Abstract
The distinguishing feature of a flexible electronic device is that it maintains its function even when the shape changes repeatedly. As the degree of integration of flexible devices increases, revealing failure mechanisms and extending the lifetime of the flexible devices are getting more difficult. One of the potential damage zones is the interface of heterogeneous material components, where strain can be localized due to the mismatch of mechanical properties. In this study, we investigate the mechanically reliable interconnect design of the flexible printed circuit board (FPCB) system in which the packaging chip is integrated. When the FPCB was bent, folding occurred at the edge of the packaging chip due to the high bending rigidity compared with the plastic substrate and resulted in high strain concentration. By introducing interconnect architecture that bypassed the strain concentration area around the packaging chip, mechanical damage of the interconnects was successfully reduced. Through finite element simulation, the strain applied to the interconnect crossing the strain-concentrated region was predicted to be 2 times larger than that bypassing the strain-concentrated region, from 8.32 to 4.64%. In addition, the strain gap of these two interconnects could be increased as the Young’s modulus mismatch between the packaging chip and the substrate increased. This study is expected to improve the design guidelines to mechanically reliable interconnects in highly integrated flexible electronics.
Graphical Abstract
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
1 Introduction
With the advent of the internet of things, the demand for electronic devices that are compact and portable, flexible, and can be wearable for the human body is increasing in various fields, including flexible and stretchable interconnects [1], displays [2], batteries[3, 4], sensors [5], optics [6], and robotics [7]. Among the flexible device modules, interconnects are one of the most important component that determine the lifetime of the entire flexible device. Since the interconnects are deposited over the entire region of the device compared with others, they can be exposed to mechanical deformation and failures more easily during operation. To make the interconnects mechanically reliable, researchers have developed conductive materials that are intrinsically soft, such as conductive polymers [8, 9], gels [10], and liquid metals [11]. Strechable interconnects can also be realized by introducing micro/macrostructures that can accommodate large deformations without degradation, e.g., nanowires and nanofibers that can accommodate the strain from rotation and alignment of wires/fibers [12, 13], interconnects that are curved to to accommodate stretch without strain generation [3, 14, 15], and introducing voids and holes that can blunt crack propagations and further catastrophic failures [16].
Not only the innate mechanical properties of individual materials but also the mismatch of mechanical properties among individual materials that are physically contacted should be considered to guarantee the mechanical reliability of integrated devices. Previous literature reported that the high level of mechanical strain is localized at the interface between heterogeneous material components during the stretching or bending of a device, which can damage and degrade the whole device [15, 17]. In other words, even if a globally uniform deformation is applied to a device, the strain at local interfacial sites may not be uniform and the device can be broken sooner than expected.
In this study, we proposed a guideline of architectural design of interconnects in flexible devices that minimizes the mechanical strain through the investigation of failure mechanisms of flexible printed circuit boards (FPCBs). Experimental and numerical analysis provided the lifetime and the strain distribution of the deformed FPCBs according to the alignment and geometry of the metal interconnects under the presence of a rigid chip die. Since most flexible electronics are easily exposed to repetitive deformation, a bending fatigue reliability test was conducted on the FPCBs. By designing the architecture of the interconnects to bypass the strain-concentrated region, this study revealed a way of minimizing the level and controlling the localized area of strain on the interconnects and improved the mechanical reliability without additional follow-up treatments.
2 Experimental Design
Copper (Cu) interconnects are designed as three types: first is a parallel-type design, which has a straight line connecting both ends with the shortest distance; the other two types have a vertical line in the middle of parallel lines, which connect the central die perpendicular to the bending direction with 10-mm and 15-mm lengths, respectively (Fig. 1a). The vertical line has proven to be a desirable alignment since it is aligned parallel to the bending axis so that it can avoid electrical degradation, even when mechanical cracks are formed (See the supplementary information and Fig. S1). We thermally deposited 1-µm-thick Cu interconnects on a 50 μm-thick polyimide (PI) substrate at 1 × 10− 6 Torr. The deposition rate was 0.8 nm/s. A silicon (Si) chip die with a 50-µm-thick bonding layer was attached on the deposited Cu film and was followed by 25-µm-thick cover layer. The scale of the bonded Si chip die was 5 mm × 6 mm.
Figure 1b shows our U-bending fatigue test system, which can perform repeated folding and unfolding by pushing and pulling the rigid plates on both sides [18]. The bending radius was set to be 3 mm, with the Cu interconnects and Si chip die located at the inner side of curvature, where the compressive strain is approximately 0.83%. Bending frequency was set to be 1 Hz, and the number of bending cycles were 104. The edges of the samples were clamped with metal grips to form electrical contact. The electrical resistance of Cu interconnects was measured in situ (Keithley 2700 Multimeter) during the repeated deformation. After the fatigue test, the surface of the Cu interconnects was observed by field emission scanning electron microscope (SEM) and optical microscope.
