Open Access
Add to BibSonomy
Abstract
Flip-chip bonding is a key technology for infrared focal plane array (IRFPA) detectors. Due to the high cost of device preparation, the ultra-large array infrared detector cannot be directly used for the flip-chip bonding experiment, and the connectivity rate cannot be measured. To evaluate the flip-chip bonding process, a test device which has the same interconnecting structure as current IRFPA detectors is proposed. Indium bumps are electrically extracted to test electrodes. Electrical measurements were performed to characterize the connection and adhesion of the indium bumps and to calculate the connectivity rate. The electrical connectivity characteristics of the test devices correspond to the observation results of the indium bump extrusion, effectively detecting the interconnecting anomalies such as disconnection, adhesion, overall misalignment, etc., and verifying the feasibility of the test method. The test device has similar multi-layer components and thermal properties as HgCdTe infrared detector for process evaluation and post-processing experiment. The connectivity rate of the test device is up to 100%, and remains above 99% after thermal recycle experiment. The contact resistance of the interconnecting structure is calculated to be about 31.84 Ω based on the test results.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Infrared focal plane array (IRFPA) detectors have a wide range of applications in astronomy, meteorology, medicine and other fields. At present, the third generation of IRFPA detectors are developing in the direction of ultra-high resolution [1–5], the focal plane specifications increase from 128 × 128, 320 × 256 to 2 k × 2 k, 4 k × 4 k, and the pixel size is also developing from 50 μm, 30 μm to 10 μm, 8 μm, 5 μm. The flip-chip bonding technology has been widely applied in the interconnection of photon detectors [6–14] including MEMS devices [15–17] and 3D integrated devices [18–20]. The interconnecting structure between the detector array and the readout circuit (ROIC) is not only mechanical connection, but also the only electrical contact that transmits photoelectric signals generated at detector array to ROIC at pixelated level [21–23]. Compared to the conventional IRPFA detectors, the ultra-large devices are more susceptible to interconnecting anomalies such as disconnection, adhesion, and overall misalignment as shown in Fig. 1, which limits the specification of ultra-large IRFPA detectors [24,25].
Fig. 1. Normal interconnections and interconnecting anomalies of IRFPA detectors. (a) Normal connection, indium bump is extruded and the cross section diameter increases; (b) Disconnection, the indium bump is not extruded, which still maintains the shape after deposition, and its cross-section is smaller than normally extruded ones; (c) Adhesion, the indium bump is severely extruded, flattened and connected to adjacent indium bumps, with a larger cross-section than other normally extruded indium bumps; (d) Overall misalignment, indium bumps fill into the gaps on the opposite side, and indium bumps are adhered to each other.
The electrical connectivity of flip-chip bonding process cannot be measured directly, but is generally estimated by testing complete IRFPA devices. The disconnection results in no response and an increase in the response of the surrounding pixels. However, this phenomenon can be influenced by factors within the pixel itself, such as material defects and process contamination. Adhesion and misalignment can cause electrical crosstalk between pixels, which can be identified through the IR device's imaging test. Identifying interconnecting anomalies and calculating connectivity rate through device testing requires the preparation of complete IRFPA devices [26–30], and devices screened for low connectivity rate can no longer be reworked and used for infrared detection. For ultra-large IRFFPA detectors, the fabrication of detector chips and readout circuits is costly and time-consuming, which makes IRFPA devices cannot be directly used for the flip-chip bonding experiment. In addition, the thermal stress caused by the varying thermal expansion coefficients of multi-layer components can result in significant deformation in large-scale devices [31–33] operating at low temperatures, ultimately leading to interconnecting structure failure. Fully-functional IRFPA devices are consumed for thermal recycle experiment and can’t be used for detecting application subsequently [34–37].
Recently, we proposed a method to evaluate the electrical connectivity of the flip-chip bonding process for ultra-large array HgCdTe infrared (IR) detectors. In this work, connectivity test devices have been designed and prepared, in which the internal indium bumps of the devices are electrically led out to the test electrodes, and electrical characteristics of the test devices are directly obtained by measuring the resistance between the electrodes. The electrical test results of the test devices including connection and interconnecting anomalies such as disconnection, adhesion and overall misalignment, are highly consistent with the observation of indium bump extrusion. The maximum connectivity rate of the test device reaches 100%, with over 99% after experiencing thermal recycling, and the contact resistance of interconnecting structure is calculated to be about 31.84 Ω. The test device has similar multi-layer components and thermal properties as HgCdTe IR detector, and can be used for mechanical experiments to evaluate the reliability of the interconnecting structure.
