- Email: [email protected]
Laser brazing of Inconel® 718 using Ag and Cu-Ag nanopastes as brazing materials
Accepted Manuscript Title: Laser brazing of Inconel® 718 using Ag and Cu-Ag nanopastes as brazing materials Authors: Denzel Bridges, Chaoli Ma, Zane Palmer, Shutong Wang, Zhili Feng, Anming Hu PII: DOI: Reference:
S0924-0136(17)30238-8 http://dx.doi.org/doi:10.1016/j.jmatprotec.2017.06.010 PROTEC 15263
To appear in:
Journal of Materials Processing Technology
Received date: Revised date: Accepted date:
14-2-2017 1-6-2017 7-6-2017
Please cite this article as: Bridges, Denzel, Ma, Chaoli, Palmer, Zane, Wang, Shutong, Feng, Zhili, Hu, Anming, Laser brazing of Inconel® 718 using Ag and Cu-Ag nanopastes as brazing materials.Journal of Materials Processing Technology http://dx.doi.org/10.1016/j.jmatprotec.2017.06.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Laser brazing of Inconel ® 718 using Ag and Cu-Ag nanopastes as brazing materials Denzel Bridges1, Chaoli Ma1,2, Zane Palmer3, Shutong Wang1, Zhili Feng4, Anming Hu1,6* 1
Department of Mechanical, Aerospace, and Biomedical Engineering, University of Tennessee, Knoxville 37996 2
College of Materials Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, 210016, China. 3
Department of Material Science and Engineering, University of Tennessee, Knoxville 37996
4
Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
6
Institute of Laser Engineering, Beijing University of Technology, 100 Pingle Yuan, Chaoyang District, Beijing 100124, China Corresponding author: [email protected] Graphical abstract
Abstract Ag nanopastes composed of Ag nanoparticles or Ag nanowires and Cu-Ag nanopastes with Cu-Ag coreshell nanowires are used as a new brazing material for Inconel® 718. Ag nanoparticles or Ag nanowires are further added to the core-shell paste to adjust to a eutectic composition. Microstructural analysis of the brazed joints was carried out with EDS and XRD. High bonding strength (>100 MPa) was obtained with both Ag and Cu-Ag nanopastes. It was concluded that the Cu-Ag nanopastes form stronger braze joints than the BAg-8 brazing alloy as a result of Hall-Petch strengthening. It has also been concluded that the addition Ag nanoparticles or Ag nanowires to the Cu-Ag core-shell nanowire paste have a significant impact on the bonding strength and fracture of the Cu-Ag joints.
1
Keywords: brazing; ; ; ; ; , laser, Inconel, nanopaste, nanojoining, core-shell
1.
Introduction
Brazing is a versatile joining technique for a wide range of applications including automotive, aerospace, electronics, and construction. Brazing involves minimal microstructural changes and thermal distortion to the base metal which makes it preferable to welding in many manufacturing situations. To reduce any thermally-driven changes to the base material and lower the operating cost, the heat affected zone and the brazing temperature should be minimized. Laser brazing is a brazing technique that helps achieve the former through localized heat input. Laser brazing has high heating and cooling rates with precise control of the laser energy input. It has frequently been used in automotive and aerospace applications For example, Shaowei et al. (2016) used numerical simulations to model the thermal behavior of laserbrazed galvanized steel for the purpose of optimizing the laser brazing procedure, especially in an automotive manufacturing setting. For aerospace applications, Xiong et al. (2014) used a new Cu-Au-PdV filler metal to braze a carbon-fiber reinforced silicon carbide composite which has great potential as a structural material in high temperature aerospace applications. Melting point depressant elements are commonly added to brazing filler metals including boron, and silicon, however, the addition of such elements results in the formation of brittle intermetallic phases, which reduces the joining strength and ductility. In an effort to eliminate this obstacle, Huang (2014) developed a boron-free and silicon-free brazing filler metal using a combination of zirconium and hafnium as melting point depressants which resulted in complete metallic bonding with CMSX-4 and less detrimental intermetallic formation. Nanomaterials provide a potential solution to achieving a lower processing temperature without adding detrimental melting point depressants. The processing temperature can be decreased via a well-known effect on the melting temperature in which decreasing the particle size decreases the melting temperature. After joining, the material properties of 3D sintered nanoparticles will more closely resemble bulk material properties. Peng et al. (2015) reviewed several methods and studies in which silver nanomaterials are used as low temperature joining materials and how silver nanomaterials will increase in electrical conductivity and thermal conductivity as the degree of sintering increases, approaching the bulk conductivities of silver. Li et al. (2015c) have used this property to fabricate flexible electronics using a silver nanoplate ink. To avoid damaging temperature sensitive components in a SiC power electronic device, Manikam et al. (2013) mixed aluminum and silver nanoparticles into a paste for low temperature die attach and demonstrated its mechanical reliability after several thermal cycles. The bonding strength for low temperature nanojoining is typically limited to tens of MPa. For example, Li et al. (2015a) reports a maximum bonding strength of 29.4 MPa using silver nanoplates as a joining material. Limited studies have been conducted on joining nanomaterials at high temperatures. Hausner et al. (2016) is one of the few nanomaterial brazing studies published at the time of this paper. In this study, Hausner et al. brazed stainless steel using a nickel nanoparticle paste to achieve a maximum bonding strength of 120 MPa using induction brazing. Copper and silver based nanomaterials are popular nanomaterials for joining applications due to excellent mechanical, thermal, and electrical properties. Fu et al. (2014) used silver nanopastes as a die attach material for large area (>100 mm2) high temperature power electronics and demonstrated good bonding strength (50 MPa), good thermal conductivity (269 W/m-K), and thermal stability. Despite its
2
good chemical stability and mechanical properties, the price of silver greatly cripples its widespread application in many industries. Copper has higher mechanical strength and lower cost than silver. Hokita et al. (2015) and Kahler et al. (2012) have investigated copper nanoparticles as a cheaper alternative to silver for printable electronics and power electronics, respectively. Copper nanoparticles were combined with silver nanoparticles by Kim Seah and Kuan Yew (2014) in order to lower the cost without sacrificing the benefits of silver for an alternative die attach material . However, at elevated temperatures, copper can easily be oxidized at the nanoscale. Tsai et al. (2013) demonstrated this fact when comparing the thermal stability of copper-silver core-shell nanoparticles to bare copper nanoparticles. Organic capping layers or metallic coatings are often employed to stabilize copper nanomaterials in air. Metal coatings have the advantage over organic capping layers for high temperature applications because when organic capping layers decompose at high temperatures, decomposition byproducts and/or microvoids contaminate the joint and harm the mechanical, thermal, and electrical properties. Metal coatings will not produce any detrimental decomposition byproducts. Chen et al. (2015) used nickel to coat copper nanowires to prevent oxidation instead of an organic capping layer to fabricate transparent, flexible heaters. Lee et al. (2015) demonstrated the potential of silver as a capping layer for copper nanoparticles by using Cu-Ag core-shell nanoparticles for printing applications, in this study Ag played a functional role in the printed devices, whereas most organic capping layers do not. Currently, limited data exists on using nanomaterials as brazing filler metals or using core-shell nanomaterials for nanojoining applications. Inconel® 718 (Ni718) is a precipitation-hardened nickel superalloy with excellent mechanical properties, structural stability and corrosion resistance at elevated temperatures. Ye et al. (2015) demonstrated these qualities in their study of hot cracking of Ni718 welds. Ni718 is widely used in aerospace turbines, liquid fueled rockets, and nuclear reactors. Pouranvari et al. (2014) investigated diffusion brazing using a commercial Ni–7Cr–4.5Si–3.2B–3Fe brazing alloy to provide an alternative method of fabrication and repair of Ni718 for the aforementioned applications. A challenge when brazing nickel and high nickel content alloys with bulk brazing materials is the tendency of base materials to react with the melting point depressant as stated in the fourth edition of the AWS Brazing Handbook (2002). Khorram and Ghoreishi (2015) demonstrated that silver-based and copper-based brazing materials, such as BAg and BCu are suitable choices for brazing Ni718. To clarify the advantage of nanomaterials, the mechanical strength of the Ag and Cu-Ag nanopastes will also be compared to that of a eutectic Cu-Ag brazing material BAg-8.
2. Experimental section 2.1 Materials Silver nitrate, L-ascorbic acid, and ethylene glycol (EG) were purchased from Sigma Aldrich. Polyvinylpyrrolidone (K-30 PVP; M.W. = 30,000) and sodium chloride was purchased from Alfa Aesar. Cupric nitrate was purchased from Fisher Scientific. K-85-90 PVP (M.W. = 1,300,000) was purchased from Acros Organics. 601 and 601B silver brazing flux were purchased from Superior Flux & Mfg. Co. All chemicals were used without further purification. Ni718 was purchased from Rolled Alloys, Inc. BAg-8 is
3
was purchased from McMaster-Carr. The composition of Ni718, the silver brazing flux, and BAg-8 are reported in Table 1.
2.2 Preparation of Ag nanoparticle paste Silver nanoparticles (NPs) were synthesized by a polyol wet chemical method (Ma et al., 2016) . 10 mL of EG-based AgNO3 solution (0.9 M) and 190 mL of EG-based K-30 PVP (0.284 M) solution were prepared. All solutions were combined under magnetic stirring and heated to 150 °C for 15 min. After cooling naturally, Ag NPs were washed and concentrated using deionized (DI) water and centrifugation at 8000 rpm for 30 min two times and 12 min for the final centrifugation to remove EG and excess PVP.
2.3. Preparation of Ag nanowire paste Ag nanowires (NWs) were prepared by a similar polyol method as the Ag NPs. 10 mL of EG-based AgNO3 solution (0.9 M), 6 mL of EG-based NaCl solution, and 184 mL of EG-based K-30 PVP (0.284 M) solution were prepared. All solutions were combined under magnetic stirring and heated to 180 °C for 15 min. After cooling naturally, Ag nanowires (NWs) were washed and concentrated using deionized (DI) water and centrifugation at 5000 rpm for 30 min, then again for 12 min for the final centrifugation to remove EG and excess PVP. This synthesis method has been modified based on the synthesis method used by Li et al. (2014).
