Method for measuring medium-deep energy level recombination center of silicon
By forming a lightly doped epitaxial layer on a heavily doped substrate and preparing a Schottky junction, the problems of weak transient spectral signals and large background capacitance in deep energy levels were solved, enabling highly sensitive and accurate quantitative measurement of deep energy recombination centers in heavily doped substrates, thus improving the accuracy and signal-to-noise ratio of the measurement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-15
- Publication Date
- 2026-03-27
AI Technical Summary
In existing technologies, the direct measurement method of deep-level transient spectrum is difficult to detect due to weak signals and huge background capacitance on heavily doped substrates, resulting in a very poor signal-to-noise ratio. The carrier tunneling effect causes spectral distortion, making it impossible to effectively measure deep-level recombination centers in heavily doped semiconductor materials.
By forming an epitaxial layer on a heavily doped substrate with a doping concentration lower than that of the substrate, deep-level recombination centers are ensured to diffuse into the epitaxial layer. A Schottky junction is then fabricated, and the deep-level transient spectrum of the Schottky junction is measured to obtain the energy level, concentration, and trapping cross section of the deep-level recombination centers.
This technology enables highly sensitive and accurate quantitative electrical characterization of deep recombination centers in heavily doped substrates, overcoming the measurement limitations of DLTS technology, improving the signal-to-noise ratio, and ensuring the accuracy and sensitivity of measurements.
Smart Images
Figure CN121752038A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of integrated circuit technology, and in particular to a method for measuring deep recombination centers in silicon. Background Technology
[0002] Deep-level transient spectroscopy (DLTS) is an important technique for characterizing defects in semiconductor materials and devices. This method offers advantages such as high sensitivity, quantitative analysis, and the ability to simultaneously obtain multiple parameters including defect energy levels, concentrations, and trapping cross-sections. It is widely used to assess defects introduced into semiconductor materials, such as purity and process-induced damage.
[0003] In semiconductor manufacturing, heavily doped substrates are commonly used to fabricate various epitaxial devices. Even at extremely low concentrations (~E10 cm⁻¹), the metallic impurities in these substrates... -3 These trace metal impurities can also act as deep-level centers, becoming recombination centers or trap centers for charge carriers, significantly reducing minority carrier lifetime, exacerbating device leakage, and leading to device performance degradation or even failure. Therefore, controlling trace metal impurities in heavily doped substrates is crucial.
[0004] Existing technologies employ VPD (Vapor Phase Decomposition) to detect metallic impurities in semiconductor silicon wafers. After collecting metals from the silicon wafer surface using HF-H₂O₂, the collected solution is sent to ICP-MS for metal content analysis. This method involves chemical collection followed by mass spectrometry measurement. The collection process requires stringent control of the external environment to ensure that the measured metal content originates from the silicon wafer rather than being introduced from the environment. Controlling external contamination necessitates complete surface collection.
[0005] Existing DLTS direct measurement methods involve fabricating Schottky junctions or PN junctions directly on heavily doped substrates and performing DLTS measurements. However, the extremely high carrier concentration of heavily doped substrates results in very narrow depletion layer widths (<100 nm) and huge background capacitances. The DLTS signal intensity (ΔC / C) is directly proportional to the depletion layer width and inversely proportional to the doping concentration, leading to a drastic deterioration in the effective signal-to-noise ratio and making the signal weak and difficult to detect. Furthermore, the strong electric field in heavily doped materials induces significant carrier tunneling, which masks the thermal emission process and severely distorts the DLTS spectrum. Therefore, standard DLTS techniques are inherently unsuitable for directly measuring heavily doped semiconductor materials. Summary of the Invention
[0006] The purpose of this invention is to provide a method for measuring deep-level recombination centers in silicon, in order to solve at least one of the problems in the direct measurement method of deep-level transient spectrum: huge background capacitance, weak signal that is difficult to detect, and severe distortion of deep-level transient spectrum caused by carrier tunneling effect.
[0007] To solve the above-mentioned technical problems, the present invention provides a method for measuring deep recombination centers in silicon, comprising:
[0008] A substrate is provided, the substrate comprising doped ions and deep-level recombination centers;
[0009] An epitaxial layer is formed on the substrate, the epitaxial layer and the substrate have the same doping type and the doping concentration of the epitaxial layer is less than that of the substrate, and the deep recombination centers in the substrate are completely diffused into the epitaxial layer;
[0010] Preparation of Schottky junctions;
[0011] The deep-level transient spectrum of the Schottky junction is measured to simultaneously obtain the energy level, concentration, and trapping cross section of the deep-level recombination centers in the substrate.