2. Experiment
As shown in Fig. 2, test device is composed of the interconnection of the testing component (simulates ROIC) and the bonding component (simulates IR array), including the metal wire structure, the insulating protective layer and the indium bump interconnecting structure. The indium bump interconnecting structure is an indium bump array that has the same specifications as current infrared detectors, where the pitch size of the array is 10 μm and the diameter of indium bump is 6 μm. A section of the indium bump is linked to the metal wire beneath through the aperture in the insulating protective layer, and is electrically joined to the test electrodes.
The connectivity rate of the test device is defined as the percentage of test points that are both connected and non-adhered in relation to the total number of test points. The connection test circuit is composed of a metal wire from the common electrode to the top of the indium bump (to be tested), through the indium bump (to be tested), and a metal wire from the bottom of the indium bump (to be tested) to test electrode 1 as shown by the blue dotted line in Fig. 2. Whether connection test the circuit is closed or not reflects whether the indium bump (to be tested) is connected or not. The adhesion test circuit includes: connection from the test electrode 1 to the bottom of the indium bump (to be tested) via the metal wire, through the possible adhesion between the indium bump (to be tested) and the indium bumps (adjacent), and connection from the metal wire at the bottom of the indium bumps (adjacent) to the test electrode 2 as shown by the red dotted line in Fig. 2, and whether the circuit is closed or not responds to whether the indium bump (to be tested) is adhered to the indium bumps (adjacent) or not.
Figure 3 shows the schematic and physical diagram of a test device. Transparent substrate is chosen for the testing component, and the metal wire structure and indium bump extrusion of the test point can be directly observed through the back of testing component. At the test point shown in Fig. 3(d), the metal wires connecting one indium bump (to be tested) and eight indium bumps (adjacent) are essentially complete, and there are no obvious dislocations, adhesions, absence, or other abnormal phenomenon. The test device is equipped with 411 identical test points to identify interconnecting anomalies in different areas.
Fig. 3. Schematic and physical diagram of test device. (a) Test device design diagram (b) Structure diagram of a single test point (c) Test device physical diagram (d) Photo of a single test point
The test device was designed to maximize the proximity to the HgCdTe IRFPA detector, including the material selection and the post-processing. The substrate of the bonding component is silicon wafer with a layer of CdTe. The substrate of testing component is initially gemstone, making it possible to observe the extrusion of indium bumps, which has been replaced by silicon wafer later. After the flip-chip bonding process, the test device is underfilled with low-temperature epoxy resin and patched to a silicon carbide composite under the same process as the IR detector. With similar component structure and mechanical properties as the IR detector, test device can be used for thermal cycling experiment to simulate the repeated temperature switching process (between room temperature and liquid nitrogen temperature) and evaluate the reliability of the interconnecting structure.
3. Results and discussion
There is a distinct standard for determining the closure of the test circuit in the test device. Based on the design parameters including resistivity of wires, structure size and locations of test points, the resistance of test circuit can be calculated as:
The closed circuit's resistance ranges from 50 Ω to 3000 Ω by calculation, primarily due to the wire's resistance. As the circuit length ${l_{wire}}$ increases, there is a consistent in the increase of resistance value. The resistance value of the open test circuit is tested to be over 1 MΩ depending on the insulation of the substrates, which is three orders of magnitude different from that of the closed circuits. This provides an effective means of evaluating the connection and adhesion properties of indium bumps.
The electrical test results and Indium bump extrusion observation are present at Table 1 and Fig. 4 correspondingly. Figure 4(a) shows a connected and non-adhered test point. Figure 4(b) shows a disconnected and non-adhered test point, where the distance between the testing component and the bonding component is large, the metal wire on the bonding component is blurred and almost invisible, and the indium bumps are not connected. Figure 4(c) shows a connected and adhered test point, where the bonding pressure is higher than normal and the indium bumps are extruded and deformed to a degree significantly greater than that of the normally connected test point, and indium bumps are observed to be connected and adhered. Figure 4(d) shows an overall misaligned test point, the slip misalignment between the testing component and the bonding component is half a pixel, the top of the indium bumps on the bonding component can be observed between the gaps, and all test points of the overall misaligned device are adhered. The disconnected and adhered test point is quite rare, which is usually caused by reasons such as failures of indium bump preparation process and adhesion of wire structures.
