Optimization method for enhancing double-layer overlay error measurement mark of high-order diffracted light

Abstract

The invention provides an optimization method of a double-layer overlay error measurement mark for enhancing high-order diffracted light, which can be applied to the technical field of metering and detection in a semiconductor manufacturing process, and comprises the following steps: firstly, constructing a mark comprising two layers of same grating structures, and arranging a main grating ridge and a secondary grating ridge in a periodic unit of each layer of grating; initial line widths of a primary grating ridge and a secondary grating ridge are respectively determined on the basis of a scalar diffraction theory, then a vector diffraction algorithm is adopted to optimize the target high-order diffraction order (such as + / -3 order) efficiency, and the thickness of each layer is optimized by using a minimum asymmetric factor change method to enhance the process deformation resistance. And then simulation performance test and asymmetric deformation resistance verification are carried out on the optimized structure until preset precision and robustness conditions are met. Through the systematic design, the intensity of high-order diffraction signals such as + / -3 levels and the like can be obviously enhanced on the basis of keeping the + / -1 level diffraction signals available.

Description

Technical Field

[0001] This disclosure relates to the field of metrology and testing technology in semiconductor manufacturing processes, and in particular to an optimized method for measuring double-layer overlay error markings that enhance higher-order diffraction light. Background Technology

[0002] Integrated circuit manufacturing technology has entered the nanoscale era, where overlay accuracy directly determines device performance and yield. Overlay error refers to the positional deviation between the current layer pattern and the previous layer pattern. As technology nodes continue to shrink, according to the requirements of the international semiconductor technology roadmap, the control accuracy of overlay error needs to reach the sub-nanometer level, which poses an unprecedented challenge to overlay error measurement technology.

[0003] Currently, diffraction-based overlay error measurement technology is the mainstream high-precision measurement method in the industry. This technology extracts overlay error information by fabricating specific periodic grating marks on a substrate, illuminating the marks with measurement light, and analyzing the diffracted light signals. Its basic principle is: when there is no overlay error, the symmetry of the mark ensures that the light intensities of the positive and negative diffraction orders are equal; when an overlay error exists, the symmetry is disrupted, and the light intensities of the positive and negative orders differ. Within a small error range, this intensity difference is linearly related to the overlay error, thus the overlay error value can be derived by measuring the intensity difference.

[0004] However, traditional diffraction-based overlay error measurement techniques primarily rely on ±1st-order diffraction light for signal extraction, which has significant limitations in practical applications. First, during frequency domain detection, ±1st-order diffraction signals easily overlap with zero-order diffraction signals, leading to crosstalk and affecting measurement accuracy. Second, ±1st-order diffraction signals are highly sensitive to manufacturing errors in the grating markings (such as grating sidewall tilting or top asymmetry). These process-induced non-ideal deformations of the markings introduce significant measurement errors, reducing measurement robustness. Furthermore, because traditional methods have a single correspondence between diffraction order and wavelength, if the ±1st-order signal performance at a particular wavelength is poor, that wavelength cannot be effectively used, limiting the flexibility of the measurement system in wavelength selection.

[0005] To overcome these problems, the industry has explored various improvement solutions. For example, employing metrology tool optical proximity effect correction technology or designing miniaturized overlay error measurement marks can reduce mark size and improve process compatibility; or using continuous bias overlay error measurement technology can broaden the usable wavelength range. On the other hand, theoretical research indicates that higher-order diffraction orders (such as ±3rd order) have stronger resistance to certain asymmetric process errors in the marks, and their signals are further separated from the zero-order signal in the frequency domain, avoiding crosstalk problems. Therefore, utilizing higher-order diffraction orders for measurement is considered a promising technological approach.

[0006] However, the diffraction efficiency of higher-order diffraction orders (especially ±3rd order and above) is usually much lower than that of ±1st order, resulting in weak signal strength that is difficult for detection systems to effectively capture and use for high-precision calculations. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] To address at least one of the aforementioned technical problems in the prior art, embodiments of this disclosure provide an optimized method for measuring double-layer overlay error of higher-order diffraction light. By combining scalar diffraction theory analysis with vector diffraction algorithm optimization, the marker structure parameters can be synergistically optimized, enhancing the efficiency of the target higher-order diffraction orders while also considering resistance to process asymmetric deformation. The measurement method based on this marker can simultaneously acquire and process diffraction light signals of at least two different orders, including ±1st and ±3rd orders. Overlay error values ​​are calculated using multi-level signal fusion technology, thereby significantly improving measurement accuracy, robustness, and wavelength selection flexibility. This effectively solves the problems of signal crosstalk, high process sensitivity, and limited measurement conditions faced by traditional measurement methods based on ±1st-order diffraction.

[0009] (II) Technical Solution

[0010] To address the aforementioned technical problems, embodiments of this disclosure propose an optimized method for measuring double-layer overlay error marks that enhances higher-order diffraction light.