2.4 Preparation of Cu-Ag core-shell nanowire paste Cu-Ag core-shell nanowires (CSNWs) were synthesized by first making the copper core nanowire, then coating with silver by galvanic displacement reaction. The copper cores were fabricated by a facile hydrothermal method developed by (Zhang et al., 2006). 0.181 g of Cu(NO3)2 and 0.15 g of L-ascorbic acid were dissolved in 30 mL ultrapure DI water (resistance = 18 MΩ). After 5 minutes, 0.38 g of K-85-90 PVP was added and magnetically stirred until fully dissolved. In this reaction, ascorbic acid co-acts as a reducing agent and capping layer and PVP as a structure-directing agent. The solution was then autoclaved in a 50 mL Teflon-lined stainless steel autoclave at 120 °C for 4 hours. After cooling, the Cu NWs were collected from the autoclave and centrifuged at 2000 rpm for 5 min to remove excess reducing agent. Grouchko et al. (2009) reported that this step is critical to avoid nucleation of free standing Ag NPs. After centrifuging, the NWs were redispersed in fresh ultrapure DI water. The silver coating solution which consists of 2.5 mL ultrapure DI water, 32 mg AgNO3, and 75 mg K-85-90 PVP is based on the coating solution used by Zhao et al. (2015). The silver coating solution was slowly added to the Cu NWs at room temperature under moderate magnetic stirring overnight. The nominal Cu:Ag atomic ratio is 4:1. The resulting solution was centrifuged and concentrated three times at 2000 rpm for 5 min. The nanopastes used for laser brazing are summarized in Table 2 where the metal content is 5060%.
2.5 Laser brazing Ni718 pieces (3 mm x 0.32 mm x 30 mm) were ultrasonically cleaned in acetone for 2 min then plasma cleaned using a PDC-001 Expanded Plasma Cleaner (Harrick Plasma) at low power for 2 min. Each nanopaste (NPA) was applied to the Ni718 and most of the water was evaporated by baking on a hot plate at 75 °C. Silver brazing flux was applied on top of the nanopaste (601 silver brazing flux was used
4
for laser power <250 W and 601B flux was used for laser power ≥250 W). The two NI718 plates were placed 9 cm from a 400 W continuous wave high energy diode laser (λ = 806 nm) on a 1 mm thick copper plate with an overlap of 34 mm as shown in Fig. 1. The laser intensity profile at the focal point is similar to a top-hat intensity profile with a size of 1 mm x 22 mm. The laser irradiated the top surface of the Ni718 plate. The laser power was ramped to the target power in ~5 seconds then held at the target power for ~10 seconds/sample as seen in Fig. 1b. No pressure was applied to the sample. A lap shear test was performed using a tensile testing machine on the brazed samples.
2.6 Characterization Scanning Electron Microscopy (SEM) images were collected on a Zeiss Auriga Scanning Electron Microscope. X-ray diffraction measurements were conducted on a Panalytical Empyrean X-ray Diffractometer, Energy Dispersive X-ray Spectroscopy (EDS) measurements were performed on samples brazed with 300 W laser power using a Phenom ProX energy-dispersive spectrometer. Vickers microhardness of 300 W brazed samples was measured using a Buehler MMT3 Microhardness Tester with 5 gram-force. The hardness of Ni718 prior to brazing was measured from a cross-sectional piece of Ni718. For cross-sectional analysis, samples were embedded in an epoxy resin and polished using silica 0.04 µm paste.
3. Results and discussion 3.1 Characterization of nanopaste Fig. 2a and 2b show the bare Cu NWs and the Cu-Ag CSNWs respectively. Cu NWs have a smooth surface with a rectangular cross section and a 400-1000 nm diameter and 5-40 µm long. After the addition of AgNO3, Ag nanocrystals nucleated and grew on the surface of the Cu NW giving it a rough surface morphology. The CSNWs became hollow with a wall thickness of 40-80 nm and a pentagonal crosssection as seen in Fig. 2c. The Ag NWs have a smooth surface texture with a diameter of 40-50 nm and a length of 5-30 µm (Fig. 2d). The Ag NPs are spherical and 55-75 nm in diameter (Fig. 2e). As seen in 2f, the UV-vis spectrum of Cu NWs show that the surface plasmon resonance (SPR) peak at 582 nm with high absorbance across the spectrum of 200 nm to 960 nm; Wang and Ruan (2016) report a similar spectrum. After Ag coating, the SPR peak red shifts into the near-infrared/infrared range and the absorbance in the visible light range is greatly diminished. The Cu-Ag CSNWs reported in Wei et al. (2015) study exhibit similar behavior. The broad SPR peak of Ag NPs centered at 432 nm indicates a wide particle size distribution. As expected, Ag NPs absorb light in the red and near-infrared spectrum (550-1000 nm) is absorbed very weakly like in the report by Krishna Podagatlapalli et al. (2013). Like Cu NWs, Ag NWs have a high absorbance across the spectrum except its SPR peak is located at 378 nm and is consistent with the UV-vis spectrum in Tang and Tsuji (2010). XRD patterns for Cu NWs (Fig. 3a) only show XRD peaks characteristic to copper and no copper oxides were identified. After the Ag coating, Cu peaks do not disappear, but the peak intensity ratio for the Cu (111) and Cu (200) planes decrease from 3.44:1 to 2.85:1 (Fig. 3b). The change in relative peak intensity between Cu (111) and Cu (200) is likely caused by the Ag covering the Cu (111) plane, causing a decrease
5
in Cu (111) intensity. The ratio between the Ag (111) and Ag (200) peak intensity is 7.02:1. It is worth noting that the CSNW XRD pattern shows distinct Cu peaks and distinct Ag peaks, and no detection of copper oxides. During the centrifugation step to remove excess reducing agent, some of the organic capping layer (PVP) is also removed by washing. If the silver coating is not sufficient, the Cu NWs will easily develop an oxide layer on the surface of the NW in an aqueous environment (Kim et al., 2014). This indicates that CSNWs adopt a core-shell structure. The Ag NWs (Fig.3c) and Ag NPs (Fig. 3d) have similar peak positions but the relative intensities for the (220) and (311) peaks are much stronger for Ag NPs than Ag NWs.
3.2 Brazing joint microstructure COMSOL Multiphysics Suite was used to estimate the temperature and temperature distribution of the laser brazing samples. For simplicity, the nanopaste material is treated as dry powder with a thermal conductivity similar to a solid material. Due to the top-hat profile of the laser beam, the laser intensity is considered constant throughout the laser heated region. At the end of the brazing process at 300 W, the temperature in the laser heated region is nearly 1000 °C as seen in Fig. 4. A high temperature thermocouple was used to measure the temperature of the top plate and validate the simulation results. The average temperature in the brazing material increases to approximately 775 °C during the brazing as seen in Fig. 5b. There was no significant temperature difference between the pure Ag paste and the CuAg paste. According Fig. 5c, there is a temperature difference of ~300 °C between the top plate and the bottom plate. On the surface of the sample, the temperature varies in the overlap area between 750975 °C (Fig. 5d). Fig. 5e and 5f show that the temperature gradient across the NPA layer is extremely low with temperatures ranging between 760 °C and 800 °C. This result suggests that the nanopaste is heated evenly during the laser brazing procedure. For nanojoining, temperature gradient and heating speed are key factors as summarized by German (1996) in his sintering fundamentals textbook. A steep temperature gradient promotes dissolution of the base metal into the brazing material via thermal diffusion. The high heating speed of laser brazing helps achieve a large thermal gradient like the one demonstrated in this simulation. Even though, the thermal gradient is low within the nanopaste layer on a macroscopic scale, there is most likely the size dependent melting and liquid phase sintering drive the inter-particle joining. Surface diffusion is the dominant mass transport mechanism during surface melting of nanomaterials (at low temperature). However, due to the heating speed of the laser brazing technique, the surface melting of nanoparticles can induce the full melting of the whole nanoparticles and then a liquid phase sintering. According to Wang et al. (2016), the atoms at the Cu core of the nanostructure begin to melt at around 800-950 °C. Once the filler material is melted, the atoms easily diffuse into the base material or participate in joining of nearby particles.
After brazing at 300 W, the Ag NWs (Fig. 6a-b) and CSNWs (Fig. 6c-d) form a highly densified brazing joint and the nanowire structure has completely disappeared. This convinces a melting predicted in our simulation. High magnification images of the brazed joint reveal the presence of submicron voids in the brazed structure. The porosity of the Cu-Ag and Ag joint is much lower compared to other pressureless
6
low temperature joining techniques such as the technique used by Li et al. (2015d) to sinter a Ag nanowire-nanoplate composite paste for low temperature joining. All the EDS line scans were performed on a sample that was joined at 300 W. The line scan in Fig. 7b reveal uniform dissolution of the base material into the Ag brazing material and dissolution of Ag into the base material. The line scan reveals that there is a 3 µm diffusion layer between the base material and the brazing material. The light spots in Fig. 7c reveal the existence of secondary phases near the interface between the base material and brazing material. The presence of the diffusion layer and dissolution of the base material into the braze indicates that metallurgical bonding has occurred at the interface. Fig. 8 shows a brazing joint using the CSNWs. The EDS line scan for this joint shows that the dark colored regions are extremely Cu-rich (Fig. 8d). These Cu-rich regions contain on average 6 times more copper than silver. The light regions have a Cu:Ag ratio of 3:4. A higher resolution line scan (Fig. 8e-f) shows that there is a smooth transition between the Cu-rich phases and the Ag-rich phases with a diffusion layer of approximately 1.6 µm. The precipitation Cu-rich phases in a Ag-rich matrix is not observed in the vacuum brazing experiments conducted by Ma et al. (2016). One reason for such a phase separation is the miscibility gap in the Cu-Ag system. Another possible reason for this phase separation is the high heating and cooling speed of our laser brazing technique prevents homogenization of the brazed joint. The existence of Cu-rich phases and Ag-rich phases in the brazed joint are consistent with molecular dynamics simulations of sintering procedures of two Cu-Ag core-shell particles. Wang et al. (2016) displayed that at 727 °C the Ag shell will melt first before the Cu core melts (800-950 °C). The simulation also reveals that the Cu-core remains intact until the Ag shell has completely melted. Once the entire nanoparticle has melted, the core-shell structure disappears and alloying occurs. This is result is consistent with the microscale phase separation observed in Cu-Ag CSNW brazing. A similar amount of the Ni718 is dispersed throughout the brazing joint. However, the amount of brazing material dispersed in the base material is approximately 33% more when using CSNWs compared to using Ag NWs as a brazing material. This brazing joint has a diffusion layer of 2 µm which is 33% smaller than the diffusion layer using Ag NWs. XRD was performed on the Cu-Ag materials after brazing at 300 W (Fig. 9). The relative intensity of the Cu* peaks are significantly higher in the BAg-8 brazed joints than the Cu* peaks in the Cu-1 and Cu-3 brazed joints. However, the relative intensities in the BAg-8 joint are very similar to the Cu-2 joint except for the Cu* (220) peak (65.3°). Compared to the BAg-8 alloy, most of the major peaks from the NPAbrazed joints were broader than those of the brazing alloy. Broader XRD peaks suggest that the final grain size is smaller for NPA brazed joint than BAg-8 brazed joints. The significance of this detail will be discussed later.