Fig. 4. Indium bump extrusion observation. (a) Connected and non-adhered test point, the metal wire (marked with a blue oval) on the bonding component is observable, indium bumps and other wire are blocked. (b) Disconnected and non-adhered test point, due to a large distance between the testing component and the bonding component, the metal wire (marked with a blue oval) on the bonding component becomes blurry. This suggests that the indium bumps are not connected. (c) Connected and adhered test point, severely extruded indium bumps on the bonding component are partially visible. (d) Overall misaligned test point, the top of the indium bumps on the bonding component can be seen through the gaps.
Table 1. Typical electrical test results of test points
The electrical test results of connected and adhered test points and overall misalignment test points are similar as shown in Table 1. However, these two scenarios express differences in the test results of test devices and can be distinguished by electrical test. Adhesions are usually caused by severe extrusion related to surface warpage and cluster in a specific area, so most test points in other areas of the test device are non-adhered. Overall misalignment affects the entire test device, and all test points are adhered.
The electrical test results of connection and adhesion correspond to the microscopic observation of extrusion, implying that the test device accurately gauges the internal indium bump's connection and adhesion properties, as well as identifying interconnecting anomalies such as disconnection, adhesion and overall misalignment. Based on preliminary results, test devices with silicon substrates are fabricated to evaluate the connectivity rate (except test device 6).
Table 2 shows the test results of five test devices, test device 1, 2 and 3 have a better connectivity rate, up to 100%, and the test points of disconnection or adhesion are scattered, indicating that the interconnecting anomalies such as disconnection or adhesion are mainly related to the indium bumps themselves; Test device 4 seems to have an area where none of the indium bumps are connected, and those test points are all disconnected. The disconnection of a specific area is the main factor limiting the connectivity rate of test device 4; Some test points of test device 5 are severely extruded, and the adhered test points are concentrated in one area, indicating that the indium bumps in this area are severely extruded, limiting the connectivity rate of test device 5. The test results of test device 6 are all adhered, and observations indicate an overall misalignment as shown in Fig. 4(d), suggesting that the two components experienced a relative slip of half a pixel during the flip-chip bonding process.
Table 2. Test results of 6 test devices
The varying test results of the first five test devices are primarily attributed to surface warpage. Although the bonding and testing components were prepared using the same process, differences in surface warpage and flip-chip bonding processes affected the connectivity rate. Test device 1, 2, and 3 consist of matching components with combined PV values lower than 3 μm, resulting in a high connectivity rate. Substrates used for test device 4 and 5 exhibited more severe surface warpage, with a combined PV value of over 5 μm, leading to disconnections clustered in a specific area. Additionally, test device 5 was bonded at 100℃ while the others were at room temperature, resulting in less disconnection but more adhesions compared to test device 4.
A thermal recycling experiment has been conducted on test device 1, and the test results are displayed at Table 3. The underfilled test device is encapsulated in a liquid nitrogen dewar to ensure that the temperature process is not strongly sloping. The test device experienced thermal stress and deformation during each thermal recycle due to differential thermal expansion of its multi-layer components. Even after 30 thermal recycling cycles, the connectivity rate of test device 1 remains above 99%. The results suggest that the interconnecting structure remains reliable despite thermal deformation.
Table 3. Test results of test device 1 after thermal recycle experiment
As shown in Fig. 5, by normalizing the metal wires to the same cross-sectional area, the discounted wire length has a good linear relationship with the corresponding circuit resistance value, and the intercept of the fitted straight line is the resistance introduced by the interconnecting structure after deducting the wire resistance based on Eq. (1). The resistance of the indium bump can be obtained as:
and is calculated to be 3.55 × 10−2 Ω based on the structural dimension and electrical resistivity of indium, so the fitted intercept is the contact resistance ${R_{contact}}$ at the interconnecting interface and the interface between the indium bump and the metal wire. The contact resistances obtained from the three test devices in Table 4 are all very similar, with an average of 31.84 Ω.Table 4. Resistance fitting results of 3 test devices
4. Conclusion
In this work, connectivity test devices are designed and prepared, in which the top and bottom of indium bumps with their adjacent indium bumps are electrically connected to the test electrode through metal wires, and the resistance are measured to obtain the electrical connection and adhesion characteristics of the indium bumps and calculate the connectivity rate of the devices. The test devices have a connectivity rate of up to 100%, which remains above 99% even during thermal recycling experiments. Interconnecting anomalies such as disconnection, adhesion, and overall misalignment can be effectively detected, and corresponds to the indium bump extrusion observation. Based on the test results, the contact resistance of the interconnecting structure is calculated to be 31.84 Ω. The test device has the same specification of the indium bump interconnecting structure, similar multi-layer components and thermal properties as the current HgCdTe IRFPA detector, and the test results provide an evaluation method of the electrical connectivity of the flip-chip bonding process for ultra-large array infrared detector.