[0011] According to a first aspect of this disclosure, an optimization method for a double-layer overlay error measurement mark to enhance high-order diffraction light is provided, comprising the following steps: Step S1, constructing a mark structure for the double-layer overlay error measurement mark, the mark structure comprising two layers of spaced grating structures, the grating structures comprising a plurality of grating periodic units arranged periodically along a first direction, each grating periodic unit comprising a main grating ridge and a secondary grating ridge; Step S2, determining a first initial linewidth of the main grating ridge and a second initial linewidth of the secondary grating ridge based on scalar diffraction theory; Step S3, based on a vector diffraction algorithm, determining the linewidth of the secondary grating ridge according to diffraction efficiency... The first and second initial linewidths are optimized; in step S4, based on the preset overlay error value, the mark structure after the initial linewidth is optimized is subjected to simulation performance test. If the test result meets the first preset condition, the mark structure after the initial linewidth is optimized is output; otherwise, return to step S3; and in step S5, the asymmetric deformation of the bottom mark is introduced into the mark structure output in step S4, and the maximum measurement error under the deformation condition is evaluated. If the measurement error meets the second preset condition, the mark structure after the initial linewidth is optimized is output as the final mark structure; otherwise, return to step S3.

[0012] In some exemplary embodiments, prior to step S3, the optimization method further includes: optimizing the thickness of each layer of the double-layer overlay error measurement mark using the minimum asymmetry factor variation method.

[0013] In some exemplary embodiments, the marking structure includes, sequentially arranged along the stacking direction, a substrate, a bottom grating, an intermediate layer, and a top grating, wherein the bottom grating and the top grating have the same grating structure. The thickness of each layer of the double-layer overlay error measurement mark is optimized using the minimum asymmetry factor variation method, including: by analyzing the rate of change of the asymmetry factor of the marking structure with the incident light wavelength under the influence of process error, adjusting the thickness of each layer so that the rate of change tends to be gradual or minimized within the target wavelength range.

[0014] In some exemplary embodiments, the effects of process errors are simulated by introducing a pre-defined geometric asymmetric deformation model into the marking structure; the geometric asymmetric deformation model includes at least one of a top linear tilt deformation model and a sidewall tilt deformation model.

[0015] In some exemplary embodiments, in step S2, based on scalar diffraction theory, the duty cycle parameters of the primary grating ridge and the secondary grating ridge are optimized to enhance the diffraction efficiency of the target higher-order diffraction orders, thereby determining the first initial linewidth and the second initial linewidth.

[0016] In some exemplary embodiments, in step S3, the optimization objective of optimizing the first initial linewidth and the second initial linewidth based on the diffraction efficiency is to increase the diffraction efficiency of the target higher-order diffraction order of the marker structure to 40% or greater than or equal to the diffraction efficiency of the basic diffraction order.

[0017] In some exemplary embodiments, the target higher-order diffraction order is ±3rd order, and the fundamental diffraction order is ±1st order.

[0018] In some exemplary embodiments, step S4 includes: calculating the measurement error of the mark structure corresponding to each preset overlay error value based on a plurality of preset overlay error values; if the maximum measurement error corresponding to all preset overlay error values ​​is less than a first preset threshold, then the test result is determined to meet the first preset condition.

[0019] In some exemplary embodiments, step S5 introduces asymmetric deformation of the bottom mark, including applying at least two different types or different degrees of geometric asymmetric deformation; evaluating the maximum measurement error under the deformation conditions, including determining whether the maximum measurement error is less than a second preset threshold under all applied geometric asymmetric deformation conditions.

[0020] In some exemplary embodiments, in the marking structure of the double-layer overlay error measurement mark, the substrate and the bottom grating are made of silicon, the middle layer is made of silicon dioxide, and the top grating is made of photoresist.

[0021] (III) Beneficial Effects

[0022] As can be seen from the above technical solutions, the optimized method for measuring double-layer overlay error of higher-order diffraction light provided in this disclosure has at least the following beneficial effects:

[0023] (1) By innovatively introducing a composite structure containing a main grating ridge and a secondary grating ridge in each grating periodic unit, the energy distribution of the light field can be effectively controlled, and the intensity of ±3rd order or higher order diffraction light, which is traditionally difficult to utilize, can be improved to a level that can be used for high-precision measurement. This fundamentally solves the technical bottleneck of weak high-order diffraction signal and provides an additional signal source with higher quality for the measurement system.

[0024] (2) The designed marking structure enables the practical application of higher-order diffraction orders (such as ±3rd order), which have the advantages of being further separated from the zero-order signal and having stronger resistance to some process asymmetric deformations (such as sidewall tilting). Combined with the multi-diffraction order signal fusion measurement method proposed in this disclosure, the complementary characteristics of ±1st order and ±3rd order signals can be utilized to comprehensively calculate the overlay error through algorithms (such as weighted averaging or fitting), thereby significantly suppressing the measurement fluctuations caused by noise or systematic errors from a single signal source and obtaining more accurate and stable measurement results.

[0025] (3) The systematic design method provided in this disclosure, especially the optimization step that incorporates the "minimum asymmetry factor variation method", can pre-compensate and optimize for possible process asymmetry deformation of the marker bottom layer during the design stage. This makes the final determined marker structure itself more robust to process fluctuations, reduces the negative impact of process deviations on measurement accuracy in actual manufacturing, and improves the reliability of measurement results.

