Optimizing Ductility And Fracture Of Amorphous Metal Thin .
Int J Fract (2017) 204:129–142DOI 10.1007/s10704-016-0166-yORIGINAL PAPEROptimizing ductility and fracture of amorphous metal thinfilms on polyimide using multilayersHai T. Tran · Hesham Mraied · Sina Izadi ·Alex A. Volinsky · Wenjun CaiReceived: 20 February 2016 / Accepted: 1 November 2016 / Published online: 11 November 2016 Springer Science Business Media Dordrecht 2016Abstract Aluminum–manganese (Al–Mn) thin filmswith manganese concentration up to 20.5 at.% weredeposited on polyimide (PI) substrates. A variety ofphases, including supersaturated fcc (5.2 at.% Mn),duplex fcc and amorphous (11.5 at.% Mn), and completely amorphous phase (20.5 at.% Mn) were obtainedby adjusting alloying concentration in the film. Tensiledeformation and subsequent fracture of strained Al–Mn films on PI were investigated experimentally andby finite element simulations. Compared with crystalline and dual phase counterparts, amorphous thinfilm exhibits the highest fracture stress and fracturetoughness, but limited elongation. Based on the fracturemechanism model, a multilayer scheme was adoptedto optimize the ductility and the fracture propertiesof the amorphous film/PI system. It was found thatby sandwiching the amorphous film (20.5 at.% Mn)between two ductile Cu layers, the elongation can beimproved by more than ten times, and the interfacialfracture toughness by more than twenty times. Thisdesign provides important guidelines to obtain optiH. T. Tran · H. Mraied · S. Izadi · A. A. Volinsky ·W. Cai (B)Department of Mechanical Engineering, University ofSouth Florida, 4202 E. Fowler Ave ENB118,Tampa, FL 33620, USAe-mail: caiw@usf.eduH. T. TranDepartment of Mechanical Engineering, University of Transportand Communications, Cau Giay, Hanoi, Vietnammized mechanical properties of future flexible electronics devices.Keywords Al–Mn alloy · Thin film · Polymersubstrate · Multilayer · Fracture toughness ·Finite element simulation1 IntroductionFlexible electronics has gained great interest latelydue to potential applications as wearable electronicdevices (Wagner et al. 2004), sensor skins (Lumelskyet al. 2001), electronic textiles (Bonderover and Wagner 2004), and flexible solar cells (Brabec 2004), etc.Ductility mismatch between the metallic interconnectthin films and the flexible polymer substrate has oftenled to limited stretchability and thus hindering reliable performance of the whole system. Various strategies have been adopted to improve the stretchabilityof the metal film coated on polymer, mostly focusing on crystalline thin films, such as Cu and Cr, etc.Metallic film rupture strain has been found to be highlysensitive to film adhesion to the substrate. Improvedfilm/substrate adhesion delays interfacial decohesionand retards strain localization, such as necking or shearband formation in the film (George et al. 2005; Xianget al. 2005). Annealing at relatively low temperaturesalso improves the strechability and fracture toughnessof crystalline thin films by grain growth and otherrelated phenomena (Lu et al. 2009). Compared with123
130crystalline metallic films, various amorphous alloysstand out to be good candidates as functional materialsdue to their good metallic bonding ability (Chu et al.2009; Inoue 2001; Wang 2009), excellent mechanical,corrosion (Chu et al. 2010; Moffat et al. 1993), andmagnetic properties (McHenry et al. 1999; Phan andPeng 2008). However, amorphous alloys are intrinsically brittle. When stretched, the glassy film cannot sustain the required co-deformation with the substrate. Such considerations have motivated the authorsto improve the mechanical properties (focusing on fracture toughness, interfacial fracture toughness and elongation) of amorphous metal films coated on polymersubstrates.Most amorphous alloys are quasi-brittle due to thelack of strain hardening mechanisms or any intrinsiccrack propagation barriers, such as grain boundariesor secondary phase boundaries. Therefore, their limited ductility and low fracture toughness often lead tohigh sensitivity to structural variables, such as surfaceroughness, making them unreliable for the widespreaduse (Schuh et al. 2007). One solution for this issueis adding crystalline metal layer(s) to create a hierarchical multilayered structure (Misra et al. 2005; Wangand Anderson 2005; Zhang et al. 2014). The ductilecrystalline layer will mitigate the catastrophic shearbands propagation of the amorphous layers and localizecrack propagation. Despite current progress, there havebeen very limited attempts to