[0012] Optionally, the doping concentration of the substrate is greater than E18 atoms / cm3.
[0013] Optionally, the doping concentration of the epitaxial layer is E14 atoms / cm3 to E17 atoms / cm3.
[0014] Optionally, the deep-level recombination center includes deep-level metallic impurities and / or deep-level defects.
[0015] Optionally, the epitaxial layer is formed using an epitaxial growth process.
[0016] Optionally, the process temperature and process time of the epitaxial growth process are set according to the diffusion equation of the deep-level recombination center to be tested, so as to ensure that the deep-level recombination center diffuses from the substrate into the epitaxial layer.
[0017] Optionally, the process temperature of the epitaxial growth process is 500℃~1200℃.
[0018] Optionally, the step of preparing the Schottky junction includes:
[0019] Cut the sample to the target size;
[0020] A vapor deposition process is performed to fabricate a front electrode on the epitaxial layer;
[0021] A back electrode is applied to the back of the sample to make contact between the back of the sample and the sample stage.
[0022] Optionally, the material of the back electrode includes indium, which fills the microscopic irregularities of the back of the sample and the sample stage surface through plastic deformation to form a conformal contact.
[0023] Optionally, the epitaxial layer and the substrate have the same doping type, which is P-type, or the epitaxial layer and the substrate have the same doping type, which is N-type.
[0024] In a method for measuring deep-level recombination centers in silicon provided by this invention, a substrate is provided, comprising doped ions and deep-level recombination centers; an epitaxial layer is formed on the substrate, the epitaxial layer and the substrate have the same doping type and the doping concentration of the epitaxial layer is less than that of the substrate, and the deep-level recombination centers in the substrate are completely diffused into the epitaxial layer; a Schottky junction is fabricated; and the deep-level transient spectrum of the Schottky junction is measured to simultaneously obtain the energy level, concentration, and trapping cross-section of the deep-level recombination centers in the substrate. This invention overcomes the limitation of DLTS technology in directly measuring heavily doped substrate materials. By designing and growing a lightly doped epitaxial layer, the deep-level recombination centers in the heavily doped substrate are fully introduced into the lightly doped epitaxial layer, thereby achieving highly sensitive and accurate quantitative electrical characterization of trace deep-level recombination centers in the heavily doped substrate. Attached Figure Description
[0025] Figure 1 This is a flowchart of a method for measuring deep recombination centers in silicon according to an embodiment of the present invention.
[0026] Figures 2 to 4 This is a schematic diagram of the structural steps corresponding to the measurement method of deep recombination centers in silicon according to an embodiment of the present invention.
[0027] Figures 5 to 10 These are the deep energy level spectra of the Schottky junction at different frequencies in the experimental group of this invention.
[0028] Figure 11 for Figures 5 to 10 A composite diagram of the deep energy level spectra of the Schottky junction at different frequencies for the experimental group.
[0029] Figure 12 This is the deep energy level spectrum of the Schottky junction of the control group at a frequency of 300 Hz in this embodiment of the invention.
[0030] In the picture,
[0031] 10-Substrate; 10a-Deep level recombination center; 11-Epipolar layer; 12-Front electrode; 13-Back electrode. Detailed Implementation
[0032] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a method for measuring deep recombination centers in silicon proposed in this invention. The advantages and features of this invention will become clearer from the following description. It should be noted that the drawings are all in a very simplified form and use non-precise scales, used only to facilitate and clarify the illustration of the embodiments of this invention. Furthermore, the structures shown in the drawings are often part of the actual structures. In particular, different figures may emphasize different aspects and sometimes use different scales.
[0033] Figure 1 This is a flowchart of a method for measuring deep recombination centers in silicon according to an embodiment of the present invention. Figure 1 As shown, this embodiment provides a method for measuring deep recombination centers in silicon, including:
[0034] Step S10: Provide a substrate, wherein the substrate includes doped ions and deep-level recombination centers;
[0035] Step S20: Form an epitaxial layer located on the substrate, wherein the epitaxial layer and the substrate have the same doping type and the doping concentration of the epitaxial layer is less than that of the substrate, and the deep recombination centers in the substrate are completely diffused into the epitaxial layer.
[0036] Step S30: Prepare a Schottky junction;
[0037] Step S40: Measure the deep-level transient spectrum of the Schottky junction to simultaneously obtain the energy level, concentration, and trapping cross section of the deep-level recombination center in the substrate.
