Al2O3 Doubly Reinforced Aluminum

Sep 29, 2017 - This design strategy and model material system should guide the synthesis of bioinspired materials to achieve exceptionally high streng...
0 downloads 9 Views 5MB Size
Download PDF
Letter Cite This: Nano Lett. 2017, 17, 6907-6915

pubs.acs.org/NanoLett

Bioinspired, Graphene/Al2O3 Doubly Reinforced Aluminum Composites with High Strength and Toughness Yunya Zhang and Xiaodong Li* Department of Mechanical and Aerospace Engineering, University of Virginia, 122 Engineer’s Way, Charlottesville, Virginia 22904-4746, United States S Supporting Information *

ABSTRACT: Nacre, commonly referred to as nature’s armor, has served as a blueprint for engineering stronger and tougher bioinspired materials. Nature organizes a brick-and-mortar-like architecture in nacre, with hard bricks of aragonite sandwiched with soft biopolymer layers. However, cloning nacre’s entire reinforcing mechanisms in engineered materials remains a challenge. In this study, we employed hybrid graphene/Al2O3 platelets with surface nanointerlocks as hard bricks for primary load bearer and mechanical interlocking, along with aluminum laminates as soft mortar for load distribution and energy dissipation, to replicate nacre’s architecture and reinforcing effects in aluminum composites. Compared with aluminum, the bioinspired, graphene/Al2O3 doubly reinforced aluminum composite demonstrated an exceptional, joint improvement in hardness (210%), strength (223%), stiffness (78%), and toughness (30%), which are even superior over nacre. This design strategy and model material system should guide the synthesis of bioinspired materials to achieve exceptionally high strength and toughness. KEYWORDS: Bioinspired, aluminum composite, graphene, laminated, powder metallurgy

S

mineralization” approach. Those composites have proven that laminated structure configuration can effectively enhance the toughness of brittle materials. However, an intrinsic limitation of using synergic polymers is the low strength, which may handicap the mechanical performance of the fabricated composites. Cloning nacre’s architecture with a stronger substitute, such as aluminum, in engineered composites has proven to be a more challenging task. Due to its ultrahigh hardness and outstanding chemical stability, Al2O3 nanoparticles have long been considered as an ideal reinforcement to strengthen polymer and metallic materials.15,16 However, Al2O3 nanoparticles tend to agglomerate in the matrixone of the major roadblocks limiting their reinforcing effect and applications. Substantial efforts have been devoted to improving the dispersion of Al2O3 nanoparticles, such as surface functionalization,16 hot rolling,17 friction stirring,18 and DC plasma process.19 It has been recently recognized that orderly arranged Al2O3 nanoparticles in metallic matrix are beneficial for the joint enhancement of strength and toughness.20 Another reinforcement is graphene, as a 2-D single layer atomic crystal with extremely high strength (140 GPa) and stiffness (1 TPa),21,22 but relatively low toughness23 is expected to serve as a perfect mechanical

trength and toughness are mutually exclusive in engineered materials.1 Up to now, it is still a challenge to develop engineering materials that exhibit greatly enhanced strength without increasing brittleness or fragility. Fortunately, nature provides us with abundant and diverse inspirations which may solve this classical dilemma.2 Nacre, commonly referred to as nature’s armor, is renowned for its unusual combination of strength and toughness.3 Nature’s “wisdom” in nacre resides in its structural design strategya hierarchical structure consisting of layered hard aragonite platelets in soft organic biopolymer matrix. The aragonite component is in a layered form and serves as the primary load bearer for strength.3,4 The soft organic biopolymer layer in between the platelets plays a critical role in both load distribution and energy dissipation due to its unique deformation features.5−9 The surface nanoasperities on aragonite platelets and mineral bridges between the platelets work to interlock the platelets and prevent single platelet motion.6,7 However, this deceptively simple hard/soft/hard architecture demonstrated by nacre is actually exceptionally difficult to replicate with engineering materials and traditional manufacturing methods. Most efforts in mimicking nacre’s architecture use a ceramic/polymer configuration.10−14 For example, Ritchie et al.10 synthesized an aluminum oxide/ poly(methyl methacrylate) (Al2O3/PMMA) lamellar composite which has a similar structure to nacre. Yu et al.14 fabricated a highly nacre-like calcium carbonate/poly(acrylic acid) (CaCO3/PAA) composite via a mesoscale “assembly-and© 2017 American Chemical Society

Received: August 2, 2017 Revised: September 18, 2017 Published: September 29, 2017 6907

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters Table 1. Sample Notations and Corresponding Fabrication Processes processes notations graphene/Al2O3/Al composite pure aluminum Al semipowder metallurgy Al/graphene mix graphene/Al composite by shear mixing

surface treatment PVA surface treatment

mixing stir with GO

PVA surface treatment PVA surface treatment

freezedry

annealing

compression

sintering

cold rolling

24 h

350 °C, 1 h

50 kN

655 °C, 2 h

80% reduction

50 kN

655 °C, 2 h

80% reduction 80% reduction

50 kN 50 kN

655 °C, 2 h 655 °C, 2 h

80% reduction 80% reduction

24 h stir with GO shear mixing with graphite

24 h

350 °C, 1 h 350 °C, 1 h

Figure 1. Microstructure of the graphene/Al2O3/Al composite. (a) A nacre-like laminated structure exhibited in the graphene/Al2O3/Al composite, showing nanoscale asperities and metallic bridges (blue arrow) on interfaces. (b) EDS oxygen map of a polished graphene/Al2O3/Al composite. (c) TEM cross-sectional image of the graphene/Al2O3/Al composite sample, showing aluminum layers and nanoparticle bands. (d) Close-up inspection on the interface area, showing nanoparticles with quasi-rectangular shape. (e) The corresponding SAED pattern of panel d. (f) SEM image of a peeled graphene/Al2O3/Al composite sample, showing rod-like nanoparticles.