[0026] (4) This disclosure not only provides an optimized marker structure, but also a complete design method from theoretical analysis (scalar diffraction) to accurate simulation (vector diffraction), and then to performance verification and anti-interference testing. The method is logically clear and the steps are well-defined. It can effectively guide the design and optimization of actual markers, ensuring that the designed markers enhance the target diffraction order while achieving optimal overall measurement performance, and has high engineering practical value. Attached Figure Description

[0027] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0028] Figure 1 The diagram illustrates the ±n order light intensity difference curves as a function of offset in diffraction-based overlay error measurement.

[0029] Figure 2 This diagram schematically illustrates the positions of each diffraction order during frequency domain detection.

[0030] Figure 3 A schematic diagram of a grating structure with top linear asymmetric deformation is shown.

[0031] Figure 4 A schematic diagram illustrating the principle of multi-diffraction order measurement is shown.

[0032] Figure 5 A schematic diagram illustrating the structure of a double-layer overlay error measurement mark according to an embodiment of the present disclosure is shown.

[0033] Figure 6 A flowchart illustrating an optimized method for measuring double-layer overlay error marks for enhancing higher-order diffraction light according to an embodiment of the present disclosure is shown schematically.

[0034] Figure 7 A schematic diagram of a double-layer overlay error measurement mark structure with enhanced diffraction order according to an embodiment of the present disclosure is shown.

[0035] Figure 8 A schematic diagram illustrating asymmetric deformation of a mark according to an embodiment of the present disclosure is shown, wherein Figure 8 (a) schematically illustrates the top linear asymmetric deformation of the mark according to an embodiment of the present disclosure; Figure 8 (b) schematically illustrates the asymmetric deformation of the sidewall of the mark according to an embodiment of the present disclosure;

[0036] Figure 9 A flowchart illustrating an overlay error measurement method according to an embodiment of the present disclosure is shown schematically.

[0037] Figure 10 A schematic diagram illustrating the diffraction efficiency distribution of the optimized marker according to Embodiment 1 of this disclosure is shown.

[0038] Figure 11 A schematic diagram illustrating the theoretical performance test results of the optimized marker according to Embodiment 1 of this disclosure is shown; and

[0039] Figure 12 The maximum measurement error curve of the optimized marker according to Embodiment 1 of this disclosure is schematically shown, wherein, Figure 12 (a) schematically illustrates the maximum measurement error curve after adding the top tilt angle according to Embodiment 1 of this disclosure; Figure 12 (b) schematically illustrates the maximum measurement error curve after adding the sidewall tilt angle according to Embodiment 1 of this disclosure. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of protection of this disclosure.

[0041] This disclosure aims to propose an improved and optimized method for measuring double-layer overlay error marks, in order to solve the problems of limited measurement performance and narrow usable wavelength range caused by excessive reliance on ±1st order diffraction signals in existing diffraction-based overlay error measurement techniques.

[0042] Diffraction-based overlay error measurement technology obtains overlay error information by analyzing specific orders of diffracted light from the marker. Its basic principle is that when there is no overlay error, the symmetry of the marker structure ensures that the light intensities of positive and negative diffraction orders are equal, resulting in zero intensity difference. When an overlay error exists, the structural symmetry is disrupted, leading to a difference in light intensity between the positive and negative orders. Within a small offset range, this intensity difference exhibits a good linear relationship with the overlay error value, such as... Figure 1 As shown. The specific calculation formula is:

[0043]

[0044] Where OV is the overlay error value to be determined, and D is the preset known offset. and The preset offsets are respectively and The positive and negative nth order diffracted light intensities measured at that time. as well as .

[0045] To achieve a breakthrough in measurement performance, the core strategy of this disclosure lies in introducing and effectively utilizing higher-order diffraction orders, particularly ±3rd orders. Higher-order diffraction orders offer two significant advantages: First, in frequency domain detection, the spatial separation between ±3rd-order and zero-order diffraction signals is far greater than that of ±1st-order signals. Figure 2 As shown. This fundamentally avoids the crosstalk problem often faced by ±1-order signals due to partial overlap with the zero-order signal, which is beneficial for obtaining a purer measurement signal with less interference. Secondly, theoretical analysis and existing research confirm that higher-order diffraction orders have stronger resistance to asymmetric deformation of the marking structure caused by process fluctuations such as photolithography and etching. The amplitude function expression of the diffraction after a unit plane wave incident on a grating structure with asymmetric deformation at the top is as follows:

[0046]

[0047] Its nth-order component is represented as:

[0048]

[0049] The above formula can be understood as the amplitude of the nth diffraction order. The error coefficient is represented by p, the grating period is p, the center position of the grating ridge is x, and j represents the imaginary unit, i.e., j 2 =-1 According to the principle of overlay error measurement, the error signal can be expressed as:

[0050]

[0051] The asymmetric deformation of the markings will cause the intensity difference of the ±n-order diffracted light to no longer be 0. Discussing the above equations separately will make the results clearer, that is, the ±3rd order has a stronger resistance to process errors than the ±1st order:

[0052]

[0053] For example, for grating structures with asymmetrical deformation at the top (such as...) Figure 3 As shown in the figure, theoretical analysis of the diffraction amplitude distribution indicates that a specific error coefficient characterizing the influence of process errors... It decreases as the diffraction order n increases, that is... This means that under the same process deformation conditions, the overlay error value calculated based on ±3 level signals is less affected by process errors, thus significantly improving the robustness and reliability of the measurement.