3.3 Mechanical Properties 3.3.1 Hardness
The hardness profile of the Ag and Cu-Ag brazed joint is plotted in Fig. 10. The Ag brazing material (Fig. 10a) has an average Vickers hardness of 71 HV within the brazing region. The plate that was directly irradiated by the laser is noted in the figure. The hardness of the interface between the brazing material and the base material is slightly lower on the laser irradiated side than the hardness of Ni718 prior to
7
brazing (indicated by the red dotted line). The hardness of the base material returns to the original hardness of Ni718 after a brief increase in hardness. The hardness of interface between the non-laser irradiated Ni718 and the Ag braze is greater than the hardness at the laser-irradiated interface. Hardness values for 100 μm outside the brazing region return to the original hardness value for Ni718 on both sides of the brazing region. The Cu-Ag brazing material is on average almost twice as hard as the Ag joint (Fig. 10b). The hardness of the Ni718, especially at the interface is much more greatly affected by the Cu-Ag CSNWs as well. This change in hardness is likely caused more significant compositional change when using Cu-1 compared to Ag-2 (as seen in Fig. 7b and 8b). Like the Ag joint in Fig. 10a, the laser irradiated interface has a lower hardness than the non-laser irradiated interface. The only difference is how the non-laser irradiated plate is affected. On laser irradiated side, the hardness values regress to the original hardness of Ni718 100 μm outside the brazing region. On the non-laser side, the hardness values regress to the original hardness of Ni718 200 μm outside the brazing region. On both sides of the brazing region the hardness is larger than the original Ni718 value before returning to normal values. 3.3.2 Bonding strength
Fig. 11 presents the bonding strength of each NPA as a function of laser power. High bonding strength (>100 MPa) was obtained for all NPAs except the Ag-1. Both silver nanomaterials have a laser power threshold of 250 W, however the Ag-2 NPA has superior bonding strength (116.3 MPa) compared to Ag-1 (49.1 MPa). The pure Cu-1 paste exhibits a steady increase in bonding strength as a function of laser power until it reaches a 224 MPa bonding strength at 300 W. The Cu-2 NPA has the highest bonding strength with a maximum bonding strength of 249.6 MPa. The increase in strength compared to the pure CSNWs is because the Ag NPs increase the green density of the paste prior to brazing. Guo et al. (2017) demonstrated that mixing NPs into a NW paste increases the green density for sintering by using a mixed Ag NW and Cu NP or Ag NP paste. The large increase in bonding strength of the Cu-2 paste displays a between 200 W and 250 W corresponds with minimum laser power for joining of Ag NPs, suggesting that the NPs melting in the paste is a key occurrence that strengthens the Cu-Ag joint. However, the effect of adding Ag NPs is limited because the bonding strength of the Cu-2 paste is greater than the pure CSNWs paste only at 250 W and 300 W laser power and only by a small margin. Mixing Ag NWs into the CSNWs (i.e. Cu-3) effectively lowers the laser power required for successful joining (150 W). Cu-3 has a lower threshold laser power compared to Ag-1, Ag-2, and Cu-1 due to the Cu:Ag ratio shifting to the eutectic composition. Even in a bulk material, the eutectic temperature of the Cu-Ag system is 183.7 °C lower than the melting point of pure silver and 61.9 °C lower than the liquidus temperature of bulk Cu-1 (i.e., 80 at% Cu, 20 at% Ag). The possible reason for why Cu-2 does not have the same bonding threshold as Cu-3 is that Ag NWs have been shown in numerous studies to have a lower minimum bonding temperature than Ag NPs via the shape effect. The shape effect, as the name suggests is the effect of shape on the melting temperature of a nanomaterial. presents a model via the Lindemann melting criterion that NWs have a lower specific heat than NPs. Stacey and Irvine (1977) provide a thermodynamic basis for the Lindemann melting criterion which they mention is a theory that explains the melting mechanism of a material using the vibration of atoms in a crystal. Li et al. (2015b) demonstrated that the bonding strength of Ag NWs becomes constant when a certain bonding temperature even when mixed with a differently shaped nanomaterial which is possibly why the laser power has very little effect on the bonding strength of the Cu-3 paste (~160 MPa).
8
The Cu-Ag NPAs have higher bonding strength than the BAg-8 primarily due to Hall-Petch strengthening. According to the Hall-Petch relation, the yield strength of a material increases as the grain size decreases. Ebrahimi et al. (1999) reports an increase in yield strength in electrodeposited nanocrystalline nickel when the grain size decreases. Using the Scherrer formula ( τ= 0.9λ/(β cosθ) ), the grain sizes were calculated based on the XRD peaks in Fig. 9. τ is the mean grain size, λ is the x-ray wavelength, β is the full-width half maximum, and θ is the Bragg angle. The results of these calculations are presented in Table 3. For the NPAs, the calculated grain sizes are smaller than BAg-8 on average except for a couple of exceptions. The Cu-1 joint contains Ag* (111) and Cu* (220) oriented grains (38.7° and 65.3°, respectively) that are roughly the same size as the Ag* (111) and Cu* (220) oriented grains in the BAg-8 joint. Based on calculations of the Ag* (311) peak (located at 74.1°), some of the Ag-rich grains in the Cu3 joint are larger than those found in the BAg-8 joint. The Cu-2 joint contains larger Cu-rich grains than the BAg-8 joint based on the Cu* (111) and Cu* (311) peaks (43.5° and 78.4°, respectively). Overall, the NPA brazed joints have much smaller grain sizes than BAg-8 brazed joints. Therefore, the superior bonding strength of the Cu-Ag NPAs can be heavily attributed to Hall-Petch strengthening of the brazing material. There are a few possible explanations for this: 1. The NPAs contain small amounts of organic compounds leftover after the washing procedure. Even though the organic compounds in question decomposed before solidification, the decomposition byproducts still remain and provide a source of defects within the brazing material. These defects may provide additional nucleation sites for the molten brazing material during solidification. BAg-8 does not contain any organic compounds within the material so these impurity-based defects would not exist in a brazed joint made by BAg-8. 2. The porosity of the NPAs may also have an effect on the nucleation and growth behavior of the NPAs. The initial porosity may have an effect on the interdiffusion behavior of Cu, Ag, and the base material. In the early stages of brazing, before the material fully melts and spreads, interdiffusion only occurs in the areas that the NMs are in direct contact with the base material. This would create local variations in the chemical composition on the Ni718 surface and thus change the nucleation and growth behavior of the base material. A more in-depth investigation of the melting and solidification behavior of NMs as a brazing filler metal is underway. The higher yield strength was confirmed by the stress-strain curves in Fig. 12 The Cu-3 brazed joint has an offset yield strength of 80 MPa and the BAg-8 brazed joint has an offset yield strength of 50 MPa.
9
According to the stress strain curve the BAg-8experiences more necking during fracture than the CSNWs, implying a more ductile failure mechanism in the BAg-8 joint. The fracture surfaces shown in Fig. 13 indicate that fracture occurs within the paste, not at the interface between the braze material and the base material. The Ag joints (Fig. 13b-c) have dimple fracture surface, indicating strong bonding. The dimples from the Ag-2 paste are much larger than the dimples in the Ag NP fracture surface. The Cu-1 joint (Fig. 13d) has a dimple-cleavage fracture surface. The Cu-2 paste fracture surface (Fig. 13e) is very similar to the fracture surface the Ag-1 paste (Fig. 13b). The similarities in the Cu-2 paste and Ag-1 paste fracture surfaces further supports the conclusion that the addition of Ag NPs greatly influences the bonding behavior of Cu-Ag CSNWs. The main difference is that the fracture surface of the Cu-2 paste has larger, deeper dimples than the pure Ag-1 paste structure. The Cu-3 paste fracture surface (Fig. 13f) primarily resembles the fracture surface of the CSNWs, however the dimple features become more apparent after Ag NWs are added to the CSNWs paste. The emphasized dimple features in the Cu-3 paste fracture surface demonstrate that the Ag NWs also have a dominant effect on the bonding/fracture of the Cu-Ag joint.
4. Conclusions Cu-Ag and Ag nanomaterials exhibit high bonding strength and higher joint density than previous nanojoining studies. Elemental analysis reveals dissolution of the base material throughout the brazing material. For Cu-2 and Cu-3 nanopastes, the Ag NPs and Ag NWs play an important role in the bonding performance and fracture mechanism of joints fabricated by mixed Cu-Ag NPAs. The superior bonding strength of Cu-Ag NPAs compared to the BAg-8 filler metal is attributed to Hall-Petch strengthening.
Acknowledgements This project is jointly supported by Institute of Public Service, University of Tennessee through I’UCRC project of National Science Foundation and a seed grant from Oak Ridge National Laboratory, Department of Energy, USA. Chris Wetteland and Maneel Bharadwaj assisted in the collection of EDS and microhardness measurements. John Dunlap, Maulik Patel, and the Joint Institute of Advanced Materials permitted use of their electron microscopy facilities and X-ray diffraction equipment provided the training needed to operate those devices.
References Chen, J., Chen, J., Li, Y., Zhou, W., Feng, X., Huang, Q., Zheng, J., Liu, R., Ma, Y., Huang, W., 2015. Enhanced oxidation-resistant Cu-Ni core-shell nanowires: controllable one-pot synthesis and solution processing to transparent flexible heaters. Nanoscale 7, 16874-16879. Ebrahimi, F., Bourne, G.R., Kelly, M.S., Matthews, T.E., 1999. Mechanical properties of nanocrystalline nickel produced by electrodeposition. Nanostruct Mater 11, 343-350.
