that at 1 μW, the signal-to-noise ratio was low, with a peak identification error of ±0.1 μm; the results for 50 μW and 10 μW were consistent; at 100 μW, slight thermal broadening occurred, with the lateral diffusion width increasing by 0.2 μm. Therefore, this invention uses a power of 10–50 μW.

Claims

1. A method for detecting laser beam-induced current in the PN junction region profile of an InGaAs photodetector, characterized in that, Includes the following steps: A photodetector chip with an InGaAs absorption layer that has completed the front-side process is diced and cleaved to obtain a smooth profile perpendicular to the PN junction interface. The cleaved chip is fixed to the soldering substrate with the cross-section facing upward, and the electrical interconnection between the chip's N electrode, P electrode and the corresponding conductive area of ​​the substrate is achieved by conductive silver paste and ultrasonic bonding, respectively. The cross-section was scanned in two dimensions using an LBIC testing system. The laser beam was incident vertically onto the exposed cross-section, and the photocurrent signal of each pixel was collected simultaneously to obtain a two-dimensional distribution map of the photocurrent of the cross-section. Based on the peak position of photocurrent, the precise boundaries of the PN junction in the depth and lateral directions are determined, and the junction depth, lateral diffusion width and uniformity parameters are quantitatively extracted to achieve a profile characterization of the doping distribution and carrier collection efficiency in the PN junction region.

2. The method according to claim 1, characterized in that, The cleavage occurs along the natural cleavage plane of the InP-based material, and the roughness of the cleavage surface Ra ≤ 5 nm.

3. The method according to claim 1, characterized in that, The conductive substrate is a composite sapphire structure, including a main sapphire sheet and a secondary sapphire sheet fixed to the left side of the main sapphire sheet by photoresist. The upper surface of the secondary sapphire sheet is provided with an Au layer and is connected to the N electrode of the chip by conductive silver paste. The upper surface of the right side of the main sapphire sheet is provided with an Au layer and is connected to the P electrode of the chip by a wire. There is an isolation gap between the two Au layers to prevent silver paste overflow and short circuit.

4. The method according to claim 3, characterized in that, The Au layer has a thickness of 300nm–500nm and an isolation spacing width of 100μm–300μm.

5. The method according to claim 1, characterized in that, The LBIC scanning parameters are: laser wavelength 900nm–1100nm, beam spot diameter ≤5μm, step interval 0.1μm–1μm, incident power 1μW–100μW, and scanning range covering the entire profile depth and at least one complete photosensitive element period.

6. The method according to claim 1, characterized in that, Also includes: The two-dimensional photocurrent distribution map obtained by scanning is integrated along the depth direction to obtain the lateral diffusion curve, and the lateral diffusion width is extracted by the full width at half maximum (FWHM) method. Connect the peak photocurrent positions of multiple pixels in the same photosensitive element array direction to evaluate junction depth consistency.

7. The method according to claim 1, characterized in that, After the test is completed, the entire conductive substrate is immersed in acetone to dissolve the photoresist, so that the chip can be separated from the substrate without damage. The substrate can be reused after cleaning.

8. The method according to any one of claims 1 to 7, characterized in that, The InGaAs photodetector is a planar PIN structure, which, from bottom to top, includes an N-type InP substrate, an N-type InP buffer layer, an intrinsic InGaAs absorber layer, an N-type InP cap layer, a passivation layer, and a P electrode.

9. A conductive substrate for use in the method of claim 1, characterized in that, include: The main sapphire wafer has an Au layer on its upper right surface. The secondary sapphire wafer is cured onto the left side of the primary sapphire wafer using photoresist, and its upper surface has an Au layer. An insulating trench is provided between the Au layer on the upper right side of the main sapphire sheet and the Au layer on the upper surface of the secondary sapphire sheet to prevent the conductive adhesive from overflowing. The Au layer on the upper surface of the secondary sapphire wafer is used to interconnect with the N electrode of the chip using conductive silver paste, and the Au layer on the upper right side of the primary sapphire wafer is used to bond with the P electrode of the chip via wires to achieve electrical lead-out during LBIC testing.

10. A LBIC detection system for the PN junction region profile of an InGaAs photodetector, characterized in that, include: The conductive substrate as described in claim 9; The LBIC scanning unit is used to output a focusable laser with a wavelength of 900nm–1100nm and perform two-dimensional scanning of the cross-section of a chip fixed to a soldering substrate. The current detection unit is connected to the positive and negative terminals of the conductive soldering substrate and is used to synchronously acquire photocurrent signals. The data processing unit is used to convert the photocurrent signal into a two-dimensional distribution map and extract parameters such as junction depth, lateral diffusion width, and uniformity.

Images