[0054] However, directly utilizing the ±3 order for measurement faces a fundamental challenge: in conventional grating mark designs, the diffraction efficiency of the ±3 order is typically much lower than that of the ±1 order, resulting in excessively weak signal intensity that is difficult for the detection system to reliably capture and use for high-precision calculations. This is the main bottleneck restricting its practical application. Furthermore, this disclosure further clarifies the measurement principle and synergistic value of using multiple diffraction orders, such as... Figure 1 As shown. In practical applications, the ideal linear relationship between light intensity difference and overlay error will introduce an inherent nonlinear error term OV due to factors such as wavelength, diffraction order, and the specific structure of the marking. M .

[0055]

[0056] in, and These are the slopes of the straight line formed by the point on the light intensity difference curve and the origin, respectively, while k is the slope of the straight line under the linear assumption.

[0057] By simultaneously acquiring and utilizing multiple diffraction signals, such as ±1st and ±3rd orders, a "self-reference" or internal correction mechanism similar to multi-wavelength technology can be constructed. Since different diffraction orders exhibit varying response characteristics (such as sensitivity and inherent error terms) to the same calibration error, fusion processing of these multi-order signals through algorithms can effectively compensate for or offset some inherent systematic errors in single-order measurements, thereby correcting measurement performance and achieving more accurate and reliable measurements over a wider wavelength range.

[0058] In summary, the core technical challenge addressed by this disclosure lies in overcoming the bottleneck of insufficient intensity of higher-order diffraction signals such as ±3rd order, and systematically designing an overlay error measurement mark and its design method that can effectively excite and utilize these higher-order signals while simultaneously optimizing their performance with ±1st order signals. To this end, the specific optimization design method proposed in this disclosure will be described in detail below.

[0059] Figure 5 A schematic diagram of the structure of a double-layer overlay error measurement mark according to an embodiment of the present disclosure is shown.

[0060] like Figure 5 As shown, the double-layer overlay error measurement mark according to an embodiment of the present disclosure includes: a substrate, a bottom grating formed on the substrate, the bottom grating including a plurality of bottom grating periodic units arranged periodically along a first direction; an intermediate layer covering the bottom grating; and a top grating formed on the intermediate layer, the top grating including a plurality of top grating periodic units arranged periodically along the first direction; wherein the bottom grating unit and / or the top grating unit includes a main grating ridge and at least one secondary grating ridge in a cross section parallel to the first direction, and the structural parameters of the main grating ridge and the secondary grating ridge are configured to enhance the intensity of ±3rd order or higher order diffraction light generated by the mark.

[0061] By simultaneously setting both the primary grating ridge and the secondary grating within a grating periodic unit, the simple periodicity of the traditional single grating ridge is broken, providing additional, finely tunable structural degrees of freedom. By optimizing the structural parameters of these grating ridges (such as width, position, and duty cycle), the diffraction behavior of incident light within the marker can be controlled, particularly by selectively coupling the light field energy to ±3rd order or higher odd-order diffraction orders. This fundamentally solves the core bottleneck problem of insufficient intensity of high-order diffraction signals, making them unsuitable for high-precision measurements.

[0062] In the embodiments of this disclosure, the bottom grating and the top grating have the same grating period and linewidth structure. This arrangement ensures that the bottom and top gratings are perfectly matched in spatial frequency. When the incident light meets a specific phase relationship, this periodic consistency can guide the light to undergo constructive interference or resonance enhancement between the two grating layers, thereby synergistically amplifying the output efficiency of higher-order diffraction orders (such as ±3rd order) of the target and further improving its signal strength. In addition, consistency also helps to reduce additional measurement errors or signal complexity caused by mismatch in the parameters of the upper and lower grating layers.

[0063] based on Figure 5 The structure of the double-layer overlay error measurement mark shown below is combined with... Figures 6-8 The optimization method for double-layer overlay error measurement marks for enhancing higher-order diffraction light according to embodiments of this disclosure is described in detail.

[0064] Figure 6 A flowchart illustrating an optimized method for measuring double-layer overlay error marks of enhanced higher-order diffraction light according to an embodiment of the present disclosure is shown.

[0065] like Figure 6 As shown, the optimization method for double-layer overlay error measurement marks for enhancing higher-order diffraction light according to an embodiment of the present disclosure includes steps S1 to S5.

[0066] In step S1, a marking structure for double-layer overlay error measurement marking is constructed. The marking structure includes two layers of grating structures spaced apart. The grating structure includes multiple grating periodic units arranged periodically along a first direction. Each grating periodic unit contains a main grating ridge and a secondary grating ridge.