To see the relationship between the orientation angle of Cu interconnects and bending direction, 1-mm-long Cu interconnects with 100-µm width were patterned with orientation angles varying: 0° (vertically aligned to the bending axis), 30°, 45°, 60°, and 90° (aligned parallel to the bending axis). The angle was defined as how much tilting was made to the Cu interconnects from the bending direction. The bending radius was 3 mm, and the Cu interconnects were placed facing inwards of the curvature so that compressive strain was applied.
Finite element (FE) simulation was conducted with ABAQUS (Dassault Systems, France) to figure out how the strain distributed when the specimens were deformed. Two different calculations were performed. First, we investigated the dependence of Young’s modulus mismatch between chip packaging and substrate on strain level. Applying different modulus values to a chip, four types of packaging materials were assumed: anisotropic conductive films (ACF), PI, epoxy molding compound (EMC), and Si. The Young’s modulus of ACF, PI, EMC, and Si were set to be 0.9, 2.8, 10, and 130 GPa, and the Poisson’s ratio was set to be 0.5 0.34, 0.2, and 0.28, respectively [19,20,21,22]. Line profiles of maximum principal strain were obtained along the line crossing the substrate from center to edge of the substrate around the chip edge. Second, we conducted the simulation including Cu interconnects with the same geometries prepared in the experiments. Young’s modulus and Poisson’s ratio of Cu were set to be 130 GPa and 0.36, respectively, using a perfect plastic model whose yield stress was set to be 271 MPa [23]. C3D8 meshes were applied to each element. The bonding layer and cover layer were excluded to reduce the cost of computation. U-bending deformation was simulated by narrowing the interval of both sides of the PI edges that were confined to the rigid plate by displacement control to have a 4.5-mm bending radius with a 1.5-mm margin.
3 Results and Discussion
First, U-bending fatigue tests were conducted to investigate the dependence of the presence of a Si die on fatigue of the Cu interconnects, as well as tests varying the interconnect design. Without the Si die, the specimen was uniformly bent (Fig. 1b). On the other hand, when the specimen included the Si die, local folding of Cu/PI occurred around the edge of the Si die. Young’s modulus and bending rigidity are 46 times larger for Si compared with PI substrate, making the substrate flat where the Si chip is placed. We hypothesized that the level of strain that would be generated at the locally folded sites would be higher than the strain derived by a uniform bending curvature, which would make the FPCB more vulnerable to the fatigue.
Electrical resistance of Cu interconnects was monitored in situ during repeated bending to compare the fatigue lifetime. The compressive strain applied to the Cu interconnects was calculated to be 0.83%, following thin film bending theory, which can be approximated by t/2R, where t and R are the thickness and bending radius, respectively [24]. When a Si die was not included, the change in electrical resistance of Cu the interconnects were negligible (Fig. 2a), while a gradual increase in resistance increase was observed in Cu interconnects with a Si die (Fig. 2b) during 104 cycles of repeated bending. The increase in electrical resistance of Cu interconnects was related to the formation of mechanical cracks that blocked the conduction path [25], indicating that mechanical damage occurred in Cu interconnects with a Si chip. In addition, from the resistance curve in Fig. 2b, we observed that when the length of the vertical interconnect line increased, the increment of resistance decreased. For example, the final resistance of the parallel-type interconnect design increased to 105% compared with the initial design, while those of the parallel type with a short vertical line and the parallel type with long vertical line increased 45% and 25%, respectively.
The SEM images in Fig. 3 shows the crack damage morphologies of Cu interconnects with a Si die. Extrusions and intrusions are found around the cracks, indicating that the mechanism of crack formation was derived by the slip of piled-up dislocations [24,25,26,27]. In all three cases, cracks were formed at the parallel lines of interconnects where folding occurred. From the SEM images, we observed that the crack densities decreased as the length of the vertical line increased, explaining the difference in the change of electrical resistance in Fig. 2b.