reinforcement for metal-based composites.24−34 Ball milling was widely used to mix graphene flakes with metal powders to enhance the dispersion uniformity.24−28 Semipowder metallurgy, which mixes graphene or graphene oxide (GO) sheets with metal powders in liquid, was found to be more effective in dispersing reinforcements and enhancing graphene/metal bonding.29−31 Several new methods such as electrostatic adsorption32 and chemical adsorption33 were recently exploited. Guo et al.34,35 utilized a magnetoresistance technique to disperse graphene flakes and carbon nanofibers in polymer matrix, which has the potential to be grafted in metal-matrix composites. After high temperature sintering, severe plastic deformation, such as high-ratio differential speed rolling,28 hot extrusion,29 and uniaxial compression,36 was often adopted to densify the composite with the goal to improve material strength and align graphene sheets. However, considering the amazing mechanical properties of graphene, the synthesized graphene/metal composites are still lower than the expectation, likely due to three roadblocks hindering the development of graphene/metal composites: weak interfacial bonding between graphene and metals, aggregation of graphene, and degradation of graphene. Here, we converted low-cost, low-purity aluminum flakes and graphene oxide sheets into a durable and noteworthy

composite. Through a series of bottom-up assembling procedures, including semipowder metallurgy, ice-templating, sintering, and densification process, typical nacre features, such as a laminated structure, nanoasperities, and mineral bridges, were effectively replicated. In particular, Al2O3 nanoparticles were introduced as a coreinforcement with graphene to further strengthen the composite. Such a multireinforcement (graphene and Al2O3) strategy enabled a joint improvement in both strength and toughness in Al composites which have significant applications in lightweight structures, such as aircrafts and electric vehicles. Since graphene and metal oxides are widely used in energy harvesting and storage systems, such as solar cells,37 thus in addition to structural materials, the graphene/ metal oxide involved, laminated configuration may find more applications. Table 1 outlines the preparation procedures of graphene/ Al2O3/Al composites and reference samples. Figure S1 (see Supporting Information, SI) schematically demonstrates the preparation procedures for graphene/Al2O3/Al composites. Specifically, the graphene oxide (GO) suspension was prepared by the Hummers method.38,39 The functional groups on the GO sheets make them hydrophilic and hence dispersible in an aqueous solvent without aggregation, forming a brown aqueous suspension (Figure S2a inset). Most as-obtained GO flakes 6908

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters were measured to be about 0.8 nm thick and highly flexible and transparent, indicative of high quality monolayered GO nanosheets (Figure S2a−d).39 A solution of 3 wt % polyvinyl acetate (PVA) was used to coat a hydrophilic layer on the surface of the Al flakes with the intention of enhancing the bonding between the aluminum flakes and GO sheets (Figure S3).29,40 The reason why we chose PVA over other polymers, such as ethyl oleate,41 polyvinylpyrrolidone,42 acrylonitrile− butadiene−styrene,43 polydimethylsiloxane,44 and poly(L-lactide),45 is because PVA has some irreplaceable features. For instance, PVA can be homogeneously dissolved into water at room temperature. Such water solution has a low viscosity. In addition, PVA can be frozen together with water in a freezer and vaporized at a temperature lower than 300 °C. After the subsequent sintering and densification processes, graphene/ Al2O3/Al composites were successfully fabricated. Excitingly, a very well arranged layered structure was successfully built throughout the composite (Figure 1a); atomic force microscopy (AFM) observations (Figure S4) further corroborated the presence of the laminated structure. Nanoasperities were observed on the aluminum flake surfaces, and metallic bridges (indicated by the blue arrow in Figure 1a) were found to connect layers together. Therefore, nacre’s three identifying features (laminated structure, nanoasperities, and mineral bridges) were effectively emulated in the composite. Yu et al.14 achieved a high resemblance to nacre in their CaCO3/PAA composite; however, the biological features that are essential for high strength and toughness have never been realized in metal/ ceramic composites. Interestingly, no other reference samples exhibited similar morphological features (Figure S5). The graphene/Al composite, which was fabricated by shear mixing graphite powders with aluminum flakes,46 processed a laminated structure which was similar to Li’s work (Figures S5d and S6).36 However, a key difference of this sample from the graphene/Al2O3/Al composite in Figure 1a was the absence of nanoasperities. Thusly, the GO addition, PVA surface treatment, and ice-templating process are essential for successfully reproducing nacre-like lamellar structure. We then proceeded to answer the question: what were the nanoasperities/nanoparticles in the graphene/Al2O3/Al composite? Energy-dispersive X-ray spectroscopy (EDS) element maps of the polished graphene/Al2O3/Al composite cross section exhibited oxygen and aluminum alternating layers (Figure 1b and Figure S7). The transmission electron microscopy (TEM) cross-sectional image revealed that nanoparticle bands embedded between two aluminum layers (Figure 1c). The close-up inspection on the interface area (Figure 1d) and the statistical results of the particles (Table S1) validated that the nanoasperities were in fact numerous nanoscale particles with quasi-rectangular shape. Selected area electron diffraction (SAED) pattern of the nanoscale asperities uncovered that these particles were actually Al2O3 (Figure 1e), which was further proven by TEM dark field image (Figure S8). The average thickness of aluminum layers and Al2O3 nanoparticle bands, calculated from scanning electron microscopy (SEM), TEM, and AFM results, were 1.05 μm and ∼200 nm, respectively. Compared with the spontaneously formed amorphous Al2O3 in ref 36, this Al2O3 nanoparticle band was much thicker and had well-defined crystalline structure. SEM images of the peeled graphene/Al2O3/Al composite surface showed rod-like Al2O3 nanoparticles (Figure 1f), and the EDS element maps suggested that the surface was covered by a carbon layer (Figure S9). Conversely, graphene/Al composites

fabricated by shear mixing method did not show any nanoasperities on the peeled surface (Figure S10). An instant question arises: where was the graphene in the composite? To address this question, atomic resolution EDS scanning and high resolution TEM (HRTEM) inspection were performed. Low magnification scanning TEM (STEM) image and the corresponding EDS element map (Figure 2a and b)