[0067] In step S2, based on scalar diffraction theory, the first initial linewidth of the main grating ridge and the second initial linewidth of the secondary grating ridge are determined respectively.

[0068] In the embodiments of this disclosure, step S2 may specifically include, based on scalar diffraction theory, optimizing the duty cycle parameters of the primary grating ridge and the secondary grating ridge to enhance the diffraction efficiency of the target higher-order diffraction orders, thereby determining the first initial linewidth and the second initial linewidth.

[0069] In the embodiments of this disclosure, after the marker is split into two layers of grating structures with the same linewidth, scalar diffraction theory is used to analyze the effect of the linewidths of the primary and secondary grating ridges on the grating diffraction efficiency after adding the secondary grating ridges. The diffraction efficiency of each diffraction order with two grating ridges in one period can be expressed as:

[0070]

[0071] Among them, f iThe denoting factor represents the duty cycle of the i-th grating ridge, p is the grating period, λ is the incident wavelength, h is the grating ridge height, n is the diffraction order, and x represents the grating ridge height. i Let be the center position of the i-th grating ridge. The marker linewidth structure with ±3 enhancement levels can be easily obtained from the above formula. The upper and lower grating linewidth structures are identical. The height of each layer initially adopts the classic value of the marker, and the optimized marker structure is designed as follows: Figure 7 As shown.

[0072] In step S3, the first initial linewidth and the second initial linewidth are optimized based on the vector diffraction algorithm and the diffraction efficiency.

[0073] In the embodiments of this disclosure, the optimization objective for optimizing the first initial linewidth and the second initial linewidth based on diffraction efficiency is to increase the diffraction efficiency of the target higher-order diffraction orders of the marker structure to 40% or greater than or equal to the diffraction efficiency of the basic diffraction order. The target higher-order diffraction order is ±3 orders, and the basic diffraction order is ±1 order.

[0074] exist Figure 6 Based on the method shown, the optimization method for enhancing high-order diffraction light double-layer overlay error measurement marks according to the embodiments of this disclosure may further include, before step S3, optimizing the thickness of each layer of the double-layer overlay error measurement mark using the minimum asymmetry factor variation method. Specifically, by analyzing the rate of change of the asymmetry factor of the mark structure with the incident light wavelength under the influence of process errors, the thickness of each layer is adjusted so that the rate of change tends to be gradual or minimized within the target wavelength range. The influence of process errors is simulated by introducing a preset geometric asymmetry deformation model into the mark structure; the geometric asymmetry deformation model includes at least one of a top linear tilt deformation model and a sidewall tilt deformation model.

[0075] For example, the minimum asymmetry factor variation method is used to optimize the thickness of each layer of the marker. This method minimizes the impact of measurement errors caused by asymmetric deformation of the marker on wavelength selection by introducing and analyzing the process error signal of a reference marker located near the marker that contains only the bottom layer structure.

[0076] In some exemplary embodiments, the target higher-order diffraction order is ±3 orders, and the optimization objective includes making the ratio of the diffraction efficiency of ±3 orders to that of ±1 orders within the range of 0.4 to 2.0.

[0077] In the embodiments of this disclosure, the theory of the minimum asymmetric factor method is as follows:

[0078]

[0079] OV represents a preset, relatively large overlay error value. This represents the overlay error value obtained from simulation measurements. and The preset deviations are respectively The difference in light intensity during micro-nano fabrication. Alignment between two layers requires overlay marks. Overlay marks on the bottom layer are more susceptible to asymmetric deformation due to process errors, leading to larger measurement errors. Overlay marks on the upper layers, excluding the bottom layer, can be reduced after exposure through a series of planarization processes to mitigate the deformation caused by asymmetric deformation. Therefore, when discussing the impact of process errors on overlay marks, the focus should be on the bottom layer overlay marks. This is a defined error signal, generated by fabricating another set of grating structures on the substrate. These grating structures are identical to those on the substrate of the double-layer overlay error measurement mark. Due to their proximity to the overlay error measurement mark, they can be considered to be subject to the same process influence. By introducing a certain process error into the mark, its AF(λ) value is calculated. The minimum AF(λ) value is obtained by adjusting the thickness of each layer. Curves, two typical bottom mark variations, such as Figure 8 As shown, α is the deformation angle. This step may not be necessary in some processes because the thickness of each layer is process-dependent and cannot be adjusted in some scenarios.

[0080] After determining the thickness of each layer of the marker, the linewidths of the primary and secondary grating ridges of the entire marker were optimized, with ±1 and ±3 order diffraction efficiencies as the focus. The performance of the resulting structure was verified by using preset overlay error values. The performance test index here is the measured value OV. M Error between the value and the preset value OV.

[0081]

[0082] In step S4, based on the preset overlay error value, the simulation performance test is performed on the mark structure after the optimized initial line width obtained in step S3. If the test result meets the first preset condition, the mark structure after the optimized initial line width is output; otherwise, return to step S3.

