Open Access
Add to BibSonomy
Abstract
In recent years, microsphere-assisted microscopy (MAM) and atomic force microscope (AFM) have been rapidly developed to meet the measurement needs of microstructures. However, the positioning of microspheres, the inability of AFM to touch the underlying sample through the transparent insulating layer, and the challenge of AFM fast positioning limit their use in practical measurements. In this paper, we propose a method that combines MAM with AFM by adhering the microsphere to the cantilever. This method allows MAM and AFM to work in parallel, and their imaging positions can correspond with each other. We use this method to measure memory devices, and the results show that MAM and AFM yield complementary advantages. This approach provides a new tool for analyzing complex structures in devices and has potential for wide application.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The rapid progression in microsphere-assisted microscopy (MAM) [1,2] and atomic force microscopy (AFM) [3,4] technologies in recent years has significantly addressed the measurement demands of intricate microstructures [5,6], especially within semiconductor devices [7]. The microsphere enhances the resolution of optical microcopy (OM) to 50 nm [8] with the advantages of non-contact surface measurement [9] and the simplicity of its manufacturing process [10]. AFM is also widely used because of its high resolution and surface profile measurement capability [11]. Nevertheless, these technological advancements come with their own set of challenges. The microsphere requires precise positioning to accurately select the region of interest [12,13]. Moreover, the resolution procured by the microsphere may not meet the increasingly demanding needs of microstructures within semiconductor devices [14]. AFM's limitations are seen in the measurement of multi-layer samples due to its contact measurement technique, which prohibits the AFM from analyzing the structure beneath the transparent insulating layer of the devices. Moreover, the point-scanning method limits AFM's scanning efficiency, making it unsuitable for applications requiring a detailed search within a large field of view [15].
Comparing these two techniques can yield complementary advantages. Researchers have adopted a strategy wherein the microsphere is adhered to the cantilever, and AFM is used to control the microsphere's position for optical imaging [7,12]. This approach procures high-quality and efficient imaging [16], yet it does not permit AFM measurements. Another strategy involves combining the microsphere with AFM by growing the probe on the microsphere using a focused ion beam (FIB) [17], which successfully allows OM and AFM to work in parallel. However, the tip is grown on the bottom of the microsphere resulting in the microsphere only being used for imaging but not for measurement. Additionally, ensuring the quality of non-standard AFM probes presents a significant challenge.
In this paper, we propose a technique that combines MAM with AFM by the microsphere probe. An AFM thin head is adopted to accommodate a high magnification objective [18]. The microsphere is adhered to the cantilever using UV-curable adhesive, with the original tip extending five hundred nanometers beyond the bottom of the microsphere to ensure accurate AFM measurements. Meanwhile, the force between the microsphere and the sample remains negligible when the microsphere is detached several hundred nanometers, thus ensuring the stability of the microsphere. This technique accomplishes precise positioning of the microsphere, enhancing the imaging quality, and resolving the rapid AFM localization issue. It also enables the observation of microstructures beneath the transparent insulating layer that cannot be reached by the tip. The efficacy of this method is corroborated by measuring memory devices.
2. Experimental setup
The combination of MAM with AFM presents various challenges. First, a conventional AFM system with a vertical light path is structurally unable to accommodate a high magnification objective. Second, for efficient operation, both MAM and AFM need to work in parallel and their target areas need to correspond. Finally, the microsphere must maintain a stable position during scanning measurements as its distance from the sample influences the imaging quality of the microsphere [19]. To overcome these challenges, a skillful design of the optical path system, the AFM head, and the microsphere probe is paramount. A vertically illuminated optical path system is employed to guarantee imaging quality. An AFM thin head is adopted to accommodate a higher magnification objective. Furthermore, a unique adhesion method ensures the concurrent operation of MAM and AFM. FIG. 1(a) shows the combined system of MAM and AFM. A standard reflective optical path is established vertically to capture the optical signal reflected from the sample. The optical path's primary components include a 50× objective lens, zoom lens, and CCD. A white light source illuminates the sample and functions as the measurement light for the optical signal. FIG. 1(b) shows the working structure of MAM and AFM, where the MAM illumination light and the AFM laser operate without mutual interference. FIG. 1(c) shows the microsphere probe along the cantilever's long-axis view. The microspheres can form a magnified virtual image [20] of the sample when detached several hundred nanometers [17], and the tip can make contact with the sample. FIG. 1(d) shows the AFM thin head, which accommodates short working distances and high magnification objectives [18]. By utilizing a multi-reflector folding optical path, the total thickness of the head is reduced to 7.3 mm. Samples are positioned on a piezoelectric actuator with triaxial displacement to ensure precise control.