Figure 2. STEM and atomic resolution EDS element maps of the graphene/Al2O3/Al composite. (a) STEM image of graphene/Al2O3/ Al composite. (b) Corresponding EDS map of panel a, indicating that the interfaces were rich in oxygen and carbon. (c) Close-up STEM inspection of the interface with three HRTEM images, suggesting that multilayered graphene embedded in α-Al2O3 at the interface and beyond the interface was aluminum. (d) EDS map of aluminum. (e) EDS map of oxygen. (f) EDS map of carbon. (g) EDS overlapped map, showing that a layer of carbon separated the aluminum and Al2O3 nanoparticle layers.

demonstrated similar results to the SEM image and EDS maps in Figure 1b and Figure S7: only lamellar interfaces were found to be rich in oxygen. A close-up observation of the interface area and atomic resolution EDS maps of aluminum, oxygen, carbon, and a combined element map are respectively shown in Figure 2c−g. A carbon film with a thickness ranging from 5 to 13 nm was clearly embedded in the Al2O3 band, separating the aluminum lamellae and forming an Al/Al2O3/graphene sandwiched structure. The thickness of monolayer graphene was measured to be 0.38 nm; therefore this carbon film is equivalent to approximately 13−34 layers of 2D graphene. The HRTEM images (Figure 2c) and corresponding fast Fourier transform (FFT) patterns (Figure S11) reliably validated that the carbon layer was multilayered graphene. Unlike the aluminum and Al2O3 crystals, which processed a nearly perfect lattice structure, the multilayered graphene displayed a discontinuous and tortuous pattern, indicative of the 6909

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters

After annealing this piece of GO at 350 °C then at 655 °C, the peak at 10.2° disappeared, and a small but broad peak emerged at 26°, indicating the formation of rGO (Figure 3a noted as ②). The results were consistent with previous studies about diffraction peaks of pristine GO and rGO.47−49 Subsequently, this piece of rGO was grounded and rolled in one direction severely to simulate the densification process, leading to the broad peak at 26° becoming sharp (Figure 3a noted as ③). This sharp peak was very similar to the peak in the spectrum of Figure 3a, ④. Raman tests were also carried out on these samples to validate the result. The intensity ratio between the D peak (∼1350 cm−1) and the G peak (1590 cm−1) of pristine GO was 0.93 (Figure 3b noted as ①). The high temperature treatment pushed up the value of ID/IG to 1.01, which was consistent with the work of others (Figure 3b noted as ②).27,50 After severe deformation the ratio reached 1.18 (Figure 3b noted as ③), and a hint of a defective peak near the D peak emerged. The ID/IG ratio of the peeled graphene/Al2O3/Al composite sample was 1.29 with an obvious defective peak near the D peak (Figure 3b noted as ④). Since the value of ID/IG denotes the degree of disorder and defects of graphene, it is clear that the carbon exists as defective rGO or defective graphene in the final composites, which is collaborated by the discontinuous lattice structure and poor FFT diffraction in Figure 2c and Figure S11. Pristine monolayer graphene exhibits extremely high strength and stiffness.22 However, the elastic modulus of monolayer graphene decreases abruptly from 1 TPa to 223.9 GPa after oxidization.51 Although the modulus would partially recover after reduction, defects induced during the reduction would deteriorate mechanical properties further.52 Moreover, the thickness of the carbon layer shown in Figure 2d is equivalent to 13−34 layers of graphene, which decreases the mechanical performance of graphene to a large extent.53 Therefore, the real elastic modulus of the defective graphene layer in the composite should be less than 200 GPa. Although multilayered defective graphene flakes may still be able to strengthen metallic matrix composites (MMCs), the real effect is much less than expected (graphene/Al mix by shear mixing sample and refs 19−28). Various oxygen containing functional groups, such as hydroxyl, ester, and ether, are often attached on GO. The content of oxygen in GO is 36.4%.51 During heating, functional groups gradually detach from the GO matrix and release oxygen containing gases depending on the temperature. The reaction between aluminum and oxygen is very likely to happen even when the concentration of oxygen is very low.54 Because the composite was compressed before sintering, the released oxygen containing gases could not escape, but reacted with aluminum, leading to the formation of Al2O3 nanoasperities at interfaces. A small amount of Al4C3 may form as a side product of these reactions. After the sintering process, severe plastic deformation further induced defects in the graphene sheets. As a result, an Al/Al2O3/defective-graphene/Al2O3/Al multilayered morphology has been successfully fabricated at the interface regions (Figure 3c). For the graphene/Al composite fabricated by shear mixing, a laminated structure was also created because the graphene flakes produced by shear mixing also prevented aluminum layers from melting together. However, due to the absence of oxygen-containing functional groups, no Al2O3 nanoasperities formed at the interfaces (Figure S10). The nacre-like graphene/Al2O3/Al composite exhibited superiority in mechanical properties compared to all other