[0083] In the embodiments of this disclosure, the simulation performance test includes: calculating the measurement error of the mark structure corresponding to each preset overlay error value based on a plurality of preset overlay error values; if the maximum measurement error corresponding to all preset overlay error values ​​is less than a first preset threshold, the test result is determined to meet the first preset condition.

[0084] In step S5, an asymmetric deformation of the bottom mark is introduced into the mark structure output in step S4, and the maximum measurement error under the deformation condition is evaluated. If the measurement error meets the second preset condition, the mark structure after optimizing the initial line width is output as the final mark structure; otherwise, return to step S3.

[0085] In embodiments of this disclosure, the asymmetric deformation of the bottom mark is introduced, including applying at least two different types or different degrees of geometric asymmetric deformation; the maximum measurement error under the deformation conditions is evaluated, including determining whether the maximum measurement error is less than a second preset threshold under all applied geometric asymmetric deformation conditions.

[0086] For example, simulated process asymmetric deformation is introduced into the marker structure, and the maximum measurement error under deformation conditions is evaluated. Process asymmetric deformation includes top linear tilt deformation and / or sidewall tilt deformation. Step S5 aims to further examine the marker's performance, focusing on testing the marker's resistance to asymmetric deformation using two classic asymmetric deformation methods, such as... Figure 8 As shown. If the effect is not satisfactory, you need to go back to the previous step and continue to optimize the marker line width.

[0087] In the embodiments of this disclosure, the substrate material of the double-layer overlay error measurement mark for enhancing high-order diffraction light includes silicon; the material of the bottom grating includes silicon; the material of the intermediate layer includes silicon dioxide; and the material of the top grating includes photoresist. Silicon substrate, silicon dioxide dielectric, and photoresist are common materials and structures in chip manufacturing, enabling the mark of this disclosure to be highly compatible with existing semiconductor manufacturing processes. It can be directly integrated into the wafer dicing or dedicated test area without introducing special materials or additional process steps. The appropriate refractive index difference between silicon, silicon dioxide, and photoresist allows for effective diffraction effects at the interfaces between layers, providing the necessary phase delay, thereby contributing to the generation of diffracted light signals with high signal intensity and clear contrast. From a functional perspective, the silicon bottom layer not only provides stable mechanical support but also possesses excellent electrical isolation properties; the silicon dioxide intermediate layer, as an optical spacer layer, can precisely adjust the optical path difference between the diffracted light from the upper and lower gratings, optimizing the interference enhancement effect; and the top photoresist layer can be easily formed into a periodic pattern structure with nanometer-level precision using conventional photolithography processes. The synergistic division of labor in the aforementioned material system ensures the excellent optical performance of the marking while also taking into account its manufacturability and process feasibility in existing production lines, achieving a good balance between optical design and process practice.

[0088] In some exemplary embodiments, after optimization using the above method, the duty cycle of the primary grating ridge in the double-layer overlay error measurement mark for enhancing higher-order diffraction light is 0.4 to 0.7; and the duty cycle of the secondary grating ridge is 0.05 to 0.3. The primary grating ridge maintains a relatively high duty cycle, ensuring that it, as the primary scatterer, can maintain sufficient fundamental diffraction efficiency (including ±1st-order signals). The secondary grating ridge, with its lower duty cycle, acts as a fine phase or amplitude modulation element, finely adjusting the Fourier spectrum distribution of the light field without significantly altering the overall structural symmetry. This specific combination of duty cycles, after optimization, can effectively redistribute and couple some of the diffraction energy originally concentrated in lower-order (e.g., ±1st-order) or zero-order directions to higher-order diffraction directions such as ±3rd order, thereby significantly improving the absolute intensity and signal-to-noise ratio of the ±3rd-order signals without excessively weakening the ±1st-order signals.

[0089] In some exemplary embodiments, after optimization using the above method, the height of the bottom grating ridge in the double-layer overlay error measurement mark for enhancing higher-order diffraction light is 50 nm to 150 nm; the thickness of the middle layer is 150 nm to 300 nm; and the height of the top grating ridge is 50 nm to 150 nm. The thickness parameters of each layer play a crucial role in optimizing the diffraction performance of the mark by controlling the phase accumulation and amplitude change of the light wave as it penetrates the multilayer structure. Specifically: the height of the grating ridge determines its phase modulation depth and diffraction signal intensity as an amplitude-phase diffraction element, directly affecting the relative distribution of light intensity at each diffraction order; the thickness of the middle layer (spacer layer) affects the amplitude value and signal-to-noise ratio of the probe light, influencing the signal intensity and contrast received at the probe end. The thickness combination determined through system optimization can improve the measurement performance of certain wavebands and filter that waveband. In addition, the set thickness range fully considers the processing capabilities and uniformity control level of thin film deposition and photolithography technologies in current mainstream semiconductor processes. While optimizing optical performance, it also ensures the feasibility and process repeatability of the marking structure in actual manufacturing.

[0090] Figure 9 A flowchart illustrating an overlay error measurement method according to an embodiment of the present disclosure is shown schematically.

[0091] like Figure 9 As shown, the overlay error measurement method according to an embodiment of the present disclosure includes steps S10 to S40.