10
Fu, S., Mei, Y., Lu, G.-Q., Li, X., Chen, G., Chen, X., 2014. Pressureless sintering of nanosilver paste at low temperature to join large area (≥100mm2) power chips for electronic packaging. Materials Letters 128, 42-45. German, R.M., 1996. Sintering Theory and Practice. John Wiley & Sons, Inc., New York, NY. Grouchko, M., Kamyshny, A., Magdassi, S., 2009. Formation of air-stable copper–silver core–shell nanoparticles for inkjet printing. Journal of Materials Chemistry 19, 3057. Guo, W., Zhang, H.Q., Zhang, X.Y., Liu, L., Peng, P., Zou, G.S., Zhou, Y.N., 2017. Preparation of nanoparticle and nanowire mixed pastes and their low temperature sintering. Journal of Alloys and Compounds 690, 86-94. Hausner, S., Weis, S., Wagner, G., 2016. Joining of steels at low temperatures by Ni nanoparticles, 11th International Conference on Brazing, High Temperature Brazing and Diffusion Bonding. Dusseldorf: DVS Media GmbH, Aachen, Germany, pp. 278-284. Hokita, Y., Kanzaki, M., Sugiyama, T., Arakawa, R., Kawasaki, H., 2015. HighConcentration Synthesis of Sub-10-nm Copper Nanoparticles for Application to Conductive Nanoinks. ACS Appl Mater Interfaces 7, 19382-19389. Huang, X., 2014. Brazing of CMSX-4 with a Boron- and Silicon-Free Ni-Co-Zr-Hf-Cr-TiAl Brazing Alloy. Weld J 93, 233s-242s. Kahler, J., Heuck, N., Wagner, A., Stranz, A., Peiner, E., Waag, A., 2012. Sintering of Copper Particles for Die Attach. IEEE Transactions on Components, Packaging and Manufacturing Technology 2, 1587-1591. Khorram, A., Ghoreishi, M., 2015. Comparative study on laser brazing and furnace brazing of Inconel 718 alloys with silver based filler metal. Optics & Laser Technology 68, 165-174. Kim, C.K., Lee, G.-J., Lee, M.K., Rhee, C.K., 2014. A novel method to prepare Cu@Ag core–shell nanoparticles for printed flexible electronics. Powder Technology 263, 1-6. Kim Seah, T., Kuan Yew, C., 2014. Physical and Electrical Characteristics of SilverCopper Nanopaste as Alternative Die-Attach. IEEE Transactions on Components, Packaging and Manufacturing Technology 4, 8-15. Krishna Podagatlapalli, G., Hamad, S., Tewari, S.P., Sreedhar, S., Prasad, M.D., Venugopal Rao, S., 2013. Silver nano-entities through ultrafast double ablation in aqueous media for surface enhanced Raman scattering and photonics applications. Journal of Applied Physics 113, 073106. Lee, C., Kim, N.R., Koo, J., Lee, Y.J., Lee, H.M., 2015. Cu-Ag core-shell nanoparticles with enhanced oxidation stability for printed electronics. Nanotechnology 26. Li, J., Johnson, C.M., Buttay, C., Sabbah, W., Azzopardi, S., 2015a. Bonding strength of multiple SiC die attachment prepared by sintering of Ag nanoparticles. Journal of Materials Processing Technology 215, 299-308. Li, R.-Z., Zhang, T., Hu, A., Bridges, D., 2015b. Ag Nanowire and Nanoplate Composite Paste for Low Temperature Bonding. Materials Transactions 56, 984-987. Li, R.Z., Hu, A., Zhang, T., Oakes, K.D., 2014. Direct writing on paper of foldable capacitive touch pads with silver nanowire inks. ACS Appl Mater Interfaces 6, 2172121729. Li, R.Z., Hu, A.M., Bridges, D., Zhang, T., Oakes, K.D., Peng, R., Tumuluri, U., Wu, Z.L., Feng, Z.L., 2015c. Robust Ag nanoplate ink for flexible electronics packaging. Nanoscale 7, 7368-7377.
11
Li, R.Z., Zhang, T., Hu, A.M., Bridges, D., 2015d. Ag Nanowire and Nanoplate Composite Paste for Low Temperature Bonding. Materials Transactions 56, 984-987. Ma, C., Xue, S., Bridges, D., Palmer, Z., Feng, Z., Hu, A., 2016. Low Temperature Brazing Nickel with Ag Nanoparticle and Cu-Ag Core-Shell Nanowire Nanopastes. submitted to Journal of Materials Science. Manikam, V.R., Razak, K.A., Cheong, K.Y., 2013. Reliability of sintered Ag80–Al20 die attach nanopaste for high temperature applications on SiC power devices. Microelectronics Reliability 53, 473-480. Peng, P., Hu, A.M., Gerlich, A.P., Zou, G.S., Liu, L., Zhou, Y.N., 2015. Joining of Silver Nanomaterials at Low Temperatures: Processes, Properties, and Applications. Acs Appl Mater Inter 7, 12597-12618. Pouranvari, M., Ekrami, A., Kokabi, A.H., 2014. Diffusion brazing of cast INCONEL 718 superalloy utilising standard heat treatment cycle. Mater Sci Tech-Lond 30, 109-115. Shaowei, L., Zhishui, Y., Peilei, Z., Xiaoliang, L., Dingding, Q., Xiaopeng, L., 2016. Numerical Simulation and Parameters Optimization of Laser Brazing of Galvanized Steel. ISIJ International 56, 637-646. Soldering, A.C.o.B.a., 2002. Brazing Handbook, 4 ed. American Welding Society, Miami, Florida, United States of America. Stacey, F.D., Irvine, R.D., 1977. Theory of Melting: Thermodynamic Basis of Lindemann's Law. Australian Journal of Physics 30, 631. Tang, X.L., Tsuji, M., 2010. Syntheses of Silver Nanowires in Liquid Phase. Nanowires Science and Technology, 25-42. Tsai, C.H., Chen, S.Y., Song, J.M., Chen, I.G., Lee, H.Y., 2013. Thermal stability of Cu@Ag core-shell nanoparticles. Corros Sci 74, 123-129. Wang, J.Q., Shin, S.H., Hu, A.M., 2016. Geometrical Effects on Sintering Dynamics of Cu-Ag Core-Shell Nanoparticles. J Phys Chem C 120, 17791-17800. Wang, R.L., Ruan, H.B., 2016. Synthesis of copper nanowires and its application to flexible transparent electrode. Journal of Alloys and Compounds 656, 936-943. Wei, Y., Chen, S., Lin, Y., Yang, Z.M., Liu, L., 2015. Cu-Ag core-shell nanowires for electronic skin with a petal molded microstructure. Journal of Materials Chemistry C 3, 9594-9602. Xiong, H.-P., Chen, B., Pan, Y., Zhao, H.-S., Ye, L., 2014. Joining of Cf/SiC composite with a Cu–Au–Pd–V brazing filler and interfacial reactions. Journal of the European Ceramic Society 34, 1481-1486. Ye, X., Hua, X.M., Wu, Y.X., Lou, S.N., 2015. Precipitates in coarse-grained heataffected zone of Ni-based 718 superalloy produced by tungsten inert gas welding. Journal of Materials Processing Technology 217, 13-20. Zhang, X., Zhang, D., Ni, X., Zheng, H., 2006. One-step preparation of copper nanorods with rectangular cross sections. Solid State Commun 139, 412-414. Zhao, J., Zhang, D., Zhang, X., 2015. Preparation and characterization of copper/silver bimetallic nanowires with core-shell structure. Surface and Interface Analysis 47, 529534.
12
Fig.1: (a) Schematic of the laser brazing set up and (b) laser power curve
Fig. 2: SEM images of a) Cu NW s, b) Cu -Ag CSNW s, c) magnified view of Cu -Ag CSNW s, d) Ag NW s, and e) Ag NPs; f) the UV -vis spectra of each nanomaterial.
13
Fig. 3: XRD pattern of a) Cu NW s, b) Cu -Ag CSNW s, c) Ag NW s and d) Ag NPs an inlet of each nanomaterial solution is included
14
Fig. 4: (a) Thermal distribution after a 5 second ramp and 10 seconds of heating at 300 W (b) Global maximum, minimum and average temperatures as a function of time
15
Fig. 5: (a) Locations of the cross-sectional thermal distributions taken for parts d,e, and f, (b) average temperature of the brazing layer as a function of time. Thermal distribution of (c) xz-plane cross-section in the middle of the laser brasing region (d) the surface top plate, (e) interface between the top plate and the brazing layer, (f) interface between the bottom plate and the brazing layer
16
Fig. 6: SEM of the cross section of a brazed joint using (a -b) Ag NW s and (c-d) Cu-Ag CSNW s as brazing materials. The red box in (a) and (c)indicates where a high magnification image of the microstructure (b & d).
17
Fig.7: EDS line scans of a brazing joint using Ag NW s (a&b) are for a line scan of the entire joint; c&d are for the interfacial area). The solid red arrow indicates the location of the line scan. The dotted lines in d) are boundaries for the diffusion layer
18
Fig. 8: EDS line scans of a brazing joint using Cu -Ag CSNW s (a & b) are for a line scan of the entire joint; c&d are for the interfacial area. The solid red line is the location of the line scan. The black dotted line in c) is the boundary be tween the Ni718 and the brazed joint. The dotted lines in (d) and (f) indicate where the locations of the diffusion layers. (e&f): the local line scan performed on a Ag rich region adjacent to a Cu -rich region as indicated by a short arrow bar in (e)
19
Fig. 9: XRD patterns after laser joining at 300 W for (a) BAg -8brazing alloy, (b) Cu-1, (c) Cu-3, and (d) Cu-2. Ag* is the Ag-rich phase and Cu* is the Cu-rich phase.
