2025 Volume 34 Issue 2
Article Contents

Jingtong Feng, Rao Xu, Ziruo Wu, Lihua Lei, Yingfan Xiong, Zhaohui Tang, Guangxu Xiao, Yuying Xie, Dongbai Xue, Xiao Deng, Xinbin Cheng, Tongbao Li. Grating pitch comparison measurement based on Cr atomic transition frequency and Si lattice constant[J]. Chinese Physics B, 2025, 34(2): 020601. doi: 10.1088/1674-1056/ad9e9a
Citation: Jingtong Feng, Rao Xu, Ziruo Wu, Lihua Lei, Yingfan Xiong, Zhaohui Tang, Guangxu Xiao, Yuying Xie, Dongbai Xue, Xiao Deng, Xinbin Cheng, Tongbao Li. Grating pitch comparison measurement based on Cr atomic transition frequency and Si lattice constant[J]. Chinese Physics B, 2025, 34(2): 020601. doi: 10.1088/1674-1056/ad9e9a

Grating pitch comparison measurement based on Cr atomic transition frequency and Si lattice constant

  • Received Date: 29/09/2024
    Accepted Date: 12/11/2024
    Available Online: 01/01/2025
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(8)  /  Tables(3)

Article Metrics

Article views(242) PDF downloads(0) Cited by(0)

Access History

Grating pitch comparison measurement based on Cr atomic transition frequency and Si lattice constant

Abstract: Traceability is the fundamental premise of all metrological activities. The establishment of a traceability chain characterized by a shortened structure, while simultaneously enabling on-site traceability, represents a key trend in the advancement of metrology. This study explores the periodic accuracy and overall uniformity of self-traceable gratings, employing multilayer film gratings with a nominal period of 25.00 nm as the medium. We present a comparative analysis of measurement capabilities in a self-traceable grating calibration system characterized by a ‘top-down’ calibration approach and a Si lattice constant calibration system characterized by a ‘bottom-up’ calibration approach. The results indicate that the values obtained for the multilayer film grating periods, calibrated using the self-traceable grating system, are 24.40 nm with a standard deviation of 0.11 nm. By comparing with the values derived from the Si lattice constant, which yield 24.34 nm with a standard deviation of 0.14 nm, the validity and feasibility of the self-traceable calibration system are confirmed. This system extends and complements existing metrological frameworks, offering a precise pathway for traceability in precision engineering and nanotechnology research.

1.   Introduction
  • In the field of nanometrology, the continuous miniaturization of manufacturing processes presents increasing challenges for measurements at the nanoscale. Ensuring the traceability and accuracy of various critical dimensions (CD), feature shapes, and profile characterization results is particularly crucial.[1] The precise measurement and characterization of these parameters are key steps in ensuring product quality, optimizing process parameters, and calibrating measurement equipment, holding significant importance in industries such as manufacturing, microelectronics, and integrated circuits. For instance, in integrated circuit chip manufacturing, as the logical structure size of chips reaches below 3 nm, the chip structures have expanded from two-dimensional to three-dimensional. This necessitates the use of a combination of optical, probe, electronic, and x-ray measurement devices for accurate characterization.[2] Different fusion measurement devices require various calibration methods, highlighting the need to establish distinct nanometrological systems to meet the demands of integrated measurement.

    In the existing nanometrology systems, there are primarily two methods for achieving traceable measurements: one is the ‘top-down’ traceability chain based on laser wavelength, and the other is the ‘bottom-up’ traceability chain based on the Si lattice constant. The I2-stabilized 633 nm He–Ne laser radiation serves as a typical representative of the ‘top-down’ length traceability chain, where the value transfer process is as follows: International System of Units ‘metre’ → I2-stabilized 633 nm He–Ne laser interferometer → metrological testing instruments → standard materials → user instruments. This traceability chain requires calibration through three steps, leading to issues such as value accumulation due to the length of the traceability chain and low calibration efficiency, resulting in continuously decreasing calibration accuracy[3] (Fig. 1). The second method utilizes the Si lattice constant as the reference for a ‘bottom-up’ traceability chain. In 2018, the 26th International Conference on Weights and Measures recommended using the Si lattice spacing (d220 = 192.0155714 × 10−12 m) as an auxiliary means for the definition of the meter in nanometric geometric applications.[4] This established a ‘bottom-up’ value transfer method that employs the Si lattice spacing as an internal ruler to determine the dimensions of nanostructures. In static applications, to obtain atomically resolved measurement images, transmission electron microscopy (TEM) is generally employed, which is considered a destructive measurement. Meanwhile, in dynamic applications, the displacement measurement range of x-ray lattice interferometers is typically around 10 μm, making the expansion of the measurement range a key area of research.