[0038] Figures 2 to 4 This is a schematic diagram of the corresponding steps in the method for measuring deep recombination centers in silicon according to an embodiment of the present invention; to make the above-mentioned objectives, features and beneficial effects of the present invention more apparent and understandable, the following description is provided in conjunction with the appendix to the specification. Figures 2 to 4 Specific embodiments of the present invention will be described in detail below.
[0039] like Figure 2 As shown, a substrate 10 is provided. The substrate 10 provides an operating platform for subsequent processes. It can be any substrate known to those skilled in the art for carrying semiconductor integrated circuit components, and can be a bare die. Specifically, the substrate is, for example, a silicon-on-insulator (SOI) substrate, a bulk silicon substrate, a germanium substrate, a germanium-silicon substrate, an indium phosphide (InP) substrate, a gallium arsenide (GaAs) substrate, or a germanium-on-insulator substrate, etc. In this embodiment, the substrate 10 is a silicon substrate.
[0040] Please continue to refer to this. Figure 2The substrate 10 includes doped ions and deep-level recombination centers 10a. The deep-level recombination centers 10a include deep-level metallic impurities and / or deep-level defects. In this embodiment, the deep-level recombination centers 10a are, for example, deep-level metallic impurities. The deep-level metallic impurities are elements from the third period or later in the periodic table, such as nickel, iron, cobalt, copper, and tungsten. The substrate 10 is a heavily doped substrate, and the doping concentration of the substrate is greater than E18 atoms / cm². 3 .
[0041] like Figure 3 As shown, an epitaxial layer 11 is formed on the substrate 10. The epitaxial layer 11 and the substrate 10 have the same doping type, but the doping concentration of the epitaxial layer 11 is lower than that of the substrate 10. Deep recombination centers 10a in the substrate 10 are completely diffused into the epitaxial layer 11. The epitaxial layer 11 is formed using an epitaxial growth process. The doping concentration of the epitaxial layer 11 is, for example, E14 atoms / cm². 3 ~E17 atoms / cm 3 Based on the diffusion equation of the deep-level recombination center 10a to be tested, the process temperature and time of the epitaxial growth process are set to ensure that the deep-level recombination center 10a diffuses from the substrate 10 to the epitaxial layer 11. The process temperature of the epitaxial growth process is 500℃ to 1200℃. The process time of the epitaxial growth process is generally a few seconds to several hours. In this embodiment, the epitaxial layer 11 and the substrate 10 have the same doping type and are P-type, that is, the substrate 10 is P+ type doped and the epitaxial layer 11 is P- type doped, that is, lightly doped.
[0042] In another embodiment, the epitaxial layer 11 and the substrate 10 are doped with the same type, specifically N-type. That is, the substrate 10 is N+ type doped, and the epitaxial layer 11 is N- type doped, i.e., lightly doped.
[0043] like Figure 4As shown, a Schottky junction is fabricated. The steps for fabricating the Schottky junction include: cutting the sample to a target size, for example, 1 cm * 1.5 cm; placing it in a photomask; and then placing it in a deposition machine for vapor deposition to fabricate a front electrode 12 on the epitaxial layer 11. The front electrode 12 is, for example, a titanium electrode, and the diameter of the front electrode 12 is, for example, 2 mm * 2 mm. A back electrode 13 is coated on the back side of the sample to ensure contact between the back side of the sample and the sample stage, guaranteeing sufficient temperature conduction during the heating and cooling of the sample and the sample stage. The material of the back electrode 13 is, for example, indium. Indium, as a soft metal, can fill the microscopic irregularities on the back side of the sample and the surface of the sample stage through plastic deformation, forming a large-area, low-resistance conformal contact, thereby significantly improving thermal and electrical conduction efficiency.
[0044] The deep-level transient spectrum of the Schottky junction was measured using a semilab DLS-1100 to simultaneously obtain the energy level, concentration, and trapping cross-section of the deep-level recombination centers in the substrate 10. In this embodiment, the energy level, concentration, and trapping cross-section of the metal impurities in the substrate 10 were also simultaneously obtained.
[0045] In one example, an experimental group and a control group are set up. The size of the substrate 10 in the experimental group is, for example, 12 inches, the doping type of the substrate 10 in the experimental group is, for example, P+ type, and the doping concentration of the substrate 10 in the experimental group is, for example, 1E19 atoms / cm³. 3 The resistivity of the substrate 10 in the experimental group is, for example, 0.01 ohmcm to 0.02 ohmcm. The deep-level recombination centers 10a of the substrate 10 in the experimental group are metallic impurities. Preferably, the deep-level recombination centers 10a of the substrate 10 in the experimental group are, for example, Ni. The parameters of the substrate in the control group are the same as those in the experimental group, except that the substrate in the control group does not contain deep-level recombination centers; here, "not containing" refers to a measurement within the allowable tolerance range of the process.