characterize the fracturebehavior of amorphous alloys, especially under tensileloading conditions. In this work, Cu was chosen as thecrystalline buffer layer to the amorphous film due toits relatively high elongation and tensile strength. Itwas proposed that if the strength of the well-adheredcrystalline layer is similar or higher than that of theamorphous counterpart, the amorphous layer will beconstrained by the crystalline one, making the wholemultilayered film fail at a larger strain (Chen et al.2011b; Li et al. 2007; Nieh et al. 1999). In this work,crystallinity of the Al thin films was tuned by varying Mn concentration in the alloy (Ruan and Schuh2009). Increasing Mn% from 5.2 to 20.5 at.% leads toa phase transformation from supersaturated fcc phasewith moderate ductility to brittle amorphous phase.Thin film microstructure and composition were characterized using scanning electron microscopy (SEM),energy dispersive spectra (EDS), transmission electron microscopy (TEM) and selected area diffraction(SAD). Tensile tests of metal coated polyimide (PI)123H. T. Tran et al.were performed and the strains were measured usingdigital image correlation (DIC) method. It was foundthat the amorphous Al-20.5 at.% Mn exhibits the highest fracture stress and fracture toughness, but lowestductility. Further improvement of fracture toughness,interfacial fracture toughness and elongation of theamorphous alloy/PI system was achieved by adoptingbilayered and trilayered structure using ultrathin Cubuffer layers. Finally, the failure mechanisms of the layered films were modeled using finite element analysis(FEA).2 Experimental procedureAl–Mn thin films were magnetron sputtered on 7.6 μmthick PI foils (Kapton HN by DuPont). Prior tofilm deposition, the PI substrates were ultrasonicallycleaned with acetone and ethanol. All depositions wereperformed using the CRC-100 sputtering system with70 W RF power at a base pressure of 1 10 6 Torr. Thenominal target-substrate distance was 60 mm and thedeposition rate was about 0.11 nm/s. Six sets of sampleswere prepared, as listed in Table 1. The total thicknessof metallic films on all samples was kept at about 1.2μm to minimize the film thickness effect on fracturestrain (Cordill et al. 2010). Samples M5 (Al-5.2 at.%Mn), M11 (Al-11.5 at.% Mn) and M20 (Al-20.5 at.%Mn) are 1.2 μm thick monolithic films on PI substrate.Samples B1 and B2 are bilayered films with 50 and100 nm Cu buffer layer between the PI substrate andthe Al–Mn film, respectively, with the total film thickness (i.e. thickness of both the Cu and Al–Mn layer)of 1.2 μm. Sample S is a trilayered structure on the PIsubstrate, with two Cu layers (100 nm) sandwiching theAl-20.5 at.% Mn layer (1 μm). After film deposition,coated PI exhibits negligible curvature change, thus theresidual stress in the film is neglected in this study. Thisis consistent with extensive previous reports showingthat the highly compressive residual stress in thin filmsincreases (or the absolute value decreases) rapidly withincreasing film thickness and becomes close to zeroat layer thicknesses greater than 500 nm (Frank et al.2011).Surface morphology and chemical composition ofas-deposited samples were characterized using scanning electron microscopy (Hitachi SU-70) and energydispersive spectrometer (EDAX-Phoenix). TEM samples were prepared by directly sputtering Al–Mn alloys
Optimizing ductility and fracture of amorphous metal thin films131Table 1 Composition and mechanical properties of monolithic (M5, M11, and M20), bilayer (B1 and B2), and trilayer (S) Al–Mn thinfilms deposited on PI substratesεc (%)σ f (MPa)ν–0.63 0.06199.1 8.6–0.57 0.08221.1 13.4–0.46 0.01Cu Al 20.5 at.% Mn50B2Cu Al 20.5 at.% MnSCu Al 20.5 at.% Mn CuSample IDCompositionM5Al–5.2 at.% MnM11Al–11.5 at.% MnM20Al–20.5 at.% MnB1E (GPa)K I c (MPa m 1/2 )Crit. J-int (J/m2 )0.3439.4 7.30.58–0.3362.3 5.90.78–321.7 22.70.32103.6 2.91.380.346.24 1.14220.8 11.90.3267.8 2.60.798.551002.32 0.24324.3 15.90.3290.7 11.61.212.191005.67 0.69284.1 5.60.3294.9 10.81.138.13h Cu (nm)h Cu is the thickness of the Cu layer(s) in the samples. The critical strain (εc ), fracture stress (σ f ) and elastic modulus (E) were measuredfrom the stress-strain curves of uniaxial tensile tests. The Poisson’s ratio (ν) was estimated using the rule of mixtures from pure Al andMn (Cardarelli 2008). Critical J-integrals of the film/substrate interface (interfacial fracture toughness) of sample M20, B1, B2, and Swere calculated by FEA at their corresponding critical strainson continuous carbon film grids for 15 min to reach asample thickness