Fig. 1. Combined system. (a) The overall look of the system with OM and AFM. The physical image can be seen from the Supplement 1 Fig. S3. (b) Working method of the microsphere probe with optical path and AFM thin head. (c) Details of microsphere probe and sample. The microsphere is 539.8 nm from the surface and forms a magnified virtual image of the sample. (d) Thin head made by our team. The main components include frame of the plate head, probe holder, probe, laser adjustment, quadrant photodetector (QPD) adjustment, XY nanopositioning stage, Z nanopositioning stage, head position adjustment.
The microsphere is adhered to the side of the AFM cantilever using UV-curable adhesive as shown in Fig. 1(c) and Fig. 2(a). The OTESPA-R3 AFM probe is selected due to its high spring constant (k-value) that can withstand the microsphere's weight and its shape minimally impacts optical imaging. The UV-curable (NOA63) adhesive is applied to the side of the cantilever, reducing the contact area between the adhesive and the microspheres, ensuring the cantilever minimally affects the optical imaging of the microsphere. To ensure AFM measurements, the tip must extend several hundred nanometers beyond the microsphere. Subsequent experiments on distance control between the microsphere and the sample reveal that a distance of approximately five hundred nanometers is suitable. In order to maintain this precise distance, we have devised a standardized process for adhering the microsphere. We create a slot on the silicon substrate with 3 µm in width, 100 µm in length, and 0.6 µm in depth, using FIB. We then selectively use silica microspheres with 25 µm diameter (from Tianjin Baseline Technology) for each adherence process to ensure that the tip protrudes approximately 500 nm beyond the bottom of the microsphere. The microsphere is maneuvered onto the slot by a probe controlled by AFM, and the probe intended for adhesion is first directed to scan the slot and align itself centrally within it. We subsequently use AFM to guide the probe towards the microsphere in a direction parallel to the slot, promoting adhesion. This process guarantees the consistent geometrical relationship between the probe and microsphere with each adherence process. The adhesion process of the microspheres to the probes is shown in Fig. 2(a,b). In measurements of sample devices, a fluctuation of five hundred nanometers suffices to meet most measurement needs.
Fig. 2. (a) Scanning electron microscope (SEM) of the microsphere probe from perpendicular plane of the cantilever. (b) Geometric relationship between the probe and the microspheres.
Our system can work in both tapping and contact modes, each with its own pros and cons. In contact mode, MAM and AFM can work in parallel, with the closed-loop feedback of AFM ensuring a constant distance between the microsphere and the sample. However, since the standard contact probe has a relatively low spring constant and cannot support the weight of the microsphere, we need to operate a tapping probe in contact mode, which can lead to rapid wear of the tip. We only adopt this mode when concurrent operation of both is required. In tapping mode, due to the unstable distance between the microsphere and the sample, MAM and AFM cannot work in parallel. Nevertheless, the microsphere probe can still eliminate the need for probe replacement and repositioning. We use ANSYS to simulate the connection between the microsphere and the cantilever, calculating the resonance frequency and twisting of the microsphere probe (as shown in Supplement 1, Fig. S1 and Visualization 1). The results show that the resonant frequency of the microsphere probe is 124.3 kHz, which is close to the experimentally measured 134.5 kHz, and there is no significant twisting at this frequency. The AFM images of the microsphere probe and the standard probe in tapping mode show that the introduction of the microsphere has virtually no impact on the AFM measurement results (as shown in Supplement 1, Fig. S2).