Funding
Innovative Project of Shanghai Institute of Technical Physics, Chinese Academy of Sciences (CX-457, CX-456).
Disclosures
The authors declare no conflicts of interest.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
1. A. Rogalski, Woodhead Publishing Series in Electronic and Optical Materials (Woodhead Publishing, 2020), Chap. 7.
2. A. Rogalski, P. Martyniuk, and M. Kopytko, “Challenges of small-pixel infrared detectors: a review,” Rep. Prog. Phys. 79(4), 046501 (2016). [CrossRef]
3. Z. Ye, H. Li, J. Wang, et al., “Recent hotspots and innovative trends of infrared photon detectors,” J. Infrared Millim. Waves 41(1), 2021342 (2022). [CrossRef]
4. A. Rogalski, “Recent progress in infrared detector technologies,” Infrared Phys. Technol. 54(3), 136–154 (2011). [CrossRef]
5. J. Caulfield and J. Curzan, “Small pixel infrared sensor technology,” Proc. SPIE10177, 1017725 (2017). [CrossRef]
6. S. Bernabé, C. Kopp, M. Volpert, et al., “Chip-to-chip optical interconnections between stacked self-aligned SOI photonic chips,” Opt. Express 20(7), 7886–7894 (2012). [CrossRef]
7. A. Schneider, N. Ghorbanian, J. Osborn, et al., “Bump deposition techniques for hybrid X-ray detectors,” J. Instrum. 18(06), C06009 (2023). [CrossRef]
8. C.-M. Kang, J. Rabinowitz, S. Bianconi, et al., “Toward sub-micron pixels for short-wave infrared imaging,” Semicond. Sci. Technol. 38(2), 025007 (2023). [CrossRef]
9. M. Hinderling, D. Sabonis, S. Paredes, et al., “Flip-chip-based microwave spectroscopy of andreev bound states in a planar josephson junction,” Phys. Rev. Appl. 19(5), 054026 (2023). [CrossRef]
10. A. S. Cross, Q. Zhou, A. Beling, et al., “High-power flip-chip mounted photodiode array,” Opt. Express 21(8), 9967–9973 (2013). [CrossRef]
11. P. Xiao, X. Tu, L. Kang, et al., “Design of double-slot antennas for terahertz array detectors in flip chip packaging,” Opt. Express 28(6), 8783–8791 (2020). [CrossRef]
12. E. C. C. Yeh, W. J. Choi, K. N. Tu, et al., “Current-crowding-induced electromigration failure in flip chip solder joints,” Appl. Phys. Lett. 80(4), 580–582 (2002). [CrossRef]
13. J. O. Son, S. Y. Na, Y. H. Kim, et al., “Wafer-level integration for hybrid chip of infrared detectors,” Proc. SPIE 12534, 65 (2023). [CrossRef]
14. P. Svihra, J. Braach, E. Buschmann, et al., “Development of novel single-die hybridisation processes for small-pitch pixel detectors,” J. Instrum. 18(03), C03008 (2023). [CrossRef]
15. D. A. B. Miller, “Dense two-dimensional integration of optoelectronics and electronics for interconnections,” Proc. SPIE 10292, 1029206 (1998). [CrossRef]
16. S. Tanaka, S.-H. Jeong, S. Sekiguchi, et al., “High-output-power, single-wavelength silicon hybrid laser using precise flip-chip bonding technology,” Opt. Express 20(27), 28057–28069 (2012). [CrossRef]
17. M. S. Yazici, B. Dong, D. Hasan, et al., “Integration of MEMS IR detectors with MIR waveguides for sensing applications,” Opt. Express 28(8), 11524–11537 (2020). [CrossRef]
18. C. Chen, D. Van Den Heuvel, M. Beggiato, et al., “Detection of bonding voids for 3D integration,” Proc. SPIE 12496, 43 (2023). [CrossRef]
19. Y. Bian, K. Ramachandran, Z.-J. Wu, et al., “3D integrated laser attach technology on a 300-mm monolithic CMOS silicon photonics platform,” IEEE J. Sel. Top. Quantum Electron. 29, 1–19 (2023). [CrossRef]
20. T. Mourier, N. M.-A. Berchet, J. Auffret, et al., “Advanced 3D integration TSV and flip chip technologies evaluation for the packaging of a mobile LiDAR 256 channels beam steering device designed for autonomous driving application,” in 2023 IEEE 73rd Electronic Components and Technology Conference (IEEE, 2023), pp. 239–246.