of 150 nm. Bright-field (BF), darkfield (DF) imaging, and SAD analysis were performedusing a Tecani F20 TEM operated at 200 kV with afield emission gun.Uniaxial tensile tests (DTS, National Instruments)were carried out at a constant strain rate of 4 10 4s 1 at room temperature. Rectangular samples with 4 20 mm2 gauge area were used. All results reportedwere obtained by averaging from at least four separatetests. Electrical resistance of the samples was recordedusing Tektronix 4050 multimeter during the test. Thecritical strain εc , i.e. the macroscopic strain whichcharacterizes the microcrack formation (as opposed torupture), was obtained by using electrical resistancechange method (Niu et al. 2007). Figure 1 shows atypical evolution of electrical resistance change duringa tensile test, where εc is defined at the point wherethe electrical resistance deviated from the ideal curve(Lu et al. 2007; Niu et al. 2007). The force of thefilm (Ffilm ) at a certain displacement was estimated asFfilm Ftotal Fsubstrate (Macionczyk and Brückner1999; Pei et al. 2011) (neglecting the force requiredto break the native oxide layer on the metallic film),where Ftotal and Fsubstrate are the tensile loads of thethin film-coated and uncoated PI specimen at the samedisplacement, respectively. The tensile stress of the filmwas then calculated as σ Ffilm /wt, where w andt are the width and the thickness of the film, respectively. Strain was measured during tensile tests usingdigital image correlation method by tracing the makers (sprayed ink speckles) displacements using a highdefinition camera (1920 1080 pixels, 30 fps) (BingPan et al. 2009). The strain was then calculated fromFig. 1 Evolution of electrical resistance change (defined as (R–R0 )/R0 , where R0 is the initial electrical resistance of the film) ofa monolithic Al-5.2 at.%Mn (sample M5) as a function of strainthe recorded images using a Matlab routine developedby Eberl et al. (2010).3 Microstructure of as-deposited Al–MnMicrostructure of as-deposited monolithic Al–Mn wasstudied by TEM and SAD, as shown in Fig. 2. Increasing Mn% in the alloy leads to a phase transition froma supersaturated fcc structure to a completely amorphous phase, similar to electrodeposited Al–Mn (Ruanand Schuh 2009). At 5.2 at.% Mn, Fig. 2a shows thatsample M5 contains a single fcc phase (lattice constanta 4.036 Å) with an average grain size d of 15nm. At the intermediate Mn concentration of 11.5 at.%,M11 contains a complex dual phase structure, where123
132H. T. Tran et al.Fig. 2 a–c TEM images and d, e corresponding SAD patterns of as-deposited monolithic Al–Mn films with various Mn % as definedin Table 1nanocrystalline fcc (a 4.035 Å and d 12 nm)and amorphous phases coexist. Further increasing Mnconcentration to 20.5 at.% leads to the formation of acompletely amorphous microstructure of M20, as confirmed by the diffuse halo in the SAD pattern (Fig. 2f).4 Tensile behaviors of monolithic and multilayeredsamplesTypical true stress-strain curves of the monolithic andmultilayer samples are shown in Fig. 3. The arrowsindicate the critical strains (εc ). Table 1 lists themechanical properties obtained from the stress-straincurves, including elastic modulus (E), fracture stress(σ f , i.e. the film stress at its critical strain), criticalstrain (εc ), and fracture toughness (K I c ). The mode Istress intensity factor K I c (fracture toughness) was calculated from the energy release rate G as (Freund andSuresh 2003): EGKIc ,(1)1 ν2123where E and ν are the elastic modulus and the Poisson’sratio of the film, respectively. The steady-state energyrelease rate G was calculated as (Beuth Jr 1992): π σ 2f h T 1 ν 2 g (α, β) ,(2)2Ewhere h T is the total thickness of the film ( 1.2 μm),g (α, β) is a dimensionless quantity determined by theelastic mismatch between the film and the substrate,and α and β are the two Dundurs’ parameters definedas:G α μ1 (1 2ν2 ) μ2 (1 2ν1 )Ē 1 Ē 2, and β ,2μ1 (1 ν2 ) 2μ2 (1 ν1 )Ē 1 Ē 2(3) 2where Ē E/ 1 ν is the plane strain tensile modulus and μ is the shear modulus. The subscripts 1 and 2refer to the film and the PI, respectively. In this work, αranges from 0.87 to 0.95, β α/4 and g (α, β) are linearinterpolation values obtained from reference (Beuth Jr1992). It should be noted the energy release rate calculated from Eq. (2) considers a single channeling crack