    To meet the metrological testing requirements in the field of micro-displacement measurements for advanced nanofabrication, the team led by Academician Tongbao Li from Tongji University proposed a shortened nanometrology system based on Cr self-traceable gratings.[3] The core of this system lies in the application of atomic lithography technology to materialize the physical wavelength standard corresponding to the Cr atomic transition frequency of 7S3→7P4 into self-traceable gratings, which can be directly used for calibrating measurement instruments. A value comparison with the metrological atomic force microscope (AFM) from the Physikalisch-Technische Bundesanstalt (PTB) confirmed the feasibility of the traceability chain based on Cr self-traceable gratings. This self-traceable calibration system not only shortens the traceability chain and enhances calibration efficiency but also serves as a beneficial complement to traditional ‘top-down’ calibration systems based on laser wavelength standards.[3]

    However, the comparison between different traceability chains is crucial for verifying the accuracy of the value transfer system. The research team led by Gaoliang Dai conducted comparative measurements of Si structure linewidths by contrasting the ‘top-down’ and ‘bottom-up’ traceability systems.[5] The measurement results demonstrate that each method possesses distinct advantages and limitations, with the ‘bottom-up’ approach demonstrating higher measurement resolution and accuracy. The comparative measurement of the metrological means from both traceability methods is beneficial for enhancing overall calibration capability. Based on this, we recognize the importance of conducting comparisons between different traceability chains. Currently, the self-traceable calibration system has only undergone value comparisons with the ‘top-down’ calibration system.

    To conduct a comparative study on the accuracy between Cr self-traceable gratings and Si lattice constants, it is essential to select appropriate measurement instruments and comparison measurement substrates. In terms of measurement instruments, AFM, with its capability to image at the nanoscale and even atomic level, has become a key tool for obtaining critical data such as surface roughness and morphology. However, it is well known that due to the inherent characteristics of piezoelectric ceramics, there is a nonlinear relationship between the actual displacement of the scanner and the applied voltage during the measurement process, which can introduce measurement errors of 5% to 10%.[6] Thus, selecting AFM as the comparative measurement instrument is an effective means to validate the cross-scale calibration accuracy of self-traceable gratings. Regarding the comparison measurement substrates, on the one hand, atomic-level images must be obtained for the substrates, necessitating that the grating period should generally not exceed 50 nm. On the other hand, since AFM and TEM cannot be positioned on the same area, the structural consistency and uniformity within the comparison substrate itself are critical; otherwise, the results may only reflect structural deviations within the sample. The core of multilayer film deposition technology is to translate film thickness into linewidth structures, providing exceptional uniformity while enabling the fabrication of sub-50 nm grating structures, thus offering advantages in uniformity as a comparison substrate. Additionally, the Si substrate of the Si/SiO2 multilayer film grating technology provides atomic-level interplanar spacing information for the calibration of grating period or linewidth,[7] thereby meeting the stringent requirements for both Si lattice constant and self-traceable grating comparisons.

    Inspired by the above factors, this study proposes to utilize multilayer film gratings as the comparison substrate and employ AFM and TEM as measurement tools to conduct a comparative study on the metrological calibration accuracy between self-traceable gratings based on Cr atomic transition frequency and Si lattice constants. This research leverages the periodic accuracy and consistency of self-traceable gratings to directly calibrate the AFM, ensuring the traceability and accuracy of its measurement results. By customizing the linewidth measurement of the TEM based on the Si lattice constant, the traceability and accuracy of the TEM measurements can be guaranteed. Additionally, the excellent uniformity of multilayer film gratings ensures structural consistency in the comparative measurements. The research aims to verify the accuracy and consistency levels of different value transfer methods through comparative studies. This study holds significant importance for the integrated circuit field regarding the hybird measurement and metrology of gratings and linewidths.

2.   Principal, fabrication and measurements
  • Using Cr as a case study, the self-traceable grating technology materializes the physical standard wavelength by precisely locking onto the specific transition frequency (7S3–7P4) of Cr atoms, thereby establishing traceability of the grating pitch to the wavelength value corresponding to this transition frequency.[8] The manufacturing process entails evaporating Cr in a vacuum environment at approximately 1600 °C to generate a Cr atomic beam. A laser, intentionally detuned relative to the transition frequency, is employed to cool the lateral velocity of the selected Cr atomic beam through the Doppler cooling principle. The collimated Cr atoms are subsequently converged within the laser standing wave field, where deposition occurs at the antinode or node positions according to the wavelength detuning, resulting in the formation of a periodic one-dimensional grating structure characterized by high accuracy and reliability, as illustrated in Fig. 2. Two laser beams are stabilized near the 7S3→7P4 transition frequency of Cr atoms, leading to a grating pitch that is half the wavelength of the standing wave field, which is traceable to the transition frequency of the Cr atom. Consequently, the Cr grating pitch exhibits self-traceable properties. The one-dimensional self-traceable grating produced via this method demonstrates exceptional accuracy and consistency, with measurements on the order of 0.001 nm.[3,9]

