A nondestructive testing method for rapidly judging material damage and removal

Abstract

The application discloses a kind of nondestructive testing methods for quickly judging material damage and removal, it is related to the field of semiconductor manufacturing technology, comprising the following steps;S01, provide the semiconductor material to be measured, the sample at least includes processed area, or / and processed surface;S02, using ellipsometer measurement The sample obtains the ellipsometric raw data of the sample under preset incident angle, and the curve of pseudo dielectric constant with photon energy is calculated accordingly, the pseudo dielectric constant includes real part, and / or imaginary part;S03, the curve of pseudo dielectric constant with photon energy is mathematically derived, and the derivative spectrum is obtained.The application can realize the nondestructive testing of quickly judging material damage and removal, avoids complex physical modeling, avoids tedious data fitting, avoids clutter component dependence, and can realize fast judgment only by the original ellipsometric data of test.

Description

Technical Field

[0001] This invention relates to the field of semiconductor manufacturing technology, specifically to a detection method that can quickly and non-destructively determine whether damage exists in mercury cadmium telluride materials during processing and whether the damage has been removed. Background Technology

[0002] Mercury cadmium telluride (MCH), a typical representative of group II-VI semiconductor materials, has a bandgap that can continuously vary from -0.3 eV to 1.6 eV depending on the cadmium composition (x), essentially covering the entire infrared detection band. This has led to its widespread application in infrared detectors in aerospace, infrared remote sensing, and other fields related to national economic and defense security, making it one of the most important semiconductor materials after silicon and gallium arsenide. However, this material faces severe challenges in processing. At room temperature, the Vickers hardness of crystalline silicon and gallium arsenide is approximately 12 GPa and 6.8 GPa, respectively, while the hardness of MCH is only about 0.5 GPa, and its brittleness is significantly higher than that of silicon. This soft-brittle characteristic makes MCH an extremely difficult semiconductor material to process. Therefore, after the processing steps are completed, it is essential to test the crystal quality of MCH.

[0003] In crystal quality assessment methods, X-ray diffraction (XRD), especially the rocking curves obtained by high-precision double-crystal diffraction, is a widely used standard technique. However, this technique has a significant efficiency bottleneck. XRD testing requires precise calibration of the equipment's optical path to ensure data reliability, and each sample needs similar calibration before testing, with the fine scan range adjusted based on the coarse scan results. Since the scan step and range set for the fine scan directly determine the testing time (the smaller the step and the larger the range, the longer the time required), the testing efficiency of XRD largely limits the scale of batch testing.

[0004] Elliptic polarization, a highly sensitive and non-destructive optical characterization technique, enables in-situ or ex-situ monitoring of material surfaces and thin films, particularly suitable for analyzing the physical and chemical properties of semiconductor materials, achieving a resolution of 0.1 nanometers for thickness variations. However, its application in the analysis of mercury cadmium telluride (HCdT) materials faces several challenges. First, HCdT is easily oxidized when exposed to air. Without treatment under a protective atmosphere such as nitrogen, the presence of a surface oxide layer necessitates an analytical model that includes at least three layers: an air layer, a surface oxide layer, and a HCdT substrate. This increased number of unknown parameters significantly complicates the analysis. Second, the critical point (CP) of uncontaminated pure HCdT materials varies with the cadmium component x, but different research teams have summarized the relationship between CP and composition (all expressed as E=a+bx+cx). 2Significant differences exist between these parameters, and the lack of a unified consensus makes it difficult to reduce the number of unknown parameters in the model using compositional information. Furthermore, considering the variation of composition with thickness, elliptic polarization data analysis of mercury cadmium telluride materials under nitrogen-free conditions becomes extremely complex. Summary of the Invention

[0005] The purpose of this invention is to provide a non-destructive testing method for rapidly determining and removing material damage, thereby overcoming the aforementioned shortcomings in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a non-destructive testing method for rapidly determining and removing material damage, comprising the following steps;

[0007] S01. Provide a semiconductor material to be tested, wherein the sample includes at least a processed region and / or a processed surface;

[0008] S02. The sample is measured using an ellipsometry to obtain the original ellipsometry data of the sample at a preset incident angle, and the curve of the pseudo dielectric constant as a function of photon energy is calculated accordingly. The pseudo dielectric constant includes a real part and / or an imaginary part.

[0009] S03. The curve of the pseudo-dielectric constant changing with photon energy is mathematically differentiated to obtain the derivative spectrum;

[0010] S04. Extract at least one feature value corresponding to the material's critical energy CP from the derivative spectrum, wherein the feature value includes one or more of the following: E1 feature value, E1+Δ1 feature value, and E2 feature value;

[0011] S05. The extracted feature values ​​are compared with reference feature values, where the reference feature values ​​are the corresponding feature values ​​of the same undamaged material under the same test conditions, and then an evaluation is given:

[0012] If the extracted feature value deviates from the reference feature value, it is determined that the semiconductor material under test has processing damage.