degradation of graphene. The HRTEM image and FFT pattern, coupled with SEAD pattern in Figure 1e, also confirmed that the Al2O3 particles were α-Al2O3, which has a hexagonal close packed (HCP) structure and thermodynamically tends to grow along the close packed ⟨0001⟩ direction, leading to the rod-like morphology displayed in Figure 1d and f. Unlike aluminum and oxygen that overlapped on a large scale (Figure 2d and e), carbon was concentrated along the laminar interface, indicating that only a handful of Al/C components formed in this composite (differing from the findings of others24,27). As a laminated composite, the interfaces between layers are of special importance because weak interfacial connection may trigger interfacial delamination. Figure S12 is the HRTEM images of graphene/Al2O3 and Al2O3/Al interfaces. A 2−3 nm thick amorphous interlayer was found to bond the graphene and the Al2O3 together (Figure S12a). EDS mapping (Figure 2d−g) unveiled that the amorphous layer contains aluminum, oxygen, and carbon. A previous study36 showed that an amorphous Al2O3 layer spontaneously forms on the aluminum surface. Moreover, since carbon was also found in the EDS map, we postulate that carbon atoms diffused into the amorphous layer during high-temperature annealing. No voids and defects were found at the graphene/Al2O3 interface, suggesting that a transition zone did link up the graphene and Al2O3, which can redistribute load, reducing interfacial delamination tendency. In contrast, a boundary layer was found between the Al2O3 and Al (Figure S12b). Therefore, intimate interfacial layers did exist at both the graphene/Al2O3 and the Al2O3/Al interfaces. GO was the first additive in the composite, but it experienced a complex journey on its way to the final product, including adsorption, reduction, sintering, and severe deformation. Understanding the final form of carbon in the composite is essential to help us to decipher the mechanisms attributing toward the superior mechanical performance, though previous studies have rarely addressed this need. The X-ray diffraction (XRD) spectrum performed on the peeled graphene/Al2O3/Al composite in Figure 1f is shown in Figure 3a and noted as ④. A

Figure 3. Evolution of the graphene and formation of the layered structure. (a) XRD spectra of ① GO, ② rGO, ③ deformed rGO, and ④ the Al/Al2O3/graphene composite. (b) Raman spectra. (c) Schematic diagram of the formation of Al/Al2O3/defective-graphene/Al2O3/Al multilayered structure.

small but sharp peak was found at 26°. However, this peak is neither similar to reduced GO (rGO) nor graphene because these two exhibit small but relatively broad peaks around 26°.47−50 To dismount this conundrum, we performed a set of XRD tests to simulate the evolution process of GO. Before reduction, GO exhibited a peak at 10.2° (Figure 3a noted as ①). 6910

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters

Figure 4. Mechanical properties of graphene/Al2O3/Al composite and reference samples. (a) Comparative tensile test curves. (b) Comparative bar chart of hardness, tensile strength, Young’s modulus, and fracture toughness. (c) Ashby plot of the specific values (normalized by density) of strength and stiffness. (d) Comparison of the strengthening and stiffening efficiencies of graphene/Al2O3 doubly reinforced composite with various reinforcements in Al matrix composites. (Data of panels c and d were drawn from refs 1, 2, 10, 23−26, 28, 29, 31, 33).

related; samples that have higher ultimate strengths may also be more ductile. Figure 4c shows the Ashby plot of the specific strength and stiffness (normalized by density) of different materials. The graphene/Al2O3/Al composite outperforms most metals and alloys and is located at the top of composites. The reinforcing efficiency in a metal matrix composite can be defined as the strength and modulus increment per unit volume fraction of the reinforcement, i.e., (σc − σm)/Vfσm and (Ec − Em)/VfEm, where σc and σm are the tensile strengths of the composite and the matrix, respectively, Ec and Em are the Young’s modulus of the composite and the matrix, respectively, and Vf is the volume fraction of the reinforcement. (In this study, 2 wt % of graphene was added. On the basis of the density of graphite and aluminum, the volume fraction of added reinforcement was about 2.6%). For aluminum based composites, our graphene/Al2O3/Al composite processes outstanding strengthening efficiency and stiffening efficiency that are superior to other reported composites known to date (Figure 4d.) To probe the microscopic mechanical response of the graphene/Al2O3/Al composite, a series of loading−unloading nanoindentation tests with a load increment of 0.1 mN each loading cycle were carried out on the composite. Intriguingly, as the indenter was pushed down into the composite, the corresponding reduced modulus exhibited an orderly wavelike pattern (Figure S13a), oscillating between a peak value of 93 GPa and a valley value of 73 GPa. Each indentation loading cycle corresponded to an approximately 800 nm indentation depth. The aluminum layer thickness was 1.05 μm. Thus, it is highly likely that the indenter tip encountered one aluminum layer in one cycle and penetrated through the Al2O3/graphene/ Al2O3 layer in the subsequent cycle, leading to an ordered oscillation of reduced modulus (Figure S13b). The result

samples tested (Figure 4). Tensile test samples were machined into a dog-bone shape according to the ASTM standard.55 Compared with the other four control samples, the tensile strength, yield strength, and Young’s modulus of the graphene/ Al2O3/Al composite sample were exceedingly superior (Figure 4a). Although the composite was not as ductile as pure aluminum, the tensile curve still exhibited a notable plastic deformation stage. Quantitatively, for the five specimens showed in Figure 4a, the graphene/Al2O3/Al composite exhibited 210% improvement in hardness, 223% increase in ultimate strength, 78% enhancement in Young’s modulus, and 30% raise in toughness (area under the tensile test curve) compared with pure aluminum (Figure 4b) (detailed numbers are listed in Table S2). A moderate increase in elastic modulus and strength of the graphene/Al mix sample (without PVA treatment and freeze-dry) and Al semipowder metallurgy sample (without GO additive) was arisen from impurities and defects, which, in turn, also made these samples less ductile. Since pores and defects were also found in the graphene/ Al2O3/Al composites, the laminated design is considered to be less defect-sensitive. The graphene/Al composite made by shear mixing also exhibited excellent improvements in strength and elastic modulus over aluminum, but it was still outperformed by the graphene/Al2O3/Al composite in all categories. The improvement of mechanical properties of the graphene/Al2O3/Al composite exceeds all other previous works.24−33 The ultimate tensile strength is also comparable to that of AA6061-T6 alloy.56 For the nine tensile samples taken from three separately prepared batches of the graphene/ Al2O3/Al composites, the ultimate strength ranged from 303 to 332 MPa with a mean value of 308 MPa, with the final strain at fracture ranging from 2.4% to 3.3% with an average value of 2.8%. Values of ultimate strength and strain are not linearly 6911