  • The manufacturing process of multilayer film gratings is a technique that translates the thickness values of deposited film layers into linewidth and grating pitch measurements.[10] The preparation workflow is depicted in Fig. 3. Utilizing ion beam sputtering technology, we sequentially deposit alternating layers of 15 nm Si and 10 nm SiO2 to construct the multilayer film structure. In this configuration, the 15 nm Si layer is designated as the linewidth, while the 10 nm SiO2 layer acts as a spacer, with the entire structure repeated for 60 cycles. This arrangement ensures stability and exceptional uniformity among the film layers across a broad range. Following this, two samples with identical deposited film layers undergo processes of bonding, cutting, grinding, and polishing. Wet etching is then performed using an HF solution to produce standard samples, which are subsequently mounted on a substrate for further application.

  • For grating images obtained via AFM, the gravity center method is commonly employed to calculate the grating period.[11,12] This method centers on determining a threshold line between the top and bottom lines of the grating structure based on a preset percentage p%, and subsequently calculating the centroid of the upper structure above this threshold line to accurately ascertain the average grating pitch and positional deviation, as illustrated in Fig. 4. The specific process is as follows: (i) Removal of ripple artifacts: The grating profiles obtained from AFM measurements often exhibit undesirable ripple artifacts coupled with the actual profiles, leading to erroneous identification of structural boundaries and impacting measurement accuracy. Therefore, this study first employs a phase-corrected low-pass filter to process the profile data, effectively recognizing and filtering out ripple artifacts. (ii) Profile line interpolation: Given that the grating profile is sampled as discrete data at a certain pixel density, the limitation in pixel quantity makes it challenging to accurately determine the boundary structure of the grating. We perform linear interpolation on the profile line based on the measured data points and the instrument noise level, thereby enhancing the accuracy of boundary recognition. (iii) Calculation of average grating pitch: By utilizing the spatial coordinates of the centroids of each grating structure within the scanning range, we apply a linear fitting method to determine the slope of the fitting line and subsequently compute the average grating pitch.

3.   Experiments
  • This experiment utilizes self-traceable national primary standard materials. The PTB measured the average pitch difference between any two self-traceable gratings, yielding a result of 0.001 nm with an uncertainty of 0.008 nm. The calibrated instrument is a commercial AFM placed in a temperature-controlled room with an ambient temperature of (20±1) °C, temperature fluctuations of ≤ 0.2 °C/h, humidity of 40 ± 20% RH, and humidity fluctuations of ≤ 5% RH. Prior to calibration, the AFM stage is kept level, and the self-traceable grating sample is allowed to rest on the stage for 30 minutes. When selecting the calibration position, care is taken to avoid areas with pits, scratches, or contamination on the grating.

  • In small-scale scanning ranges, measurement errors arise from issues such as the nonlinear effects of the AFM scanning probe and the probe expansion effects, creating an urgent need for calibration. The self-traceable grating exhibits extremely high precision in pitch, exceptional consistency between samples, and well-defined line edges, thereby meeting the nonlinear calibration requirements for AFM in small-scale scans. We employed a one-dimensional Cr self-traceable grating to calibrate the nonlinear errors of the commercial AFM. The images obtained from scanning a 1 μm × 1 μm area of the self-traceable grating are illustrated in Fig. 5.

    Based on the gravity center method, five positions are randomly selected from each of the four scanned images, and the average period value for each image is calculated, as shown in Table 1. This value was then compared with the average period obtained from the uncalibrated AFM within the same scanning range to derive the calibration factor K, as expressed by the following formula:

    where dCr is the period of the self-traceable grating obtained using the gravity center method, and dAFM is the period measured by the uncalibrated AFM. Given that the pitch of the self-traceable grating is d = 212.782 ± 0.008 nm (k = 2), the calibration factor K is determined to be 0.9922.

  • Following the precise calibration of the AFM, we employed it to measure the period of Si/SiO2 multilayer film grating samples with a nominal value of 25 nm. Scanning was conducted at ten randomly selected positions across the top, middle, and bottom regions of the sample. To account for the differing nonlinear effects associated with various scanning ranges, we chose a scanning area of 1 μm × 1 μm, which corresponds to the scale of the self-traceable grating used during the AFM calibration, as depicted in Fig. 6.