20
Fig. 10: Vickers hardness of 300 W brazed joints within the brazing region (BR) and 50 µm outside the brazing region (BR) using a) Ag -2 and b) Cu-1. The side of laser incidence is labeled. The red dotted line is the average hardness of Ni718 prior to brazing
Fig.11: Bonding strength vs laser powe r for the NPAs and the BAg -8 alloy
21
Fig. 12: Typical Stress-Strain curve of a Cu-3 joint (a) and a BAg-8 joint (b). The yield strength (σ y ) is also shown on each curve
22
Fig. 13: (a) SEM image of the surface of bare NI718. Fracture surface of Ni718 bonded at 300 W using (b) Ag-1, (c) Ag-2, (d) Cu-1, (e) Cu-2, and (f) Cu-3
Table 1: Composition of the alloys and fluxes used in this study Material Ni718 BAg-8
Ni 53.10 Cu 28
Fe 18.40 Ag 72
Cr 18.30
Cu 0.05
Chemical composition Al Si C Mn 0.49 0.07 0.05 0.24
B 0.004
Nb 4.95
Ti 1.07
S <0.002
Mo 3.06
23
601/601B Flux
KBF4 20-40
K2B4O7 15-25
H3BO3 30-40
H2B2FO2 7-15
Table 2: Description of the nanopastes used in this laser brazing study
Paste Ag-1 Ag-2 Cu-1 Cu-2 Cu-3
Cu (wt%) 0 0 80 28 28
Ag (wt%) 100 100 20 72 72
Description Silver nanoparticles paste Silver nanowire paste Copper-silver CSNWs paste CSNWs and silver nanoparticles composite paste CSNWs and silver nanowires composite paste
Table 3: Grain size (in nm) after brazing of Cu -Ag brazing materials based on XRD results
Material BAg-8 Cu-1 Cu-2 Cu-3
38.7 54.7 54.8 41.2 36.5
43.5 67.7 41.8 66.8 41.8
45 48.0 42.0 33.6 33.7
2θ (Degrees) 50.5 28.6 34.4 21.5 24.5
65.3 46.1 46.1 36.9 15.3
74.1 48.6 24.3 17.7 48.8
78.4 33.4 25.0 40.1 25.0
24
S0924-0136(17)30238-8 http://dx.doi.org/doi:10.1016/j.jmatprotec.2017.06.010 PROTEC 15263
To appear in:
Journal of Materials Processing Technology
Received date: Revised date: Accepted date:
14-2-2017 1-6-2017 7-6-2017
Please cite this article as: Bridges, Denzel, Ma, Chaoli, Palmer, Zane, Wang, Shutong, Feng, Zhili, Hu, Anming, Laser brazing of Inconel® 718 using Ag and Cu-Ag nanopastes as brazing materials.Journal of Materials Processing Technology http://dx.doi.org/10.1016/j.jmatprotec.2017.06.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Laser brazing of Inconel ® 718 using Ag and Cu-Ag nanopastes as brazing materials Denzel Bridges1, Chaoli Ma1,2, Zane Palmer3, Shutong Wang1, Zhili Feng4, Anming Hu1,6* 1
Department of Mechanical, Aerospace, and Biomedical Engineering, University of Tennessee, Knoxville 37996 2
College of Materials Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, 210016, China. 3
Department of Material Science and Engineering, University of Tennessee, Knoxville 37996
4
Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
6
Institute of Laser Engineering, Beijing University of Technology, 100 Pingle Yuan, Chaoyang District, Beijing 100124, China Corresponding author: [email protected] Graphical abstract
Abstract Ag nanopastes composed of Ag nanoparticles or Ag nanowires and Cu-Ag nanopastes with Cu-Ag coreshell nanowires are used as a new brazing material for Inconel® 718. Ag nanoparticles or Ag nanowires are further added to the core-shell paste to adjust to a eutectic composition. Microstructural analysis of the brazed joints was carried out with EDS and XRD. High bonding strength (>100 MPa) was obtained with both Ag and Cu-Ag nanopastes. It was concluded that the Cu-Ag nanopastes form stronger braze joints than the BAg-8 brazing alloy as a result of Hall-Petch strengthening. It has also been concluded that the addition Ag nanoparticles or Ag nanowires to the Cu-Ag core-shell nanowire paste have a significant impact on the bonding strength and fracture of the Cu-Ag joints.
1
Keywords: brazing; ; ; ; ; , laser, Inconel, nanopaste, nanojoining, core-shell
1.
Introduction
Brazing is a versatile joining technique for a wide range of applications including automotive, aerospace, electronics, and construction. Brazing involves minimal microstructural changes and thermal distortion to the base metal which makes it preferable to welding in many manufacturing situations. To reduce any thermally-driven changes to the base material and lower the operating cost, the heat affected zone and the brazing temperature should be minimized. Laser brazing is a brazing technique that helps achieve the former through localized heat input. Laser brazing has high heating and cooling rates with precise control of the laser energy input. It has frequently been used in automotive and aerospace applications For example, Shaowei et al. (2016) used numerical simulations to model the thermal behavior of laserbrazed galvanized steel for the purpose of optimizing the laser brazing procedure, especially in an automotive manufacturing setting. For aerospace applications, Xiong et al. (2014) used a new Cu-Au-PdV filler metal to braze a carbon-fiber reinforced silicon carbide composite which has great potential as a structural material in high temperature aerospace applications. Melting point depressant elements are commonly added to brazing filler metals including boron, and silicon, however, the addition of such elements results in the formation of brittle intermetallic phases, which reduces the joining strength and ductility. In an effort to eliminate this obstacle, Huang (2014) developed a boron-free and silicon-free brazing filler metal using a combination of zirconium and hafnium as melting point depressants which resulted in complete metallic bonding with CMSX-4 and less detrimental intermetallic formation. Nanomaterials provide a potential solution to achieving a lower processing temperature without adding detrimental melting point depressants. The processing temperature can be decreased via a well-known effect on the melting temperature in which decreasing the particle size decreases the melting temperature. After joining, the material properties of 3D sintered nanoparticles will more closely resemble bulk material properties. Peng et al. (2015) reviewed several methods and studies in which silver nanomaterials are used as low temperature joining materials and how silver nanomaterials will increase in electrical conductivity and thermal conductivity as the degree of sintering increases, approaching the bulk conductivities of silver. Li et al. (2015c) have used this property to fabricate flexible electronics using a silver nanoplate ink. To avoid damaging temperature sensitive components in a SiC power electronic device, Manikam et al. (2013) mixed aluminum and silver nanoparticles into a paste for low temperature die attach and demonstrated its mechanical reliability after several thermal cycles. The bonding strength for low temperature nanojoining is typically limited to tens of MPa. For example, Li et al. (2015a) reports a maximum bonding strength of 29.4 MPa using silver nanoplates as a joining material. Limited studies have been conducted on joining nanomaterials at high temperatures. Hausner et al. (2016) is one of the few nanomaterial brazing studies published at the time of this paper. In this study, Hausner et al. brazed stainless steel using a nickel nanoparticle paste to achieve a maximum bonding strength of 120 MPa using induction brazing. Copper and silver based nanomaterials are popular nanomaterials for joining applications due to excellent mechanical, thermal, and electrical properties. Fu et al. (2014) used silver nanopastes as a die attach material for large area (>100 mm2) high temperature power electronics and demonstrated good bonding strength (50 MPa), good thermal conductivity (269 W/m-K), and thermal stability. Despite its
2
good chemical stability and mechanical properties, the price of silver greatly cripples its widespread application in many industries. Copper has higher mechanical strength and lower cost than silver. Hokita et al. (2015) and Kahler et al. (2012) have investigated copper nanoparticles as a cheaper alternative to silver for printable electronics and power electronics, respectively. Copper nanoparticles were combined with silver nanoparticles by Kim Seah and Kuan Yew (2014) in order to lower the cost without sacrificing the benefits of silver for an alternative die attach material . However, at elevated temperatures, copper can easily be oxidized at the nanoscale. Tsai et al. (2013) demonstrated this fact when comparing the thermal stability of copper-silver core-shell nanoparticles to bare copper nanoparticles. Organic capping layers or metallic coatings are often employed to stabilize copper nanomaterials in air. Metal coatings have the advantage over organic capping layers for high temperature applications because when organic capping layers decompose at high temperatures, decomposition byproducts and/or microvoids contaminate the joint and harm the mechanical, thermal, and electrical properties. Metal coatings will not produce any detrimental decomposition byproducts. Chen et al. (2015) used nickel to coat copper nanowires to prevent oxidation instead of an organic capping layer to fabricate transparent, flexible heaters. Lee et al. (2015) demonstrated the potential of silver as a capping layer for copper nanoparticles by using Cu-Ag core-shell nanoparticles for printing applications, in this study Ag played a functional role in the printed devices, whereas most organic capping layers do not. Currently, limited data exists on using nanomaterials as brazing filler metals or using core-shell nanomaterials for nanojoining applications. Inconel® 718 (Ni718) is a precipitation-hardened nickel superalloy with excellent mechanical properties, structural stability and corrosion resistance at elevated temperatures. Ye et al. (2015) demonstrated these qualities in their study of hot cracking of Ni718 welds. Ni718 is widely used in aerospace turbines, liquid fueled rockets, and nuclear reactors. Pouranvari et al. (2014) investigated diffusion brazing using a commercial Ni–7Cr–4.5Si–3.2B–3Fe brazing alloy to provide an alternative method of fabrication and repair of Ni718 for the aforementioned applications. A challenge when brazing nickel and high nickel content alloys with bulk brazing materials is the tendency of base materials to react with the melting point depressant as stated in the fourth edition of the AWS Brazing Handbook (2002). Khorram and Ghoreishi (2015) demonstrated that silver-based and copper-based brazing materials, such as BAg and BCu are suitable choices for brazing Ni718. To clarify the advantage of nanomaterials, the mechanical strength of the Ag and Cu-Ag nanopastes will also be compared to that of a eutectic Cu-Ag brazing material BAg-8.
2. Experimental section 2.1 Materials Silver nitrate, L-ascorbic acid, and ethylene glycol (EG) were purchased from Sigma Aldrich. Polyvinylpyrrolidone (K-30 PVP; M.W. = 30,000) and sodium chloride was purchased from Alfa Aesar. Cupric nitrate was purchased from Fisher Scientific. K-85-90 PVP (M.W. = 1,300,000) was purchased from Acros Organics. 601 and 601B silver brazing flux were purchased from Superior Flux & Mfg. Co. All chemicals were used without further purification. Ni718 was purchased from Rolled Alloys, Inc. BAg-8 is
3
was purchased from McMaster-Carr. The composition of Ni718, the silver brazing flux, and BAg-8 are reported in Table 1.