[0046] According to the formula for the diffusion coefficient D:
[0047] D=D0*exp (1);
[0048] Where Ea is the diffusion activation energy, k is the gas constant, T is the diffusion temperature, and D0 is the pre-exponential factor.
[0049] The formula for diffusion length DL is:
[0050] diffusion length DL= (2);
[0051] Where D is the diffusion coefficient and t is the diffusion time.
[0052] The EPI (epitaxy growth process) for the experimental group was designed with a process temperature of 1100℃ and a process time of 3 minutes. Substituting the diffusion coefficient of Ni, it was found that Ni can diffuse almost completely from the substrate 10 to the epitaxial layer 11. It is worth emphasizing that "almost completely" here refers to the measurement within the allowable error range of the process.
[0053] Table 1, Diffusion coefficient of Ni D0 (cm2 / s) Ea (eV) kB (eV / K) T1 (C) T1 (K) <![CDATA[D1 (cm 2 / s)]]> t1 (s) <![CDATA[D1*t1 (cm 2 )]]> Distance diffused (um) 2.00E-03 0.47 8.62E-05 1100 1373.15 3.77E-05 180 6.79E-03 823.97
[0054] Table 1 shows the diffusion coefficients of Ni, where D0 is the pre-exponential factor, Ea is the diffusion activation energy, kB is the Boltzmann constant, T1 is the diffusion temperature (Celsius and absolute temperature), D1 is the diffusion coefficient, t1 is the diffusion time, and diffusion length is the diffusion length. It can be seen that Ni-related recombination centers can completely diffuse from the substrate 10 into the epitaxial layer within this temperature range.
[0055] Epitaxial layer 11 was formed on substrate 10 of the experimental group and on substrate of the control group, respectively. Schottky junction samples were fabricated, and the deep-level transient spectra of the Schottky junctions in the experimental group and the control group were measured using a semilab DLS-1100.
[0056] Figures 5 to 10 These are the deep energy level spectra of the Schottky junction at different frequencies in the experimental group of this invention. Figure 11 for Figures 5 to 10 A composite diagram of the deep energy level spectra of the Schottky junction at different frequencies for the experimental group. Figure 12 This is the deep energy level spectrum of the Schottky junction of the control group at a frequency of 300 Hz in this embodiment of the invention. Figures 5 to 11 As shown, Figures 5 to 10 The horizontal axis represents temperature in Kelvin (K); the vertical axis represents Ntrap, a parameter in semiconductors describing the density of trapped states, which is the number of defect states per unit volume that can trap electrons or holes, measured in units of # / cm². 3 . Figure 11 The horizontal axis represents the reciprocal of temperature, in units of 1 / K; the vertical axis represents the emission rate, ln(e / T). 2 ). Figure 12 The horizontal axis represents temperature in Kelvin (K); the vertical axis represents temperature in Ntrap (°C / cm). 3From the DLTS spectrum, the amplitude of the spectral peak (ΔC / C0) is directly proportional to the defect concentration Nt. By repeatedly performing temperature scans with varying rate windows (using different (t1, t2) pairs), the same defect will exhibit peaks at different temperatures. For each rate window, the temperature T corresponding to the peak and its corresponding emission rate e_n (determined by the rate window) are recorded. Plotting ln(e_n / T²) against 1 / T yields a straight line (Arrhenius plot). The slope of this line can be used to calculate the defect energy level depth ΔE, and the intercept yields the trapping cross-section. Therefore, Figure 11 The deep energy level spectra of the Schottky junction obtained from the experimental group at different frequencies all exhibit two peaks, namely a first peak P1 and a second peak P2. The temperature of the first peak P1 is 260K to 280K, and the temperature of the second peak P2 is 190K to 200K. The energy level of the first peak P1 is +0.55 eV, and the energy level of the second peak P2 is +0.38 eV. The trapping cross section of the first peak P1 is 2.1E-13cm. 2 The capture cross section of the second peak P2 is 1.5E-13cm. 2 The concentration of the first peak P1 was 2.15 E12# / cm³. 3 ~2.29E12 # / cm 3 The concentration of P2 at the second peak was 3.57 E11 # / cm³. 3 ~4.55E11 # / cm 3 It is worth emphasizing that in DLTS spectra, the amplitude (ΔC) of the spectral peak is proportional to the concentration NT of that deep level. The concentration can be quantitatively calculated using the formula NT ≈ 2 * (ΔC / C0) * ND (for majority traps), where ΔC is the peak amplitude, C0 is the steady-state capacitance, and ND is the background doping concentration. Figures 5 to 10 In this context, concentration is calculated as peak area relative to the baseline. For example... Figure 12 As shown, the deep-level spectrum of the Schottky junction in the control group did not show a significant deep-level peak, indicating that the metal impurities in the epitaxial layer 11 originate from the substrate 10, rather than from the formation process of the epitaxial layer 11. Therefore, the measurement method for deep-level recombination centers in silicon provided in this embodiment solves the signal-to-noise ratio problem of directly using DLTS to measure heavily doped substrates, and realizes the detection of electrically active metal impurities; it has high sensitivity and accurate quantification. Compared with chemical methods, it provides an electrical method for measuring the concentration of deep-level metals. This method is not only applicable to measuring deep-level metals, but also to measuring deep-level defects.