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters

Figure 5. In situ three-point bending test on the graphene/Al2O3/Al composite inside SEM. (a) SEM images of the crack under various degrees of sample deflection. (Yellow arrows indicate the progression of bending and white arrows indicate the primary crack propagation direction.) (b) The crack propagation of graphene/Al2O3/Al composite sample exhibited a confluence of multiple toughening mechanisms. The primary crack was detoured into a serpentine morphology. Along with the primary crack, several large secondary cracks were stimulated and propagated parallel to the lamellae. The tip of the primary crack was blunt with a large radius of curvature. The border of the primary crack displayed a zigzag shape with several small secondary cracks. Metal bridges formed behind the crack tip. (c) The crack propagation of a graphene/Al mix sample (without PVA surface treatment and freeze-dry casting).

The superior mechanical performance of our graphene/ Al2O3/Al composite stems from synergic contribution of multiple features. Here, the graphene/Al2O3/Al composite exhibited higher strength and toughness than the graphene/Al composite by shear mixing. Clearly, the difference in the strength and stiffness between the graphene/Al 2O 3/Al composite and graphene/Al composite by shear mixing originates from Al2O3 nanoasperities. Bulk Al2O3 possesses an elastic modulus of 400 GPa and a strength of 3 GPa.57 In the graphene/Al2O3/Al composite, the Al2O3/graphene/Al2O3 sandwiched layers play the same role as aragonite platelets in nacre, serving as the primary load bearer for greater strength and stiffness. These Al2O3/graphene/Al2O3 sandwiched layers also confined the immigration of grain boundaries and prevented recrystallization during high temperature heating. Al2O3 nanoparticles, on the other hand, are able to pin the dislocations by Orowan mechanism.58 Additionally, the large interfacial area stimulates the pile-up of dislocations, improving the dislocation density rapidly, leading to a high strain hardening rate.36 The aforementioned microscale mechanisms jointly strengthened and stiffened the composite. We also cannot ignore the toughening mechanisms stemming from the laminated design. Ritchie and his coworkers1,2 have worked extensively on the toughening mechanisms of layered biological structures and concluded that, unlike monolithic materials such as aluminum alloys, toughness in these materials often derived from extrinsic mechanisms (behind the crack tip, >1 μm) including crack deflection, bridging, crack tip blunting, and secondary cracks. Developed bioinspired materials also showed similar toughening mechanisms.10,14,59,60 To explore the fracture process of our graphene/Al2O3/Al composite, an in situ three-point bending

further validated the hard−soft−hard laminated architecture in our composite. In contrast, pure aluminum sample exhibited a relatively constant modulus value ranging from 65 to 70 GPa as a function of indentation depth (Figure S13c). The Al/ graphene mix sample, on the other hand, displayed a strong fluctuation of reduced modulus (Figure S13d). At some locations, the sample showed a high modulus of over 90 GPa, indicative of inhomogeneous dispersion of Al2O3 clusters. The graphene/Al2O3/Al composite also exhibited a superior stability at high temperature compared with pure aluminum. A piece of graphene/Al2O3/Al composite and a piece of pure aluminum with the same dimensions were heated to 750 °C for 1 h (much higher than the melting point of aluminum, 660 °C). The graphene/Al2O3/Al composite sheet remained the same as it was in terms of shape and size, whereas the pure aluminum piece was melted into an irregularly shaped bar (Figure S14a). Further experiments showed that graphene/Al2O3/Al composites could maintain their size and shape up to temperatures on the order of 900 °C. We postulate that this is due to the encapsulation of the aluminum flakes by Al2O3/graphene shells, which have much higher melting point than pure aluminum. During high temperature annealing, these shells surrounded the aluminum core allowing the material to retain its shape and size even when the aluminum core melted. After the 750 °C heat treatment the graphene/Al2O3/Al composite remained in a lamellar structure, but the nanoscale asperities became thicker, coarser, and less uniform (Figure S14b) because GO release more oxygen-containing gases at higher temperature. Mechanical testing showed that the 750 °C treated graphene/Al2O3/Al composite sheet had a lower strength of 148 MPa with a larger elongation of 4.1% (Figure S14c). 6912