Figure 4a shows the interconnect pattern with different alignment angles to the bending axis. The line pattern was tilted from fully vertical to parallel to the bending axis to confirm that path pattern affected reliability of whole circuit. Figure 4b shows the electrical resistance changes of the Cu pattern after 104 cycles under a 1.67% bending strain calculated by the approximated bending theory of thin film on substrate [27]. As the tilting angle increased, the change in electrical resistance decreased. To understand the correlation between the alignment of the interconnect and the resistance, we established an analytic model that shows the relationship between resistance and crack length [28]:
where R is the electrical resistance of the interconnect after the fatigue test, R0 is the initial electrical resistance of the film, C is a constant, lc is the crack length across the interconnect, and b is the exponent of crack length. If we define the angle between the bending direction and the interconnect as θ, then the Eq. (1) can be expressed as follows:
where R(θ) is the electrical resistance of the film after fatigue test depending on the pattern angle. The exponent b has a value of 2.268 in the case of a conductive film [28], and in our case of a Cu line with 100-µm width was plotted by applying the value of 3.25 to the exponent in Fig. 4b. The resistance of vertically aligned interconnects increased 4000%, while the increase in resistance of parallel-aligned interconnects was negligible. Figure 4b shows that our established analytic model was well matched with the experimental results.
Figure 4c shows SEM images of fatigue cracks of the Cu interconnects that were aligned at 0° and 90° after 104 cycles of sliding bending. The crack density in both cases was similar. In the case of the 0° tilting angle, the cracks grew perpendicular to the conduction pathway so that they blocked the electrical conduction and caused severe increase in resistance [29]. In contrast, the cracks grew parallel to the direction of the conduction pathway so that they did not interfere with electrical conduction at the 90° tilting angle. This difference indicated that for a uniaxial bending, the interconnect should be aligned parallel to the bending axis to prevent electrical performance degradation. Furthermore, in the case of a pattern perpendicular to repeated strain, the degradation of electrical conduction is not affected, allowing this design rule to be applied for bypassing stress concentration under various form factors.
Figure 5 represents the FE simulation results of packaging die/Cu line/PI substrate when bent. The deformed shape and maximum principal strain contour of packaging die/Cu line/PI substrate are displayed in Fig. 4a, when the samples were bent, varying with the Young’s modulus of chip die. Four different materials, ACF, PI, EMC, and Si, were subjected to the packaging die. To clearly show the strain distribution, the strain contours are provided with the simulated model unfolded. As Young’s modulus of the die increased, the substrate was folded more sharply and the strain level at the die edge/substrate interface increased due to the increase of the bending rigidity of the die. Figure 5b and c represent the line profiles of the maximum principal strain of parallel Cu line depending on Young’s modulus of the chip die (following the white dashed line denoted in upper illustration). When a die was used, the maximum strain value at the parallel Cu line in Fig. 5b increased from 5.59 to 8.32% as the Young’s modulus of the die increased. One more important thing to note is that the strain level was highest at the center of where the die was placed and decreased toward the longer side of the edge of the substrate. The maximum strain value at the parallel Cu line with a vertical long pattern in Fig. 5c increased from 4.09 to 4.64% as the Young’s modulus of the die increased; however, the amount of increase was only a third of that compared with the parallel-only design. This reduction can be understood as substrate bending becomes more uniform the farther it moves away from the chip. (For example, as the width of the sample approaches infinity, the strain will converge to 0.97%.) For the Cu interconnect including Si die, the strain decreased from 8.32 to 4.64% as it went from center to edge. This implies that as the metal interconnects are designed to be farther from the die, the effective strain generated on the interconnects will decrease and thus the mechanical stability can be improved. The strain gap between the center and the edge of the substrate grew larger as the rigidity ratio of chip and the substrate increased, which suggested the importance of the location and geometry of metal interconnects when developing FPCBs including rigid chips.
Finally, bending simulation of FPCBs was conducted to include the Cu interconnects with three different geometries that were used in experiments. Figure 5d represents the maximum principal strain profile of Cu interconnects along the longitudinal direction of the specimen. The decreased strain level around 15 mm on the x-axis well describes the damage-free surface of the vertical lines of Cu interconnects in Fig. 3. The peak of strain was getting broader and lower as it went toward to the edge, implying that it would be desirable for metal interconnects to be designed as far away from the localized strain area as possible.
4 Conclusion
This paper reveals the mechanical reliability of metal interconnects of FPCBs with packaging components and suggests design guideline of interconnects for improving mechanical reliability under uniaxial bending. As a result of the repeated bending deformation of the Si/Cu/PI integrated module, we proved through experiments and calculations that (1) local folding occurred at the interface of the Si die edge due to the high bending rigidity of Si and (2) the Cu interconnects were mechanically damaged due to the localized strain at the die edge. The degree of folding or strain level at the edge can be determined by the mismatch of Young’s modulus between the chip die and the substrate. For example, as the mismatch decreased, the strain level decreased. As the interconnect was located farther from the rigid die edge, the strain generated at the interconnect decreased and was able to delay mechanical deterioration. This research is expected to provide the design guidelines for the highly reliable interconnect architecture of flexible electronics.