[0092] In step S10, the above-mentioned overlay error measurement mark is used.

[0093] In the embodiments of this disclosure, an optimized double-layer overlay error measurement marker is employed, which innovatively introduces a secondary grating ridge structure into the traditional grating structure. By precisely controlling the duty cycle and relative position of the primary and secondary grating ridges, selective enhancement of higher-order diffraction light, such as ±3rd order, is achieved. This structural design improves the ±3rd order diffraction efficiency to a level comparable to ±1st order, creating physical conditions for the synergistic utilization of multi-order diffraction signals. This marker structure not only solves the technical problem of weak higher-order diffraction light signals but also lays a solid foundation for subsequent multi-signal fusion measurements.

[0094] In step S20, the mark is illuminated with a preset incident angle and polarized light.

[0095] In the embodiments of this disclosure, the incident angle and polarization state are precisely set according to the material properties, periodic parameters, and optical response characteristics of the target diffraction order. Linearly polarized light parallel to the grating lines is used for illumination to maximize the excitation of the optical response of the target diffraction order while effectively suppressing stray light and interference from non-target orders. This optimized matching of optical parameters ensures that the diffraction signal has sufficient intensity and signal-to-noise ratio, providing an ideal optical environment for high-precision signal acquisition.

[0096] In step S30, at least two different orders of diffraction light signals, including ±1 and ±3 orders, generated by the marker are simultaneously acquired.

[0097] In the embodiments of this disclosure, a high-sensitivity detection system is used to simultaneously acquire light intensity signals from multiple diffraction orders. Due to the optimized design of the marker structure, both ±1st and ±3rd order diffraction beams reach detectable intensity levels, making simultaneous acquisition of multiple signals possible. This parallel acquisition method not only improves measurement efficiency but, more importantly, provides a rich data source for subsequent signal fusion processing, enhancing the system's resistance to interference from anomalies in single-order signals.

[0098] In step S40, the overlay error value is calculated based on at least two different orders of diffraction light signals.

[0099] In some exemplary embodiments, step S40 may specifically include: acquiring at least two different orders of diffraction light signals under preset positive offset and preset negative offset respectively; calculating the light intensity difference corresponding to each diffraction order based on the diffraction light signals; and calculating the final overlay error value based on the correlation between the light intensity difference and the overlay error, and by fusing the information of at least two diffraction orders.

[0100] In some exemplary embodiments, fusing information from at least two diffraction orders includes: weighted averaging of overlay error values ​​calculated independently based on different diffraction orders, or using the intensity differences of multiple diffraction orders to jointly solve for the overlay error value through a fitting algorithm. This multi-level information fusion mechanism fully leverages the complementary advantages of different diffraction orders. The ±1 order provides basic measurement accuracy, while the ±3 order enhances resistance to process fluctuations. The synergistic effect of both significantly improves measurement accuracy to the sub-nanometer level, while also greatly enhancing the stability and reliability of the system under complex process environments.

[0101] Example 1:

[0102] like Figure 7 The optimized marking structure shown has the following specific parameter configurations: the bottom material is Si, the bottom grating material is Si, the top grating material is photoresist, and the middle layer material is SiO2; in terms of structural dimensions, the bottom grating ridge height is 100nm, the middle layer thickness is 230nm, and the top grating ridge height is 100nm; the optical parameters are: test wavelength 0.52μm, incident angle 36 degrees, and polarization direction parallel to the grating scribe line direction; the grating geometric parameters are: main grating ridge duty cycle f1=f3=0.55, secondary grating ridge duty cycle f2=f4=0.15, and grating period p=2μm.

[0103] Figure 10 A schematic diagram of the diffraction efficiency distribution of the optimized marker according to Embodiment 1 of this disclosure is shown.

[0104] like Figure 10 As shown, based on the optimized marking diffraction efficiency distribution of Embodiment 1 of this disclosure, the results show that the ±3rd order diffraction efficiency has been significantly improved to a level similar to the ±1st order, confirming the effective enhancement effect of the sub-grating structure on higher-order diffraction. In performance testing, the preset overlay error value OV was sampled at 1nm intervals within the range of 0~10nm. Figure 11 As shown, when ±1st and ±3rd order diffraction beams are used together for measurement, the maximum measurement error is reduced from nearly 0.5 nm when using only one diffraction order to less than 0.1 nm, and the measurement accuracy is significantly improved.

[0105] To further verify the robustness of the labeling in actual processes, a method was introduced. Figure 8 Two typical asymmetric deformations are shown, including top inclination angle (maximum 1°) and sidewall inclination angle (maximum 7°) for testing. Figure 12 As shown, under both deformation conditions, using a combination of ±1 and ±3 levels, the maximum measurement error remained stable below 0.1 nm, which was significantly better than the measurement results of a single diffraction level. This indicates that the marking structure has a strong resistance to asymmetric deformation caused by the process.