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters

Here, a is the pre-existing crack or void length. Thus, obviously, the stress concentration on the crack tip of the graphene/ Al2O3/Al composite sample is much lower than that of the graphene/Al mix sample, which also explains why the crack in the graphene/Al mix sample propagated more quickly under the same degree of deformation. Two other toughening mechanisms are the nanoasperities and the metal bridges. Delamination because of weak connections between different layers may cause quick fracturing in laminated materials. Close observation of the fracture surface of the graphene/Al2O3/Al composite sample showed rare evidence of “pulling out” between layers because the roughness provided by the nanoasperities interlocked the lamellae, preventing large-scale delamination. Additionally, metal bridges behind the crack tip impeded the opening of the crack, further preventing/delaying catastrophic fracture. According to previous studies,57−59,65 the fracture toughness density of Al2O3 is about 255 kJ/m3, and that of defective graphene is 223 kJ/m3 (maximum). A simple superposition of these values with the fracture toughness of pure aluminum in Figure 4 is 6133 kJ/m3, which is 18% lower than the fracture toughness of the graphene/Al2O3/Al composite. Thusly, via the cooperation of multiple phases and the unique features of the laminated structure, the laminated graphene/Al2O3/Al composite successfully reproduced the nature’s accomplishment in nacre; that is, the toughness value exceeds its individual constituents and their simple mixture. Summarily, a rational design strategy for nacre-like laminated graphene/Al2O3/Al composites was proposed in this contribution. During high-temperature processing, GO sheets prevented the melting of aluminum flakes and facilitated the formation of Al2O3 nanoasperities. The formed Al2O3/graphene/Al2O3 layers acted as the load bearer for the composite, while aluminum cemented between the Al2O3/graphene/Al2O3 layers ensured the ductility of the composite. Moreover, the Al2O3 nanoasperities and metal Al bridges together prevented largescale delamination of the layers, further delaying catastrophic fracture. Compared with pure aluminum, the obtained composite exhibited 210% improvement in hardness, 223% increase in ultimate strength, 78% enhancement in Young’s modulus, and 30% raise in toughness. The toughness value exceeded its individual constituents and their simple mixture. This smart manufacturing strategy which utilized the oxygen containing function groups on GO and the high strength of metal oxides can be grafted to other metallic materials such as Mg, Ti, Cu, and Zn. The design principles create new opportunities for developing bioinspired materials to achieve superior mechanical performance for applications in an extensive range of fields.

test on the graphene/Al2O3/Al composite sample was carried out under SEM to observe the crack evolution (Figure 5a). The crack propagation of the graphene/Al2O3/Al composite exhibited a confluence of multiple toughening mechanisms (Figure 5b). The primary crack was detoured into a serpentine morphology instead of a straight fracture. Along with the primary crack, several large secondary cracks were stimulated and propagated parallel to the plate (perpendicular to the propagation of primary crack). Moreover, the edge of the primary crack displayed a zigzag path with small secondary cracks. All of these mechanisms inherently elongated the crack length, leading to more energy required to propagate the crack, Ws:61,62 Ws = 2abγ

(1)

where a is the crack length, b is the out-of-plane thickness of the solid material, and γ is the sum of surface energy (γs) and energy related to plastic deformation (γp). The crack length in Figure 5a, ②, was measured carefully by pixels. The span between the crack tip and plate surface was measured to be 164.11 μm, but the total crack length was summed up to 512.89 μm. For comparison, an identical experiment was performed on a graphene/Al mix sample as in Figure S5c (without PVA surface treatment and freeze-dry) with the same size and shape. Figure 5c shows the crack morphology of the graphene/Al mix sample with the same bending deflection as Figure 5a, ①. Clearly, the crack propagated straight from one side of the plate to the other side without deflections. The span from the plate surface to the tip of the crack in this sample was 160.61 μm, and the total length of the crack was 227.15 μm. Since both of the two samples comprised of aluminum and defective graphene, we postulate that the values of energy γ were almost the same. Substituting all the numbers into eq 1, the work required to propagate a crack for the laminated graphene/ Al2O3/Al composite sample is 2.21 times higher than that of the graphene/Al mix sample. This result is consistent with the fracture toughness calculated from tensile test curves (Figure 4b). Apart from the deviation and elongation of the crack, the Cook−Gordon toughening mechanism also takes effect in the composite.63 When a composite processes soft layers (aluminum) embedded within hard layers (Al2O3/graphene/ Al2O3), as the crack reaches a weak interface, the stress on the crack can easily break the interface, forming a perpendicular crack ahead of the crack tip. When these cracks merge, the crack tip radius of curvature will significantly increase, expelling stress concentration on the interfaces. Comparing the images of the crack tips for the two samples, the crack tip of the graphene/Al2O3/Al composite sample had a much larger radius of curvature (ρ) (1.47 μm) than that of graphene/Al mix sample (0.20 μm). This is because in the graphene/Al2O3/Al composite sample the aluminum layers are perpendicular to the crack propagation direction; hence the crack has to “break down” the aluminum layers. However, for the graphene/Al mix sample, the crack can easily propagate through defects, which originated from the nonhomogeneously dispersed graphene and Al2O3 particles. The radius of curvature is inversely related with the stress concentration on the crack tip:63,64 ⎛ a⎞ ⎟ σtip = σa⎜1 + 2 ρ⎠ ⎝

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b03308. Experimental section; schematic diagram of sample preparation procedures; AFM and TEM inspections of GO flakes; SEM images of a GO/Al flake; AFM phase image of graphene/Al2O3/Al composite; fracture surface morphologies of reference samples; schematic diagram of fabricating graphene/Al flakes by shear mix and TEM image of a graphene sheet produced by shear mix; SEM

(3) 6913

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters

■

image, EDS maps, and dark field TEM image of transverse section of graphene/Al2O3/Al composite; SEM images and EDS element maps of a peeled graphene/Al2O3/Al composite and a peeled graphene/ Al composite by shear mixing; FFT patterns of HRTEM images in Figure 2c; HRTEM images of graphene/Al2O3 and Al/Al2O3 interfaces; nanoindentation of graphene/ Al2O3/Al composite, pure aluminum, and Al/graphene mix samples; high-temperature performance of graphene/Al2O3/Al composite; statistical size of Al2O3 nanoparticles; mechanical property values (PDF)