Optimizing ductility and fracture of amorphous metal thin filmsin a thin film on a semi-infinite substrate, while theeffect of crack spacing on G is neglected. For deformedductile thin films such as Cu on compliant substrate,parallel channel cracks are often observed perpendicular to the loading direction. It was shown that theenergy release rate of such thin films increases withcrack spacing for a given film (and substrate) thicknessand eventually reaches a maximum (saturated) valuethat corresponds to the case with a single isolated crack(Huang et al. 2003). However, as will be shown laterin Sect. 4.2, the deformed amorphous Al–Mn (M20)thin film studied here do not exhibit the typical parallel channel cracks, but rather extensive shear bandsinclined or perpendicular to the loading direction. Forsamples with Cu buffer layers such as B1, B2 and S,parallel cracks were indeed observed (Fig. 5b–d), withcrack spacing between 50 and 100 μm and crackspacing to film thickness ratio around 42–83. In suchcases, the energy release rate is approaching the saturated value as calculated here. Hence, considering thebehavior of both monolithic (M20) and layered samples(B1, B2, and S), we neglect the effect of crack spacingand adopt Eq. (2) to calculate G for simplicity, whileit should be noted that such calculations correspond toan upper limit of the steady state energy release rate forsome samples.Representative true stress-strain curves of all samples are shown in Fig. 3. For all monolithic samples, thestress reaches a maximum at the critical strain, followedby a stress relaxation due to crack and/or shear band formation. In addition, all mechanical properties, including σ f , E, and K I c were found to increase with Mn %,with M20 exhibiting the highest values of all, as shownin Fig. 4 and Table 1. However, at the highest Mn concentration (20.5 at.%), the amorphous nature of M20renders a very low tensile ductility ( 0.46%). This isnot surprising given that amorphous alloys lack sufficient intrinsic mechanisms to hinder crack propagationor strain hardening (Schuh et al. 2007). Figure 5a showsa typical SEM image of the surface of M20 at the criticalstrain. Extensive shear bands, either inclined at an angleof 45 or perpendicular to the loading direction canbe seen (indicated by white arrows), which contributeto the failure of an amorphous material (Schuh et al.2007).The critical strains remain very low (less than0.65%) for all monolithic samples regardless of theircrystallinity, similar to the behavior of brittle thin filmssuch as Cr ( 1%) (Cordill et al. 2010) and Ta (0.6%)133Fig. 3 a Representative true stress-strain curves of M5, M11and M20 and b B1, B2 and S samples. The arrows indicate thecritical strains εc(Frank et al. 2009) on PI. To improve the film/substrateadhesion and stretchability of the system, bilayered(B1 and B2) and trilayered (S) samples were designedvia extrinsic toughening mechanisms (Hofmann et al.2008; Ritchie 2011). Figure 3b shows the true stressstrain curves of all layered samples (B1, B2, and S). Itcan be seen that while εc occurred in the elastic domainin all monolithic samples (Fig. 3a), it always occurredafter extensive plastic deformation in the multilayeredsamples. Figure 4a shows that the bilayer scheme (B1)can enhance the elongation of M20 by more than tentimes (from 0.46 to 6.24%). However, at the same time,the fracture toughness decreased from 1.38 to 0.79MPa m1/2 . In contrast to B1, B2 samples maintainedhigh fracture toughness of M20 and moderate improvement of critical strain (from 0.46 to 2.32%). Finally,the trilayer sample S turned out to be the optimizedsolution, which exhibits a combination of high ductility (5.67%), fracture stress ( 284 MPa) and fracture123
134Fig. 4 a Critical strain, b fracture stress, and c fracture toughness of all monolithic, bilayer, and trilayer samples. Error barsrepresent standard deviation obtained from at least four separateteststoughness (1.13 MPa m1/2 ). The following discussionsjustify these experimental observations.4.1 Fracture mechanisms at the brittle/ductileinterface in the multilayered samplesIn a ductile/brittle layered structure, crack often initiates in the brittle layer and then travels to the duc-123H. T. Tran et al.tile/brittle interface (Wu et al. 2014). As long as thethickness (h) of the ductile layer (e.g. Cu in this work)is much larger than its Burgers vector, emitted dislocations will move away from the crack tip under tensile loading (Hsia et al. 1994). In the ductile layer,emitted dislocations will blunt the crack tip and therefore reduce the tensile stress at the crack tip. Hence,the crack propagation process is suppressed, since thecrack tip stress is unable to reach the cohesive tensile strength of the interface (Hsia et al. 1994). Moreover, if the