[0057] In summary, the method for measuring deep-level recombination centers in silicon provided in this invention involves providing a substrate containing doped ions and deep-level recombination centers; forming an epitaxial layer on the substrate, wherein the epitaxial layer and the substrate have the same doping type and the doping concentration of the epitaxial layer is lower than that of the substrate, and the deep-level recombination centers in the substrate are completely diffused into the epitaxial layer; fabricating a Schottky junction; and measuring the deep-level transient spectrum of the Schottky junction to simultaneously obtain the energy level, concentration, and trapping cross-section of the deep-level recombination centers in the substrate. This invention overcomes the limitation of DLTS technology in directly measuring heavily doped substrate materials. By designing and growing a lightly doped epitaxial layer, it fully introduces the deep-level recombination centers from the heavily doped substrate into the lightly doped epitaxial layer, thereby achieving highly sensitive and accurate quantitative electrical characterization of trace deep-level recombination centers in the heavily doped substrate. It solves the signal-to-noise ratio problem of directly using DLTS to measure heavily doped substrates, and realizes the detection of electrically active metal impurities; it offers high sensitivity and accurate quantification.
[0058] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to mutually. In addition, different parts between embodiments can also be combined with each other, and this invention does not limit this.
[0059] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the present invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure shall fall within the protection scope of the claims.
Claims
1. A method for measuring deep recombination centers in silicon, characterized in that, include: A substrate is provided, the substrate comprising doped ions and deep-level recombination centers; An epitaxial layer is formed on the substrate, the epitaxial layer and the substrate have the same doping type and the doping concentration of the epitaxial layer is less than that of the substrate, and the deep recombination centers in the substrate are completely diffused into the epitaxial layer; Preparation of Schottky junctions; The deep-level transient spectrum of the Schottky junction is measured to simultaneously obtain the energy level, concentration, and trapping cross section of the deep-level recombination centers in the substrate.
2. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The substrate has a doping concentration greater than E18 atoms / cm². 3 .
3. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The doping concentration of the epitaxial layer is E14 atoms / cm². 3 ~E17 atoms / cm 3 .
4. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The deep-level recombination center includes deep-level metallic impurities and / or deep-level defects.
5. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The epitaxial layer is formed using an epitaxial growth process.
6. The method for measuring deep recombination centers in silicon as described in claim 5, characterized in that, Based on the diffusion equation of the deep-level recombination center to be tested, the process temperature and process time of the epitaxial growth process are set to ensure that the deep-level recombination center diffuses from the substrate into the epitaxial layer.
7. The method for measuring deep recombination centers in silicon as described in claim 5, characterized in that, The process temperature for the epitaxial growth process is 500℃ ~ 1200℃.
8. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The steps for preparing the Schottky junction include: Cut the sample to the target size; A vapor deposition process is performed to fabricate a front electrode on the epitaxial layer; A back electrode is applied to the back of the sample to make contact between the back of the sample and the sample stage.
9. The method for measuring deep recombination centers in silicon as described in claim 8, characterized in that, The material of the back electrode includes indium, which fills the microscopic irregularities on the back of the sample and the sample stage surface through plastic deformation to form a conformal contact.
10. The method for measuring deep recombination centers in silicon as described in claim 1, characterized in that, The epitaxial layer and the substrate have the same doping type, which is P-type, and the epitaxial layer and the substrate have the same doping type, which is N-type.