3. Experiments and results
To investigate the imaging capabilities of the detached microsphere, an experiment is conducted to control the distance between the microsphere and the surface. This experiment also aids in determining the optimal length for the tip to exceed the microsphere during the fabrication of microsphere probes. As shown in Fig. 3(a,b), a tipless cantilever is employed, and the 25 µm microsphere is adhered to it, ensuring that the microsphere's bottom can touch the surface. The contact mode is used when engaging to keep the microsphere on the surface, which is defined as the zero-distance position. A z-directional piezoelectric actuator is used to separate the microsphere from the surface and to image the resolution target at different positions, with distances ranging from 0 to 1 µm at 50 nm intervals. The resolution target (2017a_USAF_Dmnd, Ready Optics) is imaged using a microsphere probe. Due to the discontinuous resolution target groups, the results do not visually reflect the microsphere's resolution at varying distances. Hence, we select fringe contrast as a quantitative evaluation of the imaging quality. Fringe contrast is calculated by the contrast formula (C = (I1-I2)/(I1 + I2), C: contrast, I1: maximum intensity I2: minimum intensity). Imaging results of Group 11 Element 5 with line widths of 154 nm are shown in Fig. 3(c). Experimental results show a significant drop in fringe contrast at a distance between 550 nm and 650 nm, where the image becomes noticeably blurred. For most structures, such as semiconductor devices, a 500 nm tip suffices for measurement. These findings led us to determine the distance between the tip and the microsphere to be approximately 500 nm. For a more intuitive comparison of fringe contrast, the intensity curves at distances of 500 nm and 650 nm are illustrated in Fig. 3(d-f). As the intensity in optical images has no actual physical meaning, we have normalized all the intensity images in optics. The color bars for all optical images ranges from 0 to 1, following the parula scale. These curves reveal a significant decrease in imaging quality when the distance exceeds 650 nm.
Fig. 3. Distance control experiment. (a) Detach the microsphere from the surface by Z-directional piezoelectric actuator. (b) Image changes as the microsphere is detached. (c) Fringe contrast of different distances from 0 to 1000 nm. (d) Image at 500 nm. (e) Image at 650 nm. (f) Direct intensity comparison at 500 nm and 650 nm.
To calibrate the system's resolution, Group 11 Elements 4, 5, and 6, with line widths of 173 nm, 154 nm, and 137 nm (periods of 345 nm, 308 nm, 274 nm) of the resolution target are imaged with the microsphere probe. The results are shown in Fig. 4(a,b). We first focus the OM on the sample surface and engage the probe to facilitate contact between the tip and the sample. The microsphere is delicately adjusted in the X-Y directions by piezoelectric actuator, ensuring optimal positioning over the target imaging area. The microsphere forms a magnified virtual image of the sample, which is located beneath the sample plane. The objective plane is adjusted to the position of this virtual image, transforming it into a real image that can be collected by the CCD. Due to the microsphere's inhomogeneous light field [21], we subtract the standard plane measurement at the same focal position from the sample measurement result to eliminate this error. To demonstrate the imaging effect, we apply a first-order Gaussian Blur to eliminate noise, then use bicubic interpolation to expand the pixel points. As shown in Fig. 4(b), the system can easily resolve microstructures with a linewidth of 137 nm. The OM's resolution, calculated by the Abbe diffraction limit (δ=0.61λ/(NA), where δ denotes resolution, λ is wavelength, and NA is the numerical aperture), is 545 nm. This indicates that the resolution can be significantly enhanced by the microsphere. Fig. 4(c) depicts the AFM measurement results using the same microsphere probe in tapping mode. The linewidth period data match the optical measurements, as shown in Fig. 4(d-g).
Fig. 4. Results of the microsphere probe on the resolution target. (a) Image of the microsphere probe. (b) Rectangular area in a. MAM images of Group 11 Element 4, Element 5, Element 6. (c) AFM image of the microsphere probe. (d,e) Intensity curve of Group 11 Element 5, Element 6 by the microsphere. (f,g) Height curve of Group 11 Element 5, Element 6 by AFM.
To demonstrate the microsphere probe's capabilities, a memory device is measured in Fig. 5(a-e). Fig. 5(a) displays the general shape of the periodic memory cell structure, with horizontal and vertical lines in the upper layer serving as the bus and the microstructure in the lower layer acting as the dividing line. Transparent insulating layers exist between different circuit layers in the device. OM is unable to identify the lower structure due to its resolution, as shown in Fig. 5(b). We carefully move the microsphere to the corresponding area by controlling the position of the microsphere probe. This enable the microsphere to reach the optimal imaging position, clearly identifying the structure with a minimum width of 150 nm, as shown in Fig. 5(c). The AFM tip cannot reach the structure underneath due to its contact measurement, as shown in Fig. 5(d). The height profile lines in Fig. 5(e) reveal that the transparent insulation layer is at a height of -200 nm. SEM is unable to measure the structure since its electron beam cannot penetrate the insulating layer.