21. J. John, L. Zimmermann, P. De Moor, et al., “High-density hybrid interconnect methodologies,” Nucl. Instrum. Methods Phys. Res., Sect. A 531(1-2), 202–208 (2004). [CrossRef]
22. X. Zhang, C. Meng, W. Zhang, et al., “Design rule of indium bump in infrared focal plane array for longer cycling life,” Infrared Phys. Technol. 76, 631–635 (2016). [CrossRef]
23. R. K. Bhan and V. Dhar, “Recent infrared detector technologies, applications, trends and development of HgCdTe based cooled infrared focal plane arrays and their characterization,” Opto-Electron. Rev. 27(2), 174–193 (2019). [CrossRef]
24. S. Vähänen, H. Heikkinen, H. Pohjonen, et al., “Rework of flip chip bonded radiation pixel detectors,” Nucl. Instrum. Methods Phys. Res., Sect. A 591(1), 233–236 (2008). [CrossRef]
25. S.-K. Kang, D.-Y. Shih, and W. E. Bernier, Flip-Chip Interconnections: Past, Present, and Future BT - Advanced Flip Chip Packaging, (Springer US, 2013).
26. R. K. Bhan, S. K. Lomash, and K. C. Chhabra, “A method for determining the yield of indium bump contacts for integration in hybrid focal plane arrays at room temperature,” Infrared Phys. 34(1), 43–48 (1993). [CrossRef]
27. C. T. Nguyen, N. Mould, and J. L. Regens, “Dead pixel correction techniques for dual-band infrared imagery,” Infrared Phys. Technol. 71, 227–235 (2015). [CrossRef]
28. J. Nghiem, J. Jaeck, J. Primot, et al., “MTF measurements of a type-II superlattice infrared focal plane array sealed in a cryocooler,” Opt. Express 26(8), 11034–11045 (2018). [CrossRef]
29. A. M. Algamdi and A. A. A. Adas, “An improved technique for blind pixels’ detection of infrared focal plane array detector,” in Proceedings of 2018 1st International Conference on Computer Applications & Information Security (ICCAIS, 2018), pp. 1–6.
30. C. Liu, J. Fu, Y. Hou, et al., “A self-test method of structural failures of uncooled infrared focal plane array,” IEEE Trans. Semicond. Manufact. 34(3), 235–240 (2021). [CrossRef]
31. Y. Li, L. He, J. R. Yang, et al., “Study on thermal mismatch stress of HdCdTe infrared focal plane array,” J. Infrared Millim. Waves 27(6), 409–412 (2008).
32. K. N. Tu, “Recent advances on electromigration in very-large-scale-integration of interconnects,” J. Appl. Phys. 94(9), 5451–5473 (2003). [CrossRef]
33. W. Zhang, X. Chen, and Z. Ye, “A study on the surface correction of large format infrared detectors,” Semicond. Sci. Technol. 35(12), 125007 (2020). [CrossRef]
34. M. Chen, Y. Zhang, Y. Guo, et al., “Research on flip-chip bonding process and thermal cycle reliability simulation of 3-D stacked structure,” IEEE Trans. Compon., Packag. Manufact. Technol. 12(1), 51–58 (2022). [CrossRef]
35. A. Davila-Frias, O. P. Yadav, and V. Marinov, “A review of methods for the reliability testing of flexible hybrid electronics,” IEEE Trans. Compon., Packag. Manufact. Technol. 10(11), 1902–1912 (2020). [CrossRef]
36. Z. Zhang and C. P. Wong, “Recent advances in flip-chip underfill: materials, process, and reliability,” IEEE Trans. Adv. Packag. 27(3), 515–524 (2004). [CrossRef]
37. W. Wang, Y. Fan, Y. Fu, et al., “Research on bad pixel variation of IRFPA by high temperature storage and temperature shock,” Proc. SPIE 8193, 819328 (2011). [CrossRef]