    The acquired grating scanning images were imported into Matlab software, where the gravity center method described in Subsection 2.3 was employed to calculate the period of the multilayer film gratings. To convert these measurements into final AFM results, we multiplied the measured values by the calibration factor K. The calculated measurement values for the period of the multilayer film gratings are presented in Table 2. Based on the calibration using self-traceable gratings, the final linewidth value determined by the AFM is 24.40 nm, with a standard deviation of 0.11 nm.

  • To cross-validate the integrity of the self-traceable system and the feasibility of calibration, a Si lattice-calibrated TEM was employed as the verification method. The measurements were performed using a Talos F200X G2 transmission electron microscope. Given that the multilayer film samples were composed of amorphous Si and lacked information on Si crystal planes necessary for direct measurement, it was essential to extract Si crystal plane information from the underlying Si substrate. For traceable measurements, the structure of the multilayer film Si grating and the Si substrate was displayed at a magnification of 36k, with localized measurements conducted at a magnification of 500k, as illustrated in Fig. 7(a). In the high-magnification TEM measurement mode, accurate Si lattice periods were derived from the interference pattern of the transmitted electron beam, yielding TEM images that are traceable to SI units. This approach utilizes the Si lattice spacing as a reproducible method for defining the meter at the nanoscale. Figure 7(b) illustrates the structure of the multilayer film grating at a magnification of 22.5k.

    The specific calibration procedure is as follows. The micrograph of the Si lattice spacing obtained from the TEM at a magnification of 500k was analyzed using the integrated software Digital Micrograph. Ten lattice spacings were selected from the scanned image, as shown in Fig. 8. A total of ten measurements were repeated to determine the calibration factor for the Si lattice-calibrated TEM:

    where d10 is the 10 nm calibration standard, px10 is the number of pixels corresponding to the 10 nm standard, pxSi is the number of pixels corresponding to the lattice periods of Si lattice TSi, and dSi is the lattice length of Si in the 〈111〉 crystallographic direction, with dSi = 0.314 nm. Thus, the calibration factor k = 0.987.

    At the same scale (10 nm), we conducted multiple scans of the multilayer structure using TEM. Figure 8(b) presents a single multilayer structure at a magnification of 500k, with the image imported into Matlab for analysis. By measuring the number of pixel points corresponding to the period of the single multilayer film and applying the previously calculated calibration factor, we obtained the final calibrated result for the multilayer film based on the Si lattice calibration as 24.34 nm, with a standard deviation of 0.14 nm. The specific measurement results are detailed in Table 3.

  • We employed the AFM calibrated using the self-traceable grating to ascertain the period of the multilayer film grating and conducted a comparative analysis with the results obtained from TEM calibrated against the Si lattice constant. Our results reveal a measurement value deviation of 0.06 nm between the two calibration methods, corresponding to a relative error of 0.25%. This consistency between the measurement outcomes substantiates the coherence of value transfer within both the self-traceable grating traceability chain and the Si lattice traceability chain.

4.   Conclusion and perspectives
  • In the realm of nano-manufacturing, precision measurement technology is pivotal for ensuring product quality and performance. Achieving high accuracy and reliability necessitates consistency across various measurement systems. This study introduces a novel shortened traceability chain based on the Cr atomic transition frequency (7S3→7P4) and utilizes self-traceable gratings fabricated via atomic lithography to calibrate AFM nonlinearities and probe expansion effects, specifically for determining the period of a Si/SiO2 multilayer film grating with a nominal value of 25 nm. Validation results show a measured value of 24.40 nm using the self-traceable grating calibration system, closely aligning with the TEM result of 24.34 nm calibrated against the Si lattice constant. This agreement confirms the accuracy of the self-traceable shortened traceability system.

    In this static calibration process, the method not only enhances calibration convenience but also enables direct calibration in real working environments, effectively avoiding potential errors from equipment relocation. Additionally, this approach offers nondestructive measurement, crucial for preserving high-value, fragile, or irreplaceable samples. It allows precise measurement without compromising sample integrity, thus ensuring the authenticity of measurement data. Furthermore, self-traceable gratings can be applied in dynamic calibration by integrating them into grating interferometers, enhancing measurement capability in the micro-displacement domain.[21] However, current limitations in preparation methods and principles restrict the grating area, posing certain constraints. To meet the need for larger measurement areas, the research team is exploring splicing atom lithography[22] to further expand the measurement range.

Figure (8)  Table (3) Reference (22)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return