(21) Novoselov, K. S.; Jiang, D.; Schedin, F.; Booth, T. J.; Khotkevich, V. V.; Morozov, S. V.; Geim, A. K. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 10451−10453. (22) Lee, C.; Wei, X.; Kysar, J. W.; Hone, J. Science 2008, 321, 385− 388. (23) Zhang, P.; Ma, L.; Fan, F.; Zeng, Z.; Peng, C.; Loya, P. E.; Liu, Z.; Gong, Y.; Zhang, J.; Zhang, X.; Ajayan, P. M.; Zhu, T.; Lou, J. Nat. Commun. 2014, 5, 3782. (24) Bartolucci, S. F.; Paras, J.; Rafiee, M. A.; Rafiee, J.; Lee, S.; Kapoor, D.; Koratkar, N. Mater. Sci. Eng., A 2011, 528, 7933−7937. (25) Bastwros, M.; Kim, G.-Y.; Zhu, C.; Zhang, K.; Wang, S.; Tang, X.; Wang, X. Composites, Part B 2014, 60, 111−118. (26) Yan, S. J.; Dai, S. L.; Zhang, X. Y.; Yang, C.; Hong, Q. H.; Chen, J. Z.; Lin, Z. M. Mater. Sci. Eng., A 2014, 612, 440−444. (27) Pérez-Bustamante, R.; Bolaños-Morales, D.; Bonilla-Martínez, J.; Estrada-Guel, I.; Martínez-Sánchez, R. J. Alloys Compd. 2014, 615, S578−S582. (28) Kim, W. J.; Lee, T. J.; Han, S. H. Carbon 2014, 69, 55−65. (29) Wang, J.; Li, Z.; Fan, G.; Pan, H.; Chen, Z.; Zhang, D. Scr. Mater. 2012, 66, 594−597. (30) Rashad, M.; Pan, F.; Tang, A.; Asif, M. Prog. Nat. Sci. 2014, 24, 101−108. (31) Rashad, M.; Pan, F.; Asif, M.; Tang, A. J. Ind. Eng. Chem. 2014, 20, 4250−4255. (32) Li, Z.; Fan, G.; Tan, Z.; Guo, Q.; Xiong, D.; Su, Y.; Li, Z.; Zhang, D. Nanotechnology 2014, 25, 325601. (33) Zhao, Z. Y.; Guan, R. G.; Guan, X. H.; Feng, Z. X.; Chen, H.; Chen, Y. Adv. Eng. Mater. 2015, 17, 663−668. (34) Zhang, X.; Alloul, O.; He, Q.; Zhu, J.; Verde, M. J.; Li, Y.; Wei, S.; Guo, Z. Polymer 2013, 54, 3594−3604. (35) Gu, H.; Guo, J.; Wei, H.; Guo, S.; Liu, J.; Huang, Y.; Khan, M. A.; Wang, X.; Young, D. P.; Wei, S.; Guo, Z. Adv. Mater. 2015, 27, 6277−6282. (36) Li, Z.; Guo, Q.; Li, Z.; Fan, G.; Xiong, D.-B.; Su, Y.; Zhang, J.; Zhang, D. Nano Lett. 2015, 15, 8077−8083. (37) Liu, T.; Yu, K.; Gao, L.; Chen, H.; Wang, N.; Hao, L.; Li, T.; He, H.; Guo, Z. J. Mater. Chem. A 2017, 5, 17848−17855. (38) Hummers, W. S., Jr.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80, 1339. (39) Stankovich, S.; Piner, R. D.; Nguyen, S. T.; Ruoff, R. S. Carbon 2006, 44, 3342−3347. (40) Peko, C.; Groth, B.; Nettleship, I. J. Am. Ceram. Soc. 2010, 93, 115−120. (41) Zheng, Y.; Zheng, Y.; Yang, S.; Guo, Z.; Zhang, T.; Song, H.; Shao, Q. Green Chem. Lett. Rev. 2017, 10, 202−209. (42) Zhang, L.; Yu, W.; Han, C.; Guo, J.; Zhang, Q.; Xie, H.; Shao, Q.; Sun, Z.; Guo, Z. J. Electrochem. Soc. 2017, 164, H651−H656. (43) Li, Y.; Wu, X.; Song, J.; Li, J.; Shao, Q.; Cao, N.; Lu, N.; Guo, Z. Polymer 2017, 124, 41−47. (44) Sun, K.; Xie, P.; Wang, Z.; Su, T.; Shao, Q.; Ryu, J.; Zhang, X.; Guo, J.; Shankar, A.; Li, J.; Fan, R.; Cao, D.; Guo, Z. Polymer 2017, 125, 50−57. (45) Zhang, K.; Li, G.-H.; Feng, L.-M.; Wang, N.; Guo, J.; Sun, K.; Yu, K.-X.; Zeng, J.-B.; Li, T.; Guo, Z.; Wang, M. J. Mater. Chem. C 2017, 5, 9359. (46) Paton, K. R.; et al. Nat. Mater. 2014, 13, 624−630. (47) Feng, H.; Cheng, R.; Zhao, X.; Duan, X.; Li, J. Nat. Commun. 2013, 4, 1539. (48) De Silva, K. S. B.; Gambhir, S.; Wang, X. L.; Xu, X.; Li, W. X.; Officer, D. L.; Wexler, D.; Wallace, G. G.; Dou, S. X. J. Mater. Chem. 2012, 22, 13941. (49) Ding, J.; Yan, W.; Xie, W.; Sun, S.; Bao, J.; Gao, C. Nanoscale 2014, 6, 2299−2306. (50) Yang, D.; Velamakanni, A.; Bozoklu, G.; Park, S.; Stoller, M.; Piner, R. D.; Stankovich, S.; Jung, I.; Field, D. A.; Ventrice, C. A., Jr.; Ruoff, R. S. Carbon 2009, 47, 145−152. (51) Haubner, K.; Murawski, J.; Olk, P.; Eng, L. M.; Ziegler, C.; Adolphi, B.; Jaehne, E. ChemPhysChem 2010, 11, 2131−2139.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: 434-243-7762. ORCID

Yunya Zhang: 0000-0002-9411-0184 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Financial support for this study was provided by the U.S. National Science Foundation (CMMI-1537021). The authors thank the staff members at the University of Virginia NMCF and NCSU AIM for electron microscopy technical support.