strength of the ductile material is increased,its fracture toughness will be increased because of theplastic deformation during crack propagation (Was andFoecke 1996). Therefore, adding a ductile Cu layer hasthe potential to improve the ductility as well as thefracture toughness of the amorphous Al–Mn/PI structure. However, the dislocations emitted in the ductilelayer are also confined by the brittle layer. These dislocations pile up at the interface, generating additionalstress at the crack tip, which hinders further dislocationemission and blunting process at the crack tip (Anderson and Li 1993). Gradually, the tensile stress at theblunted crack tip reaches a critical strength resulting infracture. Therefore, the fracture toughness of the filmdepends on the number of dislocations emitted, whichin turn depends on the thickness of the ductile layer(Hsia et al. 1994).What is the appropriate ductile layer thickness thatshould be added to the amorphous M20 sample to optimize its ductility and fracture toughness? We approachthis problem by evaluating the constraining effect ofthe ductile layer on the fracture behavior of a composite material consisting of alternating ductile and brittlelayer(s). In a ductile metallic layer with a crack, theplastic zone size at the crack tip can be estimated as(Hsia et al. 1994): 1 K Ic 2,(4)rp 2π σYwhere K Ic is the fracture toughness and σY is the yieldstrength of the material. When the layer thickness (h)of the ductile phase is large, the plastic zone size is onthe order of millimeters or even c
est fracture stress and fracture toughness, but lowest ductility. Further improvement of fracture toughness, interfacial fracture toughness and elongation of the amorphous alloy/PI system was achieved by adopting bilayered and trilayered structure using ultrathin Cu bufferlayers.Finally,thefailuremechanismsofthelay-
A.2 ASTM fracture toughness values 76 A.3 HDPE fracture toughness results by razor cut depth 77 A.4 PC fracture toughness results by razor cut depth 78 A.5 Fracture toughness values, with 4-point bend fixture and toughness tool. . 79 A.6 Fracture toughness values by fracture surface, .020" RC 80 A.7 Fracture toughness values by fracture surface .
Fracture Liaison/ investigation, treatment and follow-up- prevents further fracture Glasgow FLS 2000-2010 Patients with fragility fracture assessed 50,000 Hip fracture rates -7.3% England hip fracture rates 17% Effective Secondary Prevention of Fragility Fractures: Clinical Standards for Fracture Liaison Services: National Osteoporosis .
Fracture is defined as the separation of a material into pieces due to an applied stress. Based on the ability of materials to undergo plastic deformation before the fracture, two types of fracture can be observed: ductile and brittle fracture.1,2 In ductile fracture, materials have extensive plastic
the Brittle Fracture Problem Fracture is the separation of a solid body into two or more pieces under the action of stress. Fracture can be classified into two broad categories: ductile fracture and brittle fracture. As shown in the Fig. 2 comparison, ductile fractures are characterized by extensive plastic deformation prior to and during crack
6.4 Fracture of zinc 166 6.5 River lines on calcite 171 6.6 Interpretation of interference patterns on fracture surfaces 175 6.6.1 Interference at blisters and wedges 176 6.6.2 Interference at fracture surfaces of polymers that have crazed 178 6.6.3 Transient fracture surface features 180 6.7 Block fracture of gallium arsenide 180
Fracture control/Fracture Propagation in Pipelines . Fracture control is an integral part of the design of a pipeline, and is required to minimise both the likelihood of failures occurring (fracture initiation control) and to prevent or arrest long running brittle or ductile fractures (fracture propagation control).
on the fracture increases, the contact area between the two fracture surfaces also increases, increasing the sti -ness of the fracture. Fracture specific sti ness depends on the elastic properties of the rock and depends criti-cally on the amount and distribution of contact area in a fracture that arises from two rough surfaces in contact
he American Revolution simulation is designed to teach students about this important period of history by inviting them to relive that event . Over the course of five days, they will recreate some of the experiences of the people who were beginning a new nation . By taking the perspective of a historical character living through the event, students will begin to see that history is so much .