2.2 Preparation of Ag nanoparticle paste Silver nanoparticles (NPs) were synthesized by a polyol wet chemical method (Ma et al., 2016) . 10 mL of EG-based AgNO3 solution (0.9 M) and 190 mL of EG-based K-30 PVP (0.284 M) solution were prepared. All solutions were combined under magnetic stirring and heated to 150 °C for 15 min. After cooling naturally, Ag NPs were washed and concentrated using deionized (DI) water and centrifugation at 8000 rpm for 30 min two times and 12 min for the final centrifugation to remove EG and excess PVP.
2.3. Preparation of Ag nanowire paste Ag nanowires (NWs) were prepared by a similar polyol method as the Ag NPs. 10 mL of EG-based AgNO3 solution (0.9 M), 6 mL of EG-based NaCl solution, and 184 mL of EG-based K-30 PVP (0.284 M) solution were prepared. All solutions were combined under magnetic stirring and heated to 180 °C for 15 min. After cooling naturally, Ag nanowires (NWs) were washed and concentrated using deionized (DI) water and centrifugation at 5000 rpm for 30 min, then again for 12 min for the final centrifugation to remove EG and excess PVP. This synthesis method has been modified based on the synthesis method used by Li et al. (2014).
2.4 Preparation of Cu-Ag core-shell nanowire paste Cu-Ag core-shell nanowires (CSNWs) were synthesized by first making the copper core nanowire, then coating with silver by galvanic displacement reaction. The copper cores were fabricated by a facile hydrothermal method developed by (Zhang et al., 2006). 0.181 g of Cu(NO3)2 and 0.15 g of L-ascorbic acid were dissolved in 30 mL ultrapure DI water (resistance = 18 MΩ). After 5 minutes, 0.38 g of K-85-90 PVP was added and magnetically stirred until fully dissolved. In this reaction, ascorbic acid co-acts as a reducing agent and capping layer and PVP as a structure-directing agent. The solution was then autoclaved in a 50 mL Teflon-lined stainless steel autoclave at 120 °C for 4 hours. After cooling, the Cu NWs were collected from the autoclave and centrifuged at 2000 rpm for 5 min to remove excess reducing agent. Grouchko et al. (2009) reported that this step is critical to avoid nucleation of free standing Ag NPs. After centrifuging, the NWs were redispersed in fresh ultrapure DI water. The silver coating solution which consists of 2.5 mL ultrapure DI water, 32 mg AgNO3, and 75 mg K-85-90 PVP is based on the coating solution used by Zhao et al. (2015). The silver coating solution was slowly added to the Cu NWs at room temperature under moderate magnetic stirring overnight. The nominal Cu:Ag atomic ratio is 4:1. The resulting solution was centrifuged and concentrated three times at 2000 rpm for 5 min. The nanopastes used for laser brazing are summarized in Table 2 where the metal content is 5060%.
2.5 Laser brazing Ni718 pieces (3 mm x 0.32 mm x 30 mm) were ultrasonically cleaned in acetone for 2 min then plasma cleaned using a PDC-001 Expanded Plasma Cleaner (Harrick Plasma) at low power for 2 min. Each nanopaste (NPA) was applied to the Ni718 and most of the water was evaporated by baking on a hot plate at 75 °C. Silver brazing flux was applied on top of the nanopaste (601 silver brazing flux was used
4
for laser power <250 W and 601B flux was used for laser power ≥250 W). The two NI718 plates were placed 9 cm from a 400 W continuous wave high energy diode laser (λ = 806 nm) on a 1 mm thick copper plate with an overlap of 34 mm as shown in Fig. 1. The laser intensity profile at the focal point is similar to a top-hat intensity profile with a size of 1 mm x 22 mm. The laser irradiated the top surface of the Ni718 plate. The laser power was ramped to the target power in ~5 seconds then held at the target power for ~10 seconds/sample as seen in Fig. 1b. No pressure was applied to the sample. A lap shear test was performed using a tensile testing machine on the brazed samples.
2.6 Characterization Scanning Electron Microscopy (SEM) images were collected on a Zeiss Auriga Scanning Electron Microscope. X-ray diffraction measurements were conducted on a Panalytical Empyrean X-ray Diffractometer, Energy Dispersive X-ray Spectroscopy (EDS) measurements were performed on samples brazed with 300 W laser power using a Phenom ProX energy-dispersive spectrometer. Vickers microhardness of 300 W brazed samples was measured using a Buehler MMT3 Microhardness Tester with 5 gram-force. The hardness of Ni718 prior to brazing was measured from a cross-sectional piece of Ni718. For cross-sectional analysis, samples were embedded in an epoxy resin and polished using silica 0.04 µm paste.
3. Results and discussion 3.1 Characterization of nanopaste Fig. 2a and 2b show the bare Cu NWs and the Cu-Ag CSNWs respectively. Cu NWs have a smooth surface with a rectangular cross section and a 400-1000 nm diameter and 5-40 µm long. After the addition of AgNO3, Ag nanocrystals nucleated and grew on the surface of the Cu NW giving it a rough surface morphology. The CSNWs became hollow with a wall thickness of 40-80 nm and a pentagonal crosssection as seen in Fig. 2c. The Ag NWs have a smooth surface texture with a diameter of 40-50 nm and a length of 5-30 µm (Fig. 2d). The Ag NPs are spherical and 55-75 nm in diameter (Fig. 2e). As seen in 2f, the UV-vis spectrum of Cu NWs show that the surface plasmon resonance (SPR) peak at 582 nm with high absorbance across the spectrum of 200 nm to 960 nm; Wang and Ruan (2016) report a similar spectrum. After Ag coating, the SPR peak red shifts into the near-infrared/infrared range and the absorbance in the visible light range is greatly diminished. The Cu-Ag CSNWs reported in Wei et al. (2015) study exhibit similar behavior. The broad SPR peak of Ag NPs centered at 432 nm indicates a wide particle size distribution. As expected, Ag NPs absorb light in the red and near-infrared spectrum (550-1000 nm) is absorbed very weakly like in the report by Krishna Podagatlapalli et al. (2013). Like Cu NWs, Ag NWs have a high absorbance across the spectrum except its SPR peak is located at 378 nm and is consistent with the UV-vis spectrum in Tang and Tsuji (2010). XRD patterns for Cu NWs (Fig. 3a) only show XRD peaks characteristic to copper and no copper oxides were identified. After the Ag coating, Cu peaks do not disappear, but the peak intensity ratio for the Cu (111) and Cu (200) planes decrease from 3.44:1 to 2.85:1 (Fig. 3b). The change in relative peak intensity between Cu (111) and Cu (200) is likely caused by the Ag covering the Cu (111) plane, causing a decrease
5
in Cu (111) intensity. The ratio between the Ag (111) and Ag (200) peak intensity is 7.02:1. It is worth noting that the CSNW XRD pattern shows distinct Cu peaks and distinct Ag peaks, and no detection of copper oxides. During the centrifugation step to remove excess reducing agent, some of the organic capping layer (PVP) is also removed by washing. If the silver coating is not sufficient, the Cu NWs will easily develop an oxide layer on the surface of the NW in an aqueous environment (Kim et al., 2014). This indicates that CSNWs adopt a core-shell structure. The Ag NWs (Fig.3c) and Ag NPs (Fig. 3d) have similar peak positions but the relative intensities for the (220) and (311) peaks are much stronger for Ag NPs than Ag NWs.
3.2 Brazing joint microstructure COMSOL Multiphysics Suite was used to estimate the temperature and temperature distribution of the laser brazing samples. For simplicity, the nanopaste material is treated as dry powder with a thermal conductivity similar to a solid material. Due to the top-hat profile of the laser beam, the laser intensity is considered constant throughout the laser heated region. At the end of the brazing process at 300 W, the temperature in the laser heated region is nearly 1000 °C as seen in Fig. 4. A high temperature thermocouple was used to measure the temperature of the top plate and validate the simulation results. The average temperature in the brazing material increases to approximately 775 °C during the brazing as seen in Fig. 5b. There was no significant temperature difference between the pure Ag paste and the CuAg paste. According Fig. 5c, there is a temperature difference of ~300 °C between the top plate and the bottom plate. On the surface of the sample, the temperature varies in the overlap area between 750975 °C (Fig. 5d). Fig. 5e and 5f show that the temperature gradient across the NPA layer is extremely low with temperatures ranging between 760 °C and 800 °C. This result suggests that the nanopaste is heated evenly during the laser brazing procedure. For nanojoining, temperature gradient and heating speed are key factors as summarized by German (1996) in his sintering fundamentals textbook. A steep temperature gradient promotes dissolution of the base metal into the brazing material via thermal diffusion. The high heating speed of laser brazing helps achieve a large thermal gradient like the one demonstrated in this simulation. Even though, the thermal gradient is low within the nanopaste layer on a macroscopic scale, there is most likely the size dependent melting and liquid phase sintering drive the inter-particle joining. Surface diffusion is the dominant mass transport mechanism during surface melting of nanomaterials (at low temperature). However, due to the heating speed of the laser brazing technique, the surface melting of nanoparticles can induce the full melting of the whole nanoparticles and then a liquid phase sintering. According to Wang et al. (2016), the atoms at the Cu core of the nanostructure begin to melt at around 800-950 °C. Once the filler material is melted, the atoms easily diffuse into the base material or participate in joining of nearby particles.