[0013] If the extracted feature value is basically consistent with the reference feature value, it is determined that the semiconductor material under test is undamaged or the damage has been removed.

[0014] Preferably, the mathematical differentiation in step S03 is a first derivative, a second derivative, or a third derivative.

[0015] Preferably, the pseudo-dielectric constant in step S02 is the original pseudo-dielectric constant obtained by direct calculation.

[0016] Preferably, the preset incident angle in step S02 is 45°, and the entire test of the elliptic polarimeter does not require inert gas protection and is conducted directly in the air environment.

[0017] Preferably, the E1 characteristic value, E1+Δ1 characteristic value, and E2 characteristic value in step S04 correspond to the positions of characteristic peaks or valleys in the derivative spectrum caused by the key energy CP point in the band structure of the corresponding semiconductor material sample.

[0018] Preferably, the reference feature in step S05 is derived from any of the following methods:

[0019] The characteristic values ​​were obtained and calculated from an undamaged standard sample of the same material as the semiconductor material to be tested, without processing, and under the same test conditions; or

[0020] The characteristic values ​​are measured and calculated under the same test conditions in the unprocessed or undamaged regions of the semiconductor material under test.

[0021] Preferably, the shift in step S05 includes a blue shift towards a higher energy direction or a red shift towards a lower energy direction;

[0022] Among them, blue shift indicates compressive stress-type damage introduced by machining, while red shift indicates tensile stress-type damage introduced by machining.

[0023] Preferably, the semiconductor material to be tested is a direct bandgap semiconductor material with a significant critical energy (CP) characteristic, which includes at least mercury cadmium telluride.

[0024] As a preferred embodiment, the processing method of the semiconductor material to be tested includes diamond single-point turning, chemical mechanical polishing, chemical etching, or epitaxial growth.

[0025] The methods for removing the damage include diamond single-point turning, chemical mechanical polishing, or chemical etching.

[0026] Preferably, the method for removing the damage includes diamond single-point turning, chemical mechanical polishing, or chemical etching.

[0027] The non-destructive testing method for rapidly determining and removing material damage provided by the present invention, as described above, has the following beneficial effects:

[0028] Abandoning the inherent technical path of traditional elliptic polarization analysis, which requires the establishment of complex physical models, data fitting, and reliance on material composition information, this innovative approach utilizes the keen influence of processing damage on the key energy characteristic values ​​of the band structure of semiconductor materials. By directly mathematically differentiating the measured original elliptic polarization pseudo-permittivity, the CP characteristic value is extracted and its relative position change is analyzed. This allows for a rapid, intuitive, and non-destructive determination of whether processing damage exists in the material and whether the damage has been effectively removed.

[0029] Compared with existing technologies, this invention eliminates the need for testing under inert gas protection such as nitrogen, the need to establish a multi-layered physical model, the need for any data fitting, and even the need to know the specific composition information of the test material. This lowers the detection threshold and complexity, improves detection efficiency, and is particularly suitable for evaluating the processing quality and optimizing the process of soft and brittle semiconductor materials. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0031] Figure 1 A schematic diagram showing the curves of the real part of the pseudo-dielectric constant of the original elliptically polarized material as a function of energy, for two processed and unprocessed mercury cadmium telluride materials tested at a 45° angle, provided in an embodiment of the present invention.

[0032] Figure 2 A schematic diagram of the curve after taking the second derivative of the curve provided in the embodiment of the present invention;

[0033] Figure 3 The embodiments of the present invention provide a basis for... Figure 1 A schematic diagram of the real part of the second derivative curve of the pseudo-dielectric constant of the elliptic polarization after damage removal of the processed sample. Detailed Implementation

[0034] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0035] Please see Figure 1-3 This invention provides a technical solution: a non-destructive testing method for rapidly determining and removing material damage, comprising the following steps:

[0036] S01. Provide a single-point diamond-machined mercury cadmium telluride (MCH) wafer as the semiconductor material sample to be tested. This sample must contain at least one machined area or one machined surface. Simultaneously, provide an unprocessed MCH wafer from the same batch as an undamaged reference sample. All samples are directly exposed to normal air without the need for inert gas protection such as nitrogen, and their specific cadmium composition (x-value) does not need to be determined beforehand.

[0037] It is worth noting that the sample can be left in the air for any amount of time, and the presence of a natural oxide layer on the surface does not affect the judgment results of this invention.