Fig. 5. Results of the microsphere probe on the memory device. (a) The image of the bus and the dividing line without the microsphere probe. (b) Enlarged image of the rectangle in a. (c) Image of the lower layer dividing line by MAM at the same location as b. (d) Image by AFM of c. The black rectangles in c and d are the corresponding areas. (e) The height curve in d. The red line shows the position of the transparent insulation layer. (f) Complex circuit structure without the microsphere probe. (g) Image of the rectangle in f by MAM. (h) Image by AFM of g. (i) Image of the rectangular in h. The height profile of the node.
The low magnification and resolution of OM on AFM system do not allow for rapid localization of microstructures, and wide area scanning is inefficient for AFM. The microsphere probe enhances the magnification and resolution of the OM system with minimal effect on AFM, aiding AFM in rapid localization. The experiment in Fig. 5(f-i) demonstrates this advantage. As shown in Fig. 5(f), this is a complex circuit structure on a memory device, featuring nodes with a 100 nm width connected to lower layers. During a large MAM scan of the device using a microsphere probe, we observe the nodes. However, due to the small size of the nodes, MAM is unable to image them clearly, as shown in Fig. 5(g). We use AFM to precisely depict the nodes, as shown in Fig. 5(h,i).
From the above two experiments we can see the advantages of combining MAM with AFM. Utilizing a microsphere significantly enhances the OM resolution, providing superior image quality. Adhering the microsphere to the AFM cantilever allows precise positioning of the microsphere which makes it easy to find the region of interest and also the best position for microsphere imaging to increase the imaging quality. AFM is an advanced measurement methodology, and the microsphere probe preserves the original probe of the AFM. The incorporation of microspheres resolves the issues related to low magnification and resolution of OM within the AFM system, aiding the rapid localization of microstructures and the imaging of structures beneath the transparent layer. The combination of these techniques is designed to retain the distinct advantages of both, while compensating for their respective limitations. This synergistic approach to microscopy maximizes the strengths of each method and creates a powerful tool for high-resolution imaging and measurement.
4. Conclusion
In this paper, we presented the successful combination of MAM and AFM to develop a comprehensive system, capitalizing on the strengths of both techniques. The microsphere was adhered to the side of the cantilever using a UV-curable adhesive, and the probe tip was designed to exceed five hundred nanometers than the base of the microsphere. This strategic configuration ensured the contact of the original probe tip to the sample while facilitating resolution-enhanced imaging by the microsphere. We did a distance control experiment to identify the optimal imaging position when the microsphere was detached from the surface. The magnification of the microsphere is 3.2× and the resolution of the combined system was validated using a standard target to be 137 nm, which demonstrated a substantial increase in resolution compared to the system devoid of a microsphere. The experimental results, derived from the measurements of memory devices, attested to the capacity of MAM and AFM yield complementary benefits. This combined method enables the acquisition of both mechanical and optical signals from the sample, thus introducing a novel and robust method for the investigation of intricate structures, multilayered and multi-signal samples, such as MEMS devices and semiconductor devices. We are optimistic that this innovative method will hold significant implications for the future, especially in the domains of device manufacturing and inspection, facilitating improved efficiency and precision in these areas.
Funding
National Natural Science Foundation of China (52275540, 61973233).
Acknowledgments
This work was supported by the National Natural Science Foundation of China [grant numbers 52275540, 61973233].
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. X. Hao, C. Kuang, X. Liu, H. Zhang, and Y. Li, “Microsphere based microscope with optical super-resolution capability,” Appl. Phys. Lett. 99(20), 203102 (2011). [CrossRef]
2. L. Chen, Y. Zhou, Y. Li, and M. Hong, “Microsphere enhanced optical imaging and patterning: From physics to applications,” Appl. Phys. Rev. 6(2), 021304 (2019). [CrossRef]