■

REFERENCES

(1) Ritchie, R. O. Nat. Mater. 2011, 10, 817−822. (2) Wegst, U. G. K.; Bai, H.; Saiz, E.; Tomsia, A. P.; Ritchie, R. O. Nat. Mater. 2014, 14, 23−36. (3) Zhang, G.; Li, X. Cryst. Growth Des. 2012, 12, 4306−4310. (4) Li, X.; Chang, W. C.; Chao, Y. J.; Wang, R.; Chang, M. Nano Lett. 2004, 4, 613−617. (5) Xu, Z. H.; Li, X. Adv. Funct. Mater. 2011, 21, 3883−3888. (6) Song, F.; Zhang, X. H.; Bai, Y. L. J. Mater. Res. 2002, 17, 1567− 1570. (7) Sun, J.; Bhushan, B. RSC Adv. 2012, 2, 7617. (8) Wang, R. Z.; Suo, Z.; Evans, A. G.; Yao, N.; Aksay, I. A. J. Mater. Res. 2001, 16, 2485−2493. (9) Huang, Z.; Pan, Z.; Li, H.; Wei, Q.; Li, X. J. Mater. Res. 2014, 29, 1573−1578. (10) Munch, E.; Launey, M. E.; Alsem, D. H.; Saiz, E.; Tomsia, A. P.; Ritchie, R. O. Science 2008, 322, 1516−1520. (11) Wang, J.; Cheng, Q.; Tang, Z. Chem. Soc. Rev. 2012, 41, 1111. (12) Podsiadlo, P.; Kaushik, A. K.; Arruda, E. M.; Waas, A. M.; Shim, B. S.; Xu, J.; Nandivada, H.; Pumplin, B. G.; Lahann, J.; Ramamoorthy, A.; Kotov, N. A. Science 2007, 318, 80−83. (13) Sellinger, A.; Weiss, P. M.; Nguyen, A.; Lu, Y.; Assink, R. A.; Gong, W.; Brinker, C. J. Nature 1998, 394, 256−260. (14) Mao, L.-B.; Gao, H.-L.; Yao, H.-B.; Liu, L.; Cölfen, H.; Liu, G.; Chen, S.-M.; Li, S.-K.; Yan, Y.-X.; Liu, Y.-Y.; Yu, S.-Ho. Science 2016, 354, 107−110. (15) Quigley, B. F.; Abbaschian, G. J.; Wunderlin, R.; Mehrabian, R. Metall. Trans. A 1982, 13, 93−100. (16) Guo, Z.; Pereira, T.; Choi, O.; Wang, Y.; Hahn, H. T. J. Mater. Chem. 2006, 16, 2800−2808. (17) Zabihi, M.; Toroghinejad, M. R.; Shafyei, A. Mater. Sci. Eng., A 2013, 560, 567−574. (18) Hodder, K. J.; Izadi, H.; McDonald, A. G.; Gerlich, A. P. Mater. Sci. Eng., A 2012, 556, 114−121. (19) Wang, P. W.; Sui, S.; Wang, W.; Durrer, W. Thin Solid Films 1997, 295, 142−146. (20) Jamaati, R.; Toroghinejad, M. R. Mater. Sci. Eng., A 2010, 527, 4146−4151. 6914

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915

Letter

Nano Letters (52) Zandiatashbar, A.; Lee, G.-H.; An, S. J.; Lee, S.; Mathew, N.; Terrones, M.; Hayashi, T.; Picu, C. R.; Hone, J.; Koratkar, N. Nat. Commun. 2014, 5, 3186. (53) Zhang, Y.; Pan, C. Diamond Relat. Mater. 2012, 24, 1−5. (54) Massalski, T. B.; Murray, J. L.; Bennett, L. H.; Baker, H. Binary alloy phase diagrams. American Society for Metals 1986, 159−160. (55) ASTM E8/E8M standard test methods for tension testing of metallic materials 1. Annu. B. ASTM Stand. 2010, 4, 1−27; 10.1520/ E0008. (56) Aluminum 6061-T6 - ASM Material Data Sheet - MatWeb. http://asm.matweb.com/search/SpecificMaterial.asp?bassnum= MA6061O. (57) Auerkari, P. Technol. Res. Cent. Finl. 1996, 1792, 26. (58) Breslin, M. C.; Ringnalda, J.; Xu, L.; Fuller, M.; Seeger, J.; Daehn, G. S.; Otani, T.; Fraser, H. L. Mater. Sci. Eng., A 1995, 195, 113−119. (59) Bouville, F.; Maire, E.; Meille, S.; Van de Moortèle, B.; Stevenson, A. J.; Deville, S. Nat. Mater. 2014, 13, 508−514. (60) Studart, A. R. Nat. Mater. 2014, 13, 433−435. (61) Anderson, T. L. Fracture mechanics fundamentals and applications; Taylor & Francis Group: Boca Raton, FL, 2005. (62) Naleway, S. E.; Porter, M. M.; McKittrick, J.; Meyers, M. A. Adv. Mater. 2015, 27, 5455−5476. (63) Cook, J.; Gordon, J. E.; Evans, C. C.; Marsh, D. M. Proc. R. Soc. London, Ser. A 1964, 282, 508−520. (64) Chen, P.-Y.; McKittrick, J.; Meyers, M. A. Prog. Mater. Sci. 2012, 57, 1492−1704. (65) Suk, J. W.; Piner, R. D.; An, J.; Ruoff, R. S. ACS Nano 2010, 4, 6557−6564.

6915

DOI: 10.1021/acs.nanolett.7b03308 Nano Lett. 2017, 17, 6907−6915