[0038] S02. A commercial ellipsometry was used to measure the sample to be tested and the reference sample. The test conditions were set as follows: the incident angle was fixed at 45°, the scanning energy range of the light source covered the CP characteristic region of the material, and the scanning step could be set to 0.01 eV or less according to the required accuracy.

[0039] During measurement, the ellipsometer directly outputs the original measurement parameters (Ψ and Δ), and then the curve of the pseudo-dielectric constant changing with photon energy is directly calculated through the device's built-in software or conventional data processing programs. The pseudo-dielectric constant refers to the effective dielectric constant calculated directly from the reflective ellipsometric measurement data. It is not modified by any physical model (such as a multilayer structure model or an effective dielectric approximation model), nor does it require any data fitting or known material composition information. The real part (ε1) and imaginary part (ε2) of the pseudo-dielectric constant characterize the material's polarization response to electromagnetic waves and energy absorption properties, respectively.

[0040] like Figure 1 As shown, the original curves of the real part (ε1) of the pseudo-dielectric constant as a function of photon energy are presented for two mercury cadmium telluride samples, one processed and one unprocessed. Figure 1 As can be seen, the original curves do not show significant differences, making it difficult to directly determine the damage status.

[0041] S03. To enhance the discernibility of subtle features caused by the band structure CP point in the curve and eliminate the interference of background signals, the curve of the real part (ε1) (and / or imaginary part (ε2)) of the pseudo-dielectric constant obtained in step S02 as a function of energy is mathematically differentiated.

[0042] In this embodiment, the mathematical calculations employ the first, second, or third derivative, but the second derivative is the preferred choice. The differentiation operation can be automatically performed in data processing software using a numerical differentiation algorithm.

[0043] like Figure 2 As shown, for Figure 1After taking the second derivative of the original (ε1) curve, the inflection point region, which was originally a gradual change in the original curve, is transformed into clear and sharp characteristic peaks (positive or negative peaks). Specifically, the first major characteristic peak (corresponding to E1) appears in the range of approximately 2.2–2.4 eV, the second major characteristic peak (corresponding to E1+Δ1) appears in the range of approximately 2.8–3.0 eV, and the third major characteristic peak (corresponding to E2) appears in the range of approximately 4.4–4.5 eV. The positions of these characteristic peaks greatly facilitate the subsequent extraction of eigenvalues.

[0044] It should be noted that, depending on the material system and the degree of damage, those skilled in the art may also use the first or third derivative, both of which can achieve the purpose of highlighting CP characteristics and enhancing identification. The choice of the derivative order is one of the alternative solutions of this invention.

[0045] Step S04: From the obtained derivative spectrum (e.g.) Figure 2 As shown in the figure, at least one feature value corresponding to the material's critical energy CP is identified and extracted using an automatic peak identification algorithm or manual annotation. These feature values ​​mainly include:

[0046] E1 eigenvalue: corresponds to the position (energy value) of the first major characteristic peak or valley at a lower energy level in the derivative spectrum, and is usually measured in electron volts (eV).

[0047] E1+Δ1 eigenvalue: The position of the second characteristic peak or valley caused by spin-orbit splitting after the E1 peak.

[0048] E2 eigenvalues: correspond to the positions of characteristic peaks or valleys at higher energies, triggered by higher critical points in the energy band.

[0049] exist Figure 2 In the illustrated embodiment, the E1, E1+Δ1, and E2 values ​​extracted from the second derivative spectrum of the processed sample are 2.31473 eV, 2.90473 eV, and 4.49973 eV, respectively; the corresponding values ​​extracted from the second derivative spectrum of the unprocessed (undamaged) sample are 2.24573 eV, 2.83373 eV, and 4.42173 eV, respectively.

[0050] Step S05: Compare the feature values ​​extracted from the sample to be tested (after processing) with the reference feature values ​​(here, the corresponding values ​​extracted from the same undamaged material under the same test conditions) item by item. Based on the comparison results, make the following judgment:

[0051] If a shift exists: The E1, E1+Δ1, and E2 values ​​of the processed sample were observed to shift towards higher energies relative to their corresponding values ​​in the unprocessed sample (i.e., a blue shift), with shifts of 0.069 eV, 0.071 eV, and 0.078 eV, respectively. This significant blue shift clearly indicates the presence of compressive stress-induced lattice distortion and crystal quality damage within the material due to processing. Therefore, it is determined that the semiconductor material under test exhibits processing damage.

[0052] If basically consistent: If the extracted feature value is basically consistent with the reference feature value within the instrument measurement accuracy range (e.g., deviation less than ±0.01eV), then the sample is determined to be undamaged or the damage has been removed.