References
Wang, B., Facchetti, A.: Mechanically flexible conductors for stretchable and wearable E-skin and E‐textile devices. Adv. Mater. 31, 1901408 (2019)
Zhou, L., Wanga, A., Wu, S.-C., Sun, J., Park, S., Jackson, T.N.: All-organic active matrix flexible display. Appl. Phys. Lett. 88, 083502 (2006)
Xu, S., Zhang, Y., Cho, J., Lee, J., Huang, X., Jia, L., Fan, J.A., Su, Y., Su, J., Zhang, H., Cheng, H., Lu, B., Yu, C., Chuang, C., Kim, T., Song, T., Shigeta, K., Kang, S., Dagdeviren, C., Petrov, I., Braun, P.V., Huang, Y., Paik, U., Rogers, J.A.: Stretchable batteries with self-similar serpentine interconnects and integrated wireless recharging systems. Nat. Commun. 4, 1543 (2013)
Kim, K., Lee, Y., Costa, A., Lee, Y., Jang, T., Lee, M., Joo, Y., Oh, K.H., Song, J., Choi, I.: Extremely versatile deformability beyond materiality: a new material platform through simple cutting for rugged batteries. Adv. Eng. Mater. 21, 1900206 (2019)
Lee, Y.-J., Lim, S.-M., Yi, S.-M., Lee, J.-H., Kang, S., Choi, G.-M., Han, H.N., Sun, J.-Y., Choi, I.-S., Joo, Y.-C.: Auxetic elastomers: mechanically programmable meta-elastomers with an unusual Poisson’s ratio overcome the Gauge limit of a capacitive type strain sensor. Extreme Mech. Lett. 31, 100516 (2019)
Choi, C., Choi, M.K., Liu, S., Kim, M.S., Park, O.K., Im, C., Kim, J., Qin, X., Lee, G.J., Cho, K.W., Kim, M., Joh, E., Lee, J., Son, D., Kwon, S.-H., Jeon, N.L., Song, Y.M., Lu, N., Kim, D.-H.: Human eye-inspired soft optoelectronic device using high-density MoS2-graphene curved image sensor array. Nat. Commun. 8, 1664 (2017)
Byun, J., Oh, E., Lee, B., Kim, S., Lee, S., Hong, Y.: A single droplet-printed double‐side universal soft electronic platform for highly integrated stretchable hybrid electronics. Adv. Funct. Mater. 27, 1701912 (2017)
Lee, Y.-J., Lee, U., Yeon, Y., Shin, H.-W., Evans, H.-A.-S., Joo, L.A.: Influences of semiconductor morphology on the mechanical fatigue behavior of flexible organic electronics. Appl. Phys. Lett. 103, 241904 (2013)
Lee, Y.-Y., Lee, J.-H., Cho, J.-Y., Kim, N.-R., Nam, D.-H., Choi, I.-S., Nam, K.T., Joo, Y.-C.: Stretching-induced growth of PEDOT-rich cores: a new mechanism for strain-dependent resistivity change in PEDOT: PSS films. Adv. Funct. Mater. 23, 4020–4027 (2013)
Lee, Y.-Y., Kang, H.-Y., Gwon, S.H., Choi, G.M., Lim, S.-M., Sun, J.-Y., Joo, Y.-C.: A strain-insensitive stretchable electronic conductor: PEDOT:PSS/acrylamide organogels. Adv. Mater. 28, 1636–1643 (2016)
Yoon, S.G., Koo, H.-J., Chang, S.T.: Highly stretchable and transparent microfluidic strain sensors for monitoring human body motions. ACS Appl. Mater. Interfaces 7, 27562–27570 (2015)
Zhou, R., Guo, W., Yu, R., Pan, C.: Highly flexible, conductive and catalytic Pt networks as transparent counter electrodes for wearable dye-sensitized solar cells. J. Mater. Chem. A 3, 23028–23034 (2015)
Hwang, B., Shin, H.-A.-S., Kim, T., Joo, Y.-C., Han, S.M.: Highly reliable Ag nanowire flexible transparent electrode with mechanically welded junctions. Small 10, 3397–3404 (2014)
Kim, D.-H., Ahn, J.-H., Choi, W.M., Kim, H.-S., Kim, T.-H., Song, J., Huang, Y.Y., Liu, Z., Lu, C., Rogers, J.A.: Stretchable and foldable silicon integrated circuits. Science 320, 507–511 (2008)
Hsu, Y.-Y., Gonzalez, M., Bossuyt, F., Vanfleteren, J., De Wolf, I.: Polyimide-enhanced stretchable interconnects: design, fabrication, and characterization. IEEE Trans. Electron. Devices 58, 2680–2688 (2011)