After brazing at 300 W, the Ag NWs (Fig. 6a-b) and CSNWs (Fig. 6c-d) form a highly densified brazing joint and the nanowire structure has completely disappeared. This convinces a melting predicted in our simulation. High magnification images of the brazed joint reveal the presence of submicron voids in the brazed structure. The porosity of the Cu-Ag and Ag joint is much lower compared to other pressureless
6
low temperature joining techniques such as the technique used by Li et al. (2015d) to sinter a Ag nanowire-nanoplate composite paste for low temperature joining. All the EDS line scans were performed on a sample that was joined at 300 W. The line scan in Fig. 7b reveal uniform dissolution of the base material into the Ag brazing material and dissolution of Ag into the base material. The line scan reveals that there is a 3 µm diffusion layer between the base material and the brazing material. The light spots in Fig. 7c reveal the existence of secondary phases near the interface between the base material and brazing material. The presence of the diffusion layer and dissolution of the base material into the braze indicates that metallurgical bonding has occurred at the interface. Fig. 8 shows a brazing joint using the CSNWs. The EDS line scan for this joint shows that the dark colored regions are extremely Cu-rich (Fig. 8d). These Cu-rich regions contain on average 6 times more copper than silver. The light regions have a Cu:Ag ratio of 3:4. A higher resolution line scan (Fig. 8e-f) shows that there is a smooth transition between the Cu-rich phases and the Ag-rich phases with a diffusion layer of approximately 1.6 µm. The precipitation Cu-rich phases in a Ag-rich matrix is not observed in the vacuum brazing experiments conducted by Ma et al. (2016). One reason for such a phase separation is the miscibility gap in the Cu-Ag system. Another possible reason for this phase separation is the high heating and cooling speed of our laser brazing technique prevents homogenization of the brazed joint. The existence of Cu-rich phases and Ag-rich phases in the brazed joint are consistent with molecular dynamics simulations of sintering procedures of two Cu-Ag core-shell particles. Wang et al. (2016) displayed that at 727 °C the Ag shell will melt first before the Cu core melts (800-950 °C). The simulation also reveals that the Cu-core remains intact until the Ag shell has completely melted. Once the entire nanoparticle has melted, the core-shell structure disappears and alloying occurs. This is result is consistent with the microscale phase separation observed in Cu-Ag CSNW brazing. A similar amount of the Ni718 is dispersed throughout the brazing joint. However, the amount of brazing material dispersed in the base material is approximately 33% more when using CSNWs compared to using Ag NWs as a brazing material. This brazing joint has a diffusion layer of 2 µm which is 33% smaller than the diffusion layer using Ag NWs. XRD was performed on the Cu-Ag materials after brazing at 300 W (Fig. 9). The relative intensity of the Cu* peaks are significantly higher in the BAg-8 brazed joints than the Cu* peaks in the Cu-1 and Cu-3 brazed joints. However, the relative intensities in the BAg-8 joint are very similar to the Cu-2 joint except for the Cu* (220) peak (65.3°). Compared to the BAg-8 alloy, most of the major peaks from the NPAbrazed joints were broader than those of the brazing alloy. Broader XRD peaks suggest that the final grain size is smaller for NPA brazed joint than BAg-8 brazed joints. The significance of this detail will be discussed later.
3.3 Mechanical Properties 3.3.1 Hardness
The hardness profile of the Ag and Cu-Ag brazed joint is plotted in Fig. 10. The Ag brazing material (Fig. 10a) has an average Vickers hardness of 71 HV within the brazing region. The plate that was directly irradiated by the laser is noted in the figure. The hardness of the interface between the brazing material and the base material is slightly lower on the laser irradiated side than the hardness of Ni718 prior to
7
brazing (indicated by the red dotted line). The hardness of the base material returns to the original hardness of Ni718 after a brief increase in hardness. The hardness of interface between the non-laser irradiated Ni718 and the Ag braze is greater than the hardness at the laser-irradiated interface. Hardness values for 100 μm outside the brazing region return to the original hardness value for Ni718 on both sides of the brazing region. The Cu-Ag brazing material is on average almost twice as hard as the Ag joint (Fig. 10b). The hardness of the Ni718, especially at the interface is much more greatly affected by the Cu-Ag CSNWs as well. This change in hardness is likely caused more significant compositional change when using Cu-1 compared to Ag-2 (as seen in Fig. 7b and 8b). Like the Ag joint in Fig. 10a, the laser irradiated interface has a lower hardness than the non-laser irradiated interface. The only difference is how the non-laser irradiated plate is affected. On laser irradiated side, the hardness values regress to the original hardness of Ni718 100 μm outside the brazing region. On the non-laser side, the hardness values regress to the original hardness of Ni718 200 μm outside the brazing region. On both sides of the brazing region the hardness is larger than the original Ni718 value before returning to normal values. 3.3.2 Bonding strength
Fig. 11 presents the bonding strength of each NPA as a function of laser power. High bonding strength (>100 MPa) was obtained for all NPAs except the Ag-1. Both silver nanomaterials have a laser power threshold of 250 W, however the Ag-2 NPA has superior bonding strength (116.3 MPa) compared to Ag-1 (49.1 MPa). The pure Cu-1 paste exhibits a steady increase in bonding strength as a function of laser power until it reaches a 224 MPa bonding strength at 300 W. The Cu-2 NPA has the highest bonding strength with a maximum bonding strength of 249.6 MPa. The increase in strength compared to the pure CSNWs is because the Ag NPs increase the green density of the paste prior to brazing. Guo et al. (2017) demonstrated that mixing NPs into a NW paste increases the green density for sintering by using a mixed Ag NW and Cu NP or Ag NP paste. The large increase in bonding strength of the Cu-2 paste displays a between 200 W and 250 W corresponds with minimum laser power for joining of Ag NPs, suggesting that the NPs melting in the paste is a key occurrence that strengthens the Cu-Ag joint. However, the effect of adding Ag NPs is limited because the bonding strength of the Cu-2 paste is greater than the pure CSNWs paste only at 250 W and 300 W laser power and only by a small margin. Mixing Ag NWs into the CSNWs (i.e. Cu-3) effectively lowers the laser power required for successful joining (150 W). Cu-3 has a lower threshold laser power compared to Ag-1, Ag-2, and Cu-1 due to the Cu:Ag ratio shifting to the eutectic composition. Even in a bulk material, the eutectic temperature of the Cu-Ag system is 183.7 °C lower than the melting point of pure silver and 61.9 °C lower than the liquidus temperature of bulk Cu-1 (i.e., 80 at% Cu, 20 at% Ag). The possible reason for why Cu-2 does not have the same bonding threshold as Cu-3 is that Ag NWs have been shown in numerous studies to have a lower minimum bonding temperature than Ag NPs via the shape effect. The shape effect, as the name suggests is the effect of shape on the melting temperature of a nanomaterial. presents a model via the Lindemann melting criterion that NWs have a lower specific heat than NPs. Stacey and Irvine (1977) provide a thermodynamic basis for the Lindemann melting criterion which they mention is a theory that explains the melting mechanism of a material using the vibration of atoms in a crystal. Li et al. (2015b) demonstrated that the bonding strength of Ag NWs becomes constant when a certain bonding temperature even when mixed with a differently shaped nanomaterial which is possibly why the laser power has very little effect on the bonding strength of the Cu-3 paste (~160 MPa).
8
The Cu-Ag NPAs have higher bonding strength than the BAg-8 primarily due to Hall-Petch strengthening. According to the Hall-Petch relation, the yield strength of a material increases as the grain size decreases. Ebrahimi et al. (1999) reports an increase in yield strength in electrodeposited nanocrystalline nickel when the grain size decreases. Using the Scherrer formula ( τ= 0.9λ/(β cosθ) ), the grain sizes were calculated based on the XRD peaks in Fig. 9. τ is the mean grain size, λ is the x-ray wavelength, β is the full-width half maximum, and θ is the Bragg angle. The results of these calculations are presented in Table 3. For the NPAs, the calculated grain sizes are smaller than BAg-8 on average except for a couple of exceptions. The Cu-1 joint contains Ag* (111) and Cu* (220) oriented grains (38.7° and 65.3°, respectively) that are roughly the same size as the Ag* (111) and Cu* (220) oriented grains in the BAg-8 joint. Based on calculations of the Ag* (311) peak (located at 74.1°), some of the Ag-rich grains in the Cu3 joint are larger than those found in the BAg-8 joint. The Cu-2 joint contains larger Cu-rich grains than the BAg-8 joint based on the Cu* (111) and Cu* (311) peaks (43.5° and 78.4°, respectively). Overall, the NPA brazed joints have much smaller grain sizes than BAg-8 brazed joints. Therefore, the superior bonding strength of the Cu-Ag NPAs can be heavily attributed to Hall-Petch strengthening of the brazing material. There are a few possible explanations for this: 1. The NPAs contain small amounts of organic compounds leftover after the washing procedure. Even though the organic compounds in question decomposed before solidification, the decomposition byproducts still remain and provide a source of defects within the brazing material. These defects may provide additional nucleation sites for the molten brazing material during solidification. BAg-8 does not contain any organic compounds within the material so these impurity-based defects would not exist in a brazed joint made by BAg-8. 2. The porosity of the NPAs may also have an effect on the nucleation and growth behavior of the NPAs. The initial porosity may have an effect on the interdiffusion behavior of Cu, Ag, and the base material. In the early stages of brazing, before the material fully melts and spreads, interdiffusion only occurs in the areas that the NMs are in direct contact with the base material. This would create local variations in the chemical composition on the Ni718 surface and thus change the nucleation and growth behavior of the base material. A more in-depth investigation of the melting and solidification behavior of NMs as a brazing filler metal is underway. The higher yield strength was confirmed by the stress-strain curves in Fig. 12 The Cu-3 brazed joint has an offset yield strength of 80 MPa and the BAg-8 brazed joint has an offset yield strength of 50 MPa.
9
According to the stress strain curve the BAg-8experiences more necking during fracture than the CSNWs, implying a more ductile failure mechanism in the BAg-8 joint. The fracture surfaces shown in Fig. 13 indicate that fracture occurs within the paste, not at the interface between the braze material and the base material. The Ag joints (Fig. 13b-c) have dimple fracture surface, indicating strong bonding. The dimples from the Ag-2 paste are much larger than the dimples in the Ag NP fracture surface. The Cu-1 joint (Fig. 13d) has a dimple-cleavage fracture surface. The Cu-2 paste fracture surface (Fig. 13e) is very similar to the fracture surface the Ag-1 paste (Fig. 13b). The similarities in the Cu-2 paste and Ag-1 paste fracture surfaces further supports the conclusion that the addition of Ag NPs greatly influences the bonding behavior of Cu-Ag CSNWs. The main difference is that the fracture surface of the Cu-2 paste has larger, deeper dimples than the pure Ag-1 paste structure. The Cu-3 paste fracture surface (Fig. 13f) primarily resembles the fracture surface of the CSNWs, however the dimple features become more apparent after Ag NWs are added to the CSNWs paste. The emphasized dimple features in the Cu-3 paste fracture surface demonstrate that the Ag NWs also have a dominant effect on the bonding/fracture of the Cu-Ag joint.