[0053] Example 2

[0054] This embodiment illustrates how to use the method of the present invention to monitor the damage removal process in order to verify whether the damage has been completely removed.

[0055] The cadmium telluride mercury samples processed in Example 1 and determined to be damaged were subjected to damage removal treatments in sequence. In this example, chemical mechanical polishing (chemical etching followed by polishing) was used as the damage removal method. However, it should be understood that the damage removal method is not limited to this and may also include diamond single-point turning, chemical etching, or a combination of the above methods.

[0056] For samples that have undergone one or more chemical mechanical polishing treatments, repeat steps S02 to S05. For example... Figure 3 As shown, Figure 1 The second derivative curve of the real part of the pseudo-dielectric constant of the processed sample after damage removal treatment. Figure 3 It can be clearly seen that the E1, E1+Δ1, and E2 characteristic peak positions corresponding to CP have recovered to approximately 2.24573 eV, 2.83373 eV, and 4.42173 eV, respectively, which are basically consistent with the corresponding characteristic values ​​of the unprocessed reference sample. According to the judgment rule in step S05, the extracted characteristic values ​​are basically consistent with the reference characteristic values, therefore it is determined that the processing damage on the sample has been completely removed.

[0057] Example 3

[0058] The above embodiments use unprocessed samples from the same batch as a reference. If unprocessed samples are unavailable, the following alternative can be used: if the sample to be tested has unprocessed or undamaged areas (e.g., areas protected by a mask, or areas where the processing edges are unaffected), these areas can be tested under the same testing conditions, and the extracted feature values ​​can be used as reference feature values. This method eliminates the need for additional reference samples, further improving the convenience of the method.

[0059] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A non-destructive testing method for rapid assessment and removal of material damage, characterized in that, Includes the following steps; S01. Provide a semiconductor material to be tested, wherein the sample includes at least a processed region and / or a processed surface; S02. The sample is measured using an ellipsometry to obtain the original ellipsometry data of the sample at a preset incident angle, and the curve of the pseudo dielectric constant as a function of photon energy is calculated accordingly. The pseudo dielectric constant includes a real part and / or an imaginary part. S03. The curve of the pseudo-dielectric constant changing with photon energy is mathematically differentiated to obtain the derivative spectrum; S04. Extract at least one feature value corresponding to the material's critical energy CP from the derivative spectrum, wherein the feature value includes one or more of the following: E1 feature value, E1+Δ1 feature value, and E2 feature value; S05. The extracted feature values ​​are compared with reference feature values, where the reference feature values ​​are the corresponding feature values ​​of the same undamaged material under the same test conditions, and then an evaluation is given: If the extracted feature value deviates from the reference feature value, it is determined that the semiconductor material under test has processing damage. If the extracted feature value is basically consistent with the reference feature value, it is determined that the semiconductor material under test is undamaged or the damage has been removed.

2. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The mathematical differentiation in step S03 refers to the first derivative, second derivative, or third derivative.

3. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The pseudo-dielectric constant mentioned in step S02 is the original pseudo-dielectric constant obtained by direct calculation.

4. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The preset incident angle in step S02 is 45°, and the entire test of the elliptic polarimeter does not require inert gas protection and is conducted directly in the air environment.

5. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, In step S04, the E1 characteristic value, E1+Δ1 characteristic value, and E2 characteristic value correspond to the positions of characteristic peaks or valleys in the derivative spectrum caused by the key energy CP point in the band structure of the corresponding semiconductor material sample.

6. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The reference feature mentioned in step S05 is derived from any of the following methods: The characteristic values ​​were obtained and calculated from an undamaged standard sample of the same material as the semiconductor material to be tested, without processing, and under the same test conditions; or The characteristic values ​​are measured and calculated under the same test conditions in the unprocessed or undamaged regions of the semiconductor material under test.

7. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The shift mentioned in step S05 includes a blue shift towards higher energy or a red shift towards lower energy. Among them, blue shift indicates compressive stress-type damage introduced by machining, while red shift indicates tensile stress-type damage introduced by machining.

8. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The semiconductor material to be tested is a direct bandgap semiconductor material with a significant critical energy (CP) characteristic, and it includes at least mercury cadmium telluride.

9. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The processing methods of the semiconductor material to be tested include diamond single-point turning, chemical mechanical polishing, chemical etching, or epitaxial growth. The methods for removing the damage include diamond single-point turning, chemical mechanical polishing, or chemical etching.

10. The non-destructive testing method for rapid determination and removal of material damage according to claim 1, characterized in that, The methods for removing the damage include diamond single-point turning, chemical mechanical polishing, or chemical etching.

Images