[0106] In summary, the introduction of the secondary grating alters the amplitude and phase modulation characteristics compared to traditional marking. By carefully designing the duty cycles and relative positions of the primary and secondary grating ridges, selective enhancement of the diffraction efficiency of specific higher-order diffraction orders (such as ±3rd order) can be achieved. The physical essence of this enhancement lies in controlling the scattering phase distribution in different regions within the grating period, enabling multiple diffraction orders to achieve energy equilibrium at a specific wavelength, thus laying the signal foundation for the coordinated use of multi-order diffraction signals. Furthermore, higher-order diffraction orders have higher spatial frequencies and are less sensitive to local deformation and process variations of the marking, further enhancing the anti-interference capability of the measurement system.

[0107] This optimized design, by introducing a sub-grating structure into the double-layer overlay markings, successfully improved the ±3rd order diffraction efficiency to a level comparable to the ±1st order, making it suitable for practical measurements. By combining the ±1st and ±3rd order diffraction signals for joint measurement, not only was the measurement accuracy of overlay errors significantly improved, but the mutual compensation between diffraction orders also effectively suppressed measurement errors caused by fluctuations in the performance of a single wavelength or single diffraction order. This design significantly enhances the robustness of the measurement system and strengthens its adaptability to complex process environments.

[0108] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of this disclosure. It should be understood that the above are only specific embodiments of this disclosure and are not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.

Claims

1. An optimized method for measuring double-layer overlay error markings to enhance higher-order diffraction light, characterized in that, Includes the following steps: Step S1: Construct a marking structure for a double-layer overlay error measurement mark. The marking structure includes two layers of grating structures spaced apart. The two layers of grating structures are identical. Each layer of grating structure includes multiple grating periodic units arranged periodically along a first direction. Each grating periodic unit includes a main grating ridge and a secondary grating ridge. Step S2: Based on scalar diffraction theory, determine the first initial linewidth of the main grating ridge and the second initial linewidth of the secondary grating ridge, respectively. Step S3: Based on the vector diffraction algorithm, optimize the first initial linewidth and the second initial linewidth according to the diffraction efficiency; Step S4: Based on the preset overlay error value, perform simulation performance testing on the mark structure after optimizing the initial line width obtained in step S3. If the test result meets the first preset condition, output the mark structure after optimizing the initial line width. Otherwise, return to step S3; as well as Step S5: Introduce asymmetrical deformation of the bottom mark into the mark structure output in step S4, and evaluate the maximum measurement error under the deformation conditions. If the measurement error meets the second preset condition, the mark structure after optimizing the initial line width is output as the final mark structure; otherwise, return to step S3.

2. The optimization method according to claim 1, characterized in that, Prior to step S3, the optimization method further includes: The thickness of each layer of the double-layer overlay error measurement mark is optimized using the minimum asymmetry factor variation method.

3. The optimization method according to claim 2, characterized in that, The marking structure includes, sequentially arranged along the stacking direction, a substrate, a bottom grating, an intermediate layer, and a top grating, wherein the bottom grating and the top grating have the same grating structure. The optimization of the thickness of each layer of the double-layer overlay error measurement mark using the minimum asymmetry factor variation method includes: By analyzing the rate of change of the asymmetry factor of the marking structure with the incident light wavelength under the influence of process errors, the thickness of each layer is adjusted to make the rate of change tend to be gradual or minimized within the target wavelength range.

4. The optimization method according to claim 3, characterized in that, The influence of the process error is simulated by introducing a preset geometric asymmetric deformation model into the marking structure; The geometrically asymmetric deformation model includes at least one of the top linear tilt deformation model and the sidewall tilt deformation model.

5. The optimization method according to claim 1, characterized in that, In step S2, based on scalar diffraction theory, the duty cycle parameters of the main grating ridge and the secondary grating ridge are optimized to enhance the diffraction efficiency of the target higher-order diffraction orders, thereby determining the first initial linewidth and the second initial linewidth.

6. The optimization method according to claim 1, characterized in that, In step S3, the optimization objective of optimizing the first initial linewidth and the second initial linewidth based on diffraction efficiency is to increase the diffraction efficiency of the target higher-order diffraction order of the marker structure to 40% or greater than or equal to the diffraction efficiency of the basic diffraction order.

7. The optimization method according to claim 6, characterized in that, The target higher-order diffraction order is ±3, and the basic diffraction order is ±1.

8. The optimization method according to claim 1, characterized in that, In step S4, the simulation performance test includes: Based on multiple preset overlay error values, the measurement error of the marking structure corresponding to each preset overlay error value is calculated; If the maximum measurement error corresponding to all preset overlay error values ​​is less than the first preset threshold, then the test result is determined to meet the first preset condition.

9. The optimization method according to claim 1, characterized in that, In step S5, the asymmetric deformation of the bottom mark is introduced by applying at least two different types or different degrees of geometric asymmetric deformation. The evaluation of the maximum measurement error under deformation conditions includes determining whether the maximum measurement error is less than a second preset threshold under all applied geometrically asymmetric deformation conditions.

10. The optimization method according to claim 3, characterized in that, In the marking structure of the double-layer overlay error measurement mark, the substrate and the bottom grating are made of silicon, the middle layer is made of silicon dioxide, and the top grating is made of photoresist.

Images