4. Conclusions Cu-Ag and Ag nanomaterials exhibit high bonding strength and higher joint density than previous nanojoining studies. Elemental analysis reveals dissolution of the base material throughout the brazing material. For Cu-2 and Cu-3 nanopastes, the Ag NPs and Ag NWs play an important role in the bonding performance and fracture mechanism of joints fabricated by mixed Cu-Ag NPAs. The superior bonding strength of Cu-Ag NPAs compared to the BAg-8 filler metal is attributed to Hall-Petch strengthening.
Acknowledgements This project is jointly supported by Institute of Public Service, University of Tennessee through I’UCRC project of National Science Foundation and a seed grant from Oak Ridge National Laboratory, Department of Energy, USA. Chris Wetteland and Maneel Bharadwaj assisted in the collection of EDS and microhardness measurements. John Dunlap, Maulik Patel, and the Joint Institute of Advanced Materials permitted use of their electron microscopy facilities and X-ray diffraction equipment provided the training needed to operate those devices.
References Chen, J., Chen, J., Li, Y., Zhou, W., Feng, X., Huang, Q., Zheng, J., Liu, R., Ma, Y., Huang, W., 2015. Enhanced oxidation-resistant Cu-Ni core-shell nanowires: controllable one-pot synthesis and solution processing to transparent flexible heaters. Nanoscale 7, 16874-16879. Ebrahimi, F., Bourne, G.R., Kelly, M.S., Matthews, T.E., 1999. Mechanical properties of nanocrystalline nickel produced by electrodeposition. Nanostruct Mater 11, 343-350.
10
Fu, S., Mei, Y., Lu, G.-Q., Li, X., Chen, G., Chen, X., 2014. Pressureless sintering of nanosilver paste at low temperature to join large area (≥100mm2) power chips for electronic packaging. Materials Letters 128, 42-45. German, R.M., 1996. Sintering Theory and Practice. John Wiley & Sons, Inc., New York, NY. Grouchko, M., Kamyshny, A., Magdassi, S., 2009. Formation of air-stable copper–silver core–shell nanoparticles for inkjet printing. Journal of Materials Chemistry 19, 3057. Guo, W., Zhang, H.Q., Zhang, X.Y., Liu, L., Peng, P., Zou, G.S., Zhou, Y.N., 2017. Preparation of nanoparticle and nanowire mixed pastes and their low temperature sintering. Journal of Alloys and Compounds 690, 86-94. Hausner, S., Weis, S., Wagner, G., 2016. Joining of steels at low temperatures by Ni nanoparticles, 11th International Conference on Brazing, High Temperature Brazing and Diffusion Bonding. Dusseldorf: DVS Media GmbH, Aachen, Germany, pp. 278-284. Hokita, Y., Kanzaki, M., Sugiyama, T., Arakawa, R., Kawasaki, H., 2015. HighConcentration Synthesis of Sub-10-nm Copper Nanoparticles for Application to Conductive Nanoinks. ACS Appl Mater Interfaces 7, 19382-19389. Huang, X., 2014. Brazing of CMSX-4 with a Boron- and Silicon-Free Ni-Co-Zr-Hf-Cr-TiAl Brazing Alloy. Weld J 93, 233s-242s. Kahler, J., Heuck, N., Wagner, A., Stranz, A., Peiner, E., Waag, A., 2012. Sintering of Copper Particles for Die Attach. IEEE Transactions on Components, Packaging and Manufacturing Technology 2, 1587-1591. Khorram, A., Ghoreishi, M., 2015. Comparative study on laser brazing and furnace brazing of Inconel 718 alloys with silver based filler metal. Optics & Laser Technology 68, 165-174. Kim, C.K., Lee, G.-J., Lee, M.K., Rhee, C.K., 2014. A novel method to prepare Cu@Ag core–shell nanoparticles for printed flexible electronics. Powder Technology 263, 1-6. Kim Seah, T., Kuan Yew, C., 2014. Physical and Electrical Characteristics of SilverCopper Nanopaste as Alternative Die-Attach. IEEE Transactions on Components, Packaging and Manufacturing Technology 4, 8-15. Krishna Podagatlapalli, G., Hamad, S., Tewari, S.P., Sreedhar, S., Prasad, M.D., Venugopal Rao, S., 2013. Silver nano-entities through ultrafast double ablation in aqueous media for surface enhanced Raman scattering and photonics applications. Journal of Applied Physics 113, 073106. Lee, C., Kim, N.R., Koo, J., Lee, Y.J., Lee, H.M., 2015. Cu-Ag core-shell nanoparticles with enhanced oxidation stability for printed electronics. Nanotechnology 26. Li, J., Johnson, C.M., Buttay, C., Sabbah, W., Azzopardi, S., 2015a. Bonding strength of multiple SiC die attachment prepared by sintering of Ag nanoparticles. Journal of Materials Processing Technology 215, 299-308. Li, R.-Z., Zhang, T., Hu, A., Bridges, D., 2015b. Ag Nanowire and Nanoplate Composite Paste for Low Temperature Bonding. Materials Transactions 56, 984-987. Li, R.Z., Hu, A., Zhang, T., Oakes, K.D., 2014. Direct writing on paper of foldable capacitive touch pads with silver nanowire inks. ACS Appl Mater Interfaces 6, 2172121729. Li, R.Z., Hu, A.M., Bridges, D., Zhang, T., Oakes, K.D., Peng, R., Tumuluri, U., Wu, Z.L., Feng, Z.L., 2015c. Robust Ag nanoplate ink for flexible electronics packaging. Nanoscale 7, 7368-7377.
11
Li, R.Z., Zhang, T., Hu, A.M., Bridges, D., 2015d. Ag Nanowire and Nanoplate Composite Paste for Low Temperature Bonding. Materials Transactions 56, 984-987. Ma, C., Xue, S., Bridges, D., Palmer, Z., Feng, Z., Hu, A., 2016. Low Temperature Brazing Nickel with Ag Nanoparticle and Cu-Ag Core-Shell Nanowire Nanopastes. submitted to Journal of Materials Science. Manikam, V.R., Razak, K.A., Cheong, K.Y., 2013. Reliability of sintered Ag80–Al20 die attach nanopaste for high temperature applications on SiC power devices. Microelectronics Reliability 53, 473-480. Peng, P., Hu, A.M., Gerlich, A.P., Zou, G.S., Liu, L., Zhou, Y.N., 2015. Joining of Silver Nanomaterials at Low Temperatures: Processes, Properties, and Applications. Acs Appl Mater Inter 7, 12597-12618. Pouranvari, M., Ekrami, A., Kokabi, A.H., 2014. Diffusion brazing of cast INCONEL 718 superalloy utilising standard heat treatment cycle. Mater Sci Tech-Lond 30, 109-115. Shaowei, L., Zhishui, Y., Peilei, Z., Xiaoliang, L., Dingding, Q., Xiaopeng, L., 2016. Numerical Simulation and Parameters Optimization of Laser Brazing of Galvanized Steel. ISIJ International 56, 637-646. Soldering, A.C.o.B.a., 2002. Brazing Handbook, 4 ed. American Welding Society, Miami, Florida, United States of America. Stacey, F.D., Irvine, R.D., 1977. Theory of Melting: Thermodynamic Basis of Lindemann's Law. Australian Journal of Physics 30, 631. Tang, X.L., Tsuji, M., 2010. Syntheses of Silver Nanowires in Liquid Phase. Nanowires Science and Technology, 25-42. Tsai, C.H., Chen, S.Y., Song, J.M., Chen, I.G., Lee, H.Y., 2013. Thermal stability of Cu@Ag core-shell nanoparticles. Corros Sci 74, 123-129. Wang, J.Q., Shin, S.H., Hu, A.M., 2016. Geometrical Effects on Sintering Dynamics of Cu-Ag Core-Shell Nanoparticles. J Phys Chem C 120, 17791-17800. Wang, R.L., Ruan, H.B., 2016. Synthesis of copper nanowires and its application to flexible transparent electrode. Journal of Alloys and Compounds 656, 936-943. Wei, Y., Chen, S., Lin, Y., Yang, Z.M., Liu, L., 2015. Cu-Ag core-shell nanowires for electronic skin with a petal molded microstructure. Journal of Materials Chemistry C 3, 9594-9602. Xiong, H.-P., Chen, B., Pan, Y., Zhao, H.-S., Ye, L., 2014. Joining of Cf/SiC composite with a Cu–Au–Pd–V brazing filler and interfacial reactions. Journal of the European Ceramic Society 34, 1481-1486. Ye, X., Hua, X.M., Wu, Y.X., Lou, S.N., 2015. Precipitates in coarse-grained heataffected zone of Ni-based 718 superalloy produced by tungsten inert gas welding. Journal of Materials Processing Technology 217, 13-20. Zhang, X., Zhang, D., Ni, X., Zheng, H., 2006. One-step preparation of copper nanorods with rectangular cross sections. Solid State Commun 139, 412-414. Zhao, J., Zhang, D., Zhang, X., 2015. Preparation and characterization of copper/silver bimetallic nanowires with core-shell structure. Surface and Interface Analysis 47, 529534.
13
14
15
16
17
18
19
20
21
22
Table 1: Composition of the alloys and fluxes used in this study Material Ni718 BAg-8
Ni 53.10 Cu 28
Fe 18.40 Ag 72
Cr 18.30
Cu 0.05
Chemical composition Al Si C Mn 0.49 0.07 0.05 0.24
B 0.004
Nb 4.95
Ti 1.07
S <0.002
Mo 3.06
23
601/601B Flux
KBF4 20-40
K2B4O7 15-25
H3BO3 30-40
H2B2FO2 7-15
Table 2: Description of the nanopastes used in this laser brazing study
Paste Ag-1 Ag-2 Cu-1 Cu-2 Cu-3
Cu (wt%) 0 0 80 28 28
Ag (wt%) 100 100 20 72 72
Description Silver nanoparticles paste Silver nanowire paste Copper-silver CSNWs paste CSNWs and silver nanoparticles composite paste CSNWs and silver nanowires composite paste
Table 3: Grain size (in nm) after brazing of Cu -Ag brazing materials based on XRD results
Material BAg-8 Cu-1 Cu-2 Cu-3
38.7 54.7 54.8 41.2 36.5
43.5 67.7 41.8 66.8 41.8
45 48.0 42.0 33.6 33.7
2θ (Degrees) 50.5 28.6 34.4 21.5 24.5
65.3 46.1 46.1 36.9 15.3
74.1 48.6 24.3 17.7 48.8
78.4 33.4 25.0 40.1 25.0
24