3. G. Binnig, C. F. Quate, E. L. Gi, and C. Gerber, “Atomic Force Microsope,” Physical Review Letters. 56(9) (1986).
4. F. J. Giessibl, “Advances in atomic force microscopy,” Rev. Mod. Phys. 75(3), 949–983 (2003). [CrossRef]
5. Y. Yan, L. Li, C. Feng, W. Guo, S. Lee, and M. Hong, “Microsphere-coupled scanning laser confocal nanoscope for sub-diffraction-limited imaging at 25 nm lateral resolution in the visible spectrum,” ACS Nano 8(2), 1809–1816 (2014). [CrossRef]
6. S. T. P. Errin, A. U. L. E. Oï, S. Y. L. Ecler, P. I. P. Feiffer, I. V. A. N. K. Assamakov, A. N. N. Olvi, E. D. H. AEggström, and P. A. U. L. M. Ontgomery, “Microsphere-assisted phase-shifting profilometry,” Appl. Opt. 56(25), 7249–7255 (2017). [CrossRef]
7. F. Wang, L. Liu, H. Yu, Y. Wen, P. Yu, Z. Liu, Y. Wang, and W. J. Li, “Scanning superlens microscopy for non-invasive large field-of-view visible light nanoscale imaging,” Nat. Commun. 7(1), 13748 (2016). [CrossRef]
8. Z. Wang, W. Guo, L. Li, B. Luk’Yanchuk, A. Khan, Z. Liu, Z. Chen, and M. Hong, “Optical virtual imaging at 50 nm lateral resolution with a white-light nanoscope,” Nat. Commun. 2(1), 218 (2011). [CrossRef]
9. A. Darafsheh, C. Guardiola, A. Palovcak, J. C. Finlay, and A. Cárabe, “Optical super-resolution imaging by high-index microspheres embedded in elastomers,” Opt. Lett. 40(1), 5 (2015). [CrossRef]
10. B. Yan, Z. Wang, A. L. Parker, Y. Lai, P. John Thomas, L. Yue, and J. N. Monks, “Superlensing microscope objective lens,” Appl. Opt. 56(11), 3142 (2017). [CrossRef]
11. R. Zhang, S. Wu, L. Liu, X. Fu, S. T. Gao, and X. D. Hu, “Adaptive-angle scanning method for 3D measurement with atomic force microscopy,” Meas. Sci. Technol. 30(9), 095005 (2019). [CrossRef]
12. L. A. Krivitsky, J. J. Wang, Z. Wang, and B. Luk’Yanchuk, “Locomotion of microspheres for super-resolution imaging,” Sci. Rep. 3(1), 3501 (2013). [CrossRef]
13. B. Yan, Y. Song, X. Yang, D. Xiong, and Z. Wang, “Unibody microscope objective tipped with a microsphere: design, fabrication, and application in subwavelength imaging,” Appl. Opt. 59(8), 2641 (2020). [CrossRef]
14. H. Yang, R. Trouillon, G. Huszka, and M. A. M. Gijs, “Super-Resolution Imaging of a Dielectric Microsphere Is Governed by the Waist of Its Photonic Nanojet,” Nano Lett. 16(8), 4862–4870 (2016). [CrossRef]
15. Y. Zhou, G. Shang, W. Cai, and J. E. Yao, “Cantilevered bimorph-based scanner for high speed atomic force microscopy with large scanning range,” Rev. Sci. Instrum. 81(5), 053708 (2010). [CrossRef]
16. J. Zhou, Z. Lian, C. Zhou, S. Bi, and Y. Wang, “Scanning microsphere array optical microscope for efficient and large area super-resolution imaging,” Appl. Opt. 56(25), 7249 (2017). [CrossRef]
17. T. Zhang, H. Yu, J. Shi, X. Wang, H. Luo, D. Lin, Z. Liu, C. Su, Y. Wang, and L. Liu, “Correlative AFM and Scanning Microlens Microscopy for Time-Efficient Multiscale Imaging,” Adv. Sci. 42(1), 577–580 (1993). [CrossRef]
18. N. Lu, S. Xiao, R. Zhang, J. Liu, L. Ma, and S. Wu, “Thin head atomic force microscope for integration with optical microscope,” Rev. Sci. Instrum. 93(8), 083702 (2022). [CrossRef]
19. M. Duocastella, F. Tantussi, A. Haddadpour, R. P. Zaccaria, A. Jacassi, G. Veronis, A. Diaspro, and F. De Angelis, “Combination of scanning probe technology with photonic nanojets,” Sci. Rep. 7(1), 3474 (2017). [CrossRef]
20. S. Lecler, S. Perrin, A. Leong-Hoi, and P. Montgomery, “Photonic jet lens,” Sci. Rep. 9(1), 4725 (2019). [CrossRef]
21. S. Perrin, Y. J. Donie, P. Montgomery, G. Gomard, and S. Lecler, “Compensated Microsphere-Assisted Interference Microscopy,” Phys. Rev. Appl. 13(1), 014068 (2020). [CrossRef]