Kim, B.-J., Cho, Y., Jung, M.-S., Shin, H.-A.-S., Moon, M.-W., Han, H.N., Nam, K.T., Joo, Y.-C., Choi, I.-S.: Fatigue-free, electrically reliable copper electrode with nanohole array. Small 8, 3300–3306 (2012)
Naserifar, N., LeDuc, P.R., Fedder, G.K.: Material gradients in stretchable substrates toward integrated electronic functionality. Adv. Mater. 28, 3584–3591 (2016)
Yi, S.-M., Choi, I.-S., Kim, B.-J., Joo, Y.-C.: Reliability issues and solutions in flexible electronics under mechanical fatigue. Electron. Mater. Lett. 14, 387–404 (2018)
Gan, L., Ben-Nissan, B., Ben-David, A.: Modelling and finite element analysis of ultra-microhardness indentation of thin films. Thin Solid Films 290–291, 362–366 (1996)
Yu, D.Y.W., Spaepen, F.: The yield strength of thin copper films on kapton. J. Appl. Phys. 95, 2991–2997 (2004)
Kim, J.-H., Lee, T.-I., Kim, T.-S., Paik, K.-W.: Effects of ACFs adhesion on the bending reliability of chip-in-flex packages for wearable electronics applications. In: 2016 IEEE 66th Electronic Components and Technology Conference (ECTC), IEEE: Las Vegas, NV, USA, pp. 2461–2467 (2016)
Ko, C.-T., Yang, H., Lau, J.H., Li, M., Lin, C., Chang, C.-L., Pan, J.-Y., Wu, H.-H., Xu, I., Chen, T., Li, Z., Tan, K.H., Lo, P., So, R., Chen, Y.H., Fan, N., Kuah, E., Lin, M., Cheung, Y.M., Ng, E., Xi, C., Beica, R., Lim, S.P., Lee, N.C., Tao, M., Lo, J., Lee, R.: Feasibility study of fan-out panel-level packaging for heterogeneous integrations. In :2019 IEEE 69th Electronic Components and Technology Conference (ECTC); IEEE: Las Vegas, NV, USA, pp. 14–20 (2019)
Lu, N., Suo, Z., Vlassak, J.: The effect of film thickness on the failure strain of polymer-supported metal films. Acta Mater. 58, 1679–1687 (2010)
Kraft, O., Schwaiger, R., Wellner, P.: Fatigue in thin films: lifetime and damage formation. Mater. Sci. Eng. A 319–321, 919–923 (2001)
Sun, X.J., Wang, C.C., Zhang, J., Liu, G., Zhang, G.J., Ding, X.D., Zhang, G.P., Sun, J.: Thickness dependent fatigue life at microcrack nucleation for metal thin films on flexible substrates. J. Phys. D Appl. Phys. 41, 195404 (2008)
Schwaiger, R., Dehm, G., Kraft, O.: Cyclic deformation of polycrystalline Cu films. Philos. Mag. 83, 693–710 (2003)
Kim, B.-J., Shin, H.-A.-S., Jung, S.-Y., Cho, Y., Kraft, O., Choi, I.-S., Joo, Y.-C.: Crack nucleation during mechanical fatigue in thin metal films on flexible substrates. Acta Mater. 61, 3473–3481 (2013)
Deng, G., Nakanishi, T.: Advances in Ceramics-Characterization, Raw Materials, Processing, Properties, Degradation and Healing. IntechOpen, London (2011)
Lacour, S.P., Wagner, S., Huang, Z., Suo, Z.: Stretchable gold conductors on elastomeric substrates. Appl. Phys. Lett. 82, 2404–2406 (2003)
Acknowledgements
This work was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (No. 2022R1A5A7000765), MOTIE(Ministry of Trade, Industry & Energy (20019469) and KSRC(Korea Semiconductor Research Consortium) (852) support program for the development of the future semiconductor device, and Technology Innovation Program (20011119) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Lee, JS., Lee, YJ., Seol, J. et al. Vertical Pattern of Interconnects to Bypass High Strain Near a Hard Die on a Flexible Substrate Under Mechanical Bending. Electron. Mater. Lett. 20, 122–130 (2024). https://doi.org/10.1007/s13391-023-00444-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13391-023-00444-1