Materials and Design 32 (2011) 2685–2694

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

A new approach to joining of bulk copper using microwave energy

M.S. Srinath, Apurbba Kumar Sharma ⇑, Pradeep KumarDepartment of Mechanical and Industrial Engineering, Indian Institute of Technology Roorkee, Roorkee 247 667, India

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 October 2010Accepted 11 January 2011Available online 18 January 2011

Keywords:C. JoiningF. MicrostructureG. X-ray analysis

0261-3069/$ - see front matter � 2011 Elsevier Ltd Adoi:10.1016/j.matdes.2011.01.023

⇑ Corresponding author. Tel.: +91 1332 285421; faxE-mail addresses: srinadme@gmail.com (M.S. Sri

(A.K. Sharma), kumarfme@iitr.ernet.in (P. Kumar).

Metallurgical joining of high thermal conductivity materials like copper has been technically challenging.This paper illustrates a novel method for joining of bulk metallic materials through microwave heating.Joining of copper in bulk form has been carried out using microwave energy in a multimode applicator at2.45 GHz and 900 W. Charcoal was used as susceptor material to facilitate microwave hybrid heating(MHH). Copper in coin and plate forms have been successfully joined through microwave heating within900 s of exposure time. A sandwich layer of copper powder with approximately 0.5 mm thickness wasintroduced between the two candidate surfaces. Near complete melting of the powder particles in thesandwich layer does take place during the microwave exposure leading to metallurgical bonding ofthe bulk surfaces. Characterisation of the joints has been carried out through microstructure study, ele-mental analysis, phase analysis, microhardness survey, porosity measurement and tensile strength test-ing. X-ray diffraction (XRD) pattern indicates that some copper powder particles got transformed intocopper oxides. XRD analysis also reveals that the dominant orientation (3 1 1) in starting copper powdergot transformed into a preferential orientation (1 1 1) in the joint. A dense uniform microstructure withgood metallurgical bonds between the sandwich layer and the interface was obtained. The hardness ofthe joint area was observed to be 78 ± 7 Hv, while the porosity in the joint was observed to be 1.92%.Strength character of the copper joints shows approximately 29.21% elongation with an average ultimatetensile strength of 164.4 MPa.

� 2011 Elsevier Ltd All rights reserved.

1. Introduction

Permanent joining of materials has been one of the primerequirements in most of the manufacturing and assembling indus-tries. The existing techniques like welding, soldering and brazingare being widely practiced in industries; however, they have theirown limitations regarding processing time, materials to be joinedand characteristics of the joint. Further, ease of processing andenvironmental hazards, are some of the issues that need to be ad-dressed. Thus a more versatile, faster and cleaner process couldhave a huge impact on production. Investigations reveal that appli-cation of microwave energy as a tool in materials processing is notonly a green manufacturing process, but also significantly faster atrelatively low investment. Microwave materials processing cangive an alternative to high energy consumption heating techniquesthat are commonly used in industries.

In microwave processing, energy is directly transferred to thematerial through interaction of electromagnetic waves with mole-cules leading to volumetric heating. Heat is generated internallywithin the material, instead of originating from the externalsources, and gets transmitted outward. Hence, there is an inverse

ll rights reserved.

: +91 1332 285665.nath), akshafme@iitr.ernet.in

heating profile, ‘inside-out’ unlike in a conventional heating ‘out-side-in’. Several authors have shown the use of microwave energyfor wide ranging applications. In one of the premier investigations,Osepchuk has explained the basics of microwave heating and pre-sented a brief history of the applications of microwave energy [1].Later, the same author further explored the possible areas of appli-cations of microwave power in details [2]. Significance of micro-wave heating and its applications in processing of ceramics werethen analyzed by Sutton in a landmark publication in 1989 [3].The unique features of processing materials with microwave werelucidly presented. Later, Clarke et al. have shown the potentials andchallenges of using microwave energy in materials processing [4].Subsequently, application of microwave energy in material pro-cessing was reported in many areas including the new and unusualapplication like glazing of sprayed ceramic composite surfaces [5].Microwave energy has been effectively used in the processing ofdifferent materials. However, majority of these applications waslimited to processing of microwave absorbing materials (mostly,bio-materials, hydrocarbons etc.), ceramics and ceramic compos-ites. Successful sintering of alumina with nearly full density at1350 �C after 50 min has been achieved using 2.45 GHz microwaveand its comparison with conventional heating shows only 62% den-sity at this temperature [6]. Metals and alloys, on the other hand,remained outside the applications envelop of microwave energyin materials processing for an uncharacteristically longer time.

2686 M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694

Recent research activities, however, indicate that it is possibleto process metals under certain conditions. Many commercialpowder metal components and their alloys have been sinteredusing microwaves. It has been reported that the powders with acomposition of iron (Fe), copper (2%) and graphite (0:8%) havebeen sintered in a microwave field at 1200 �C for 30 min withexcellent density [7]. The same authors also reported microwavesintering of cobalt metal powder in pure H2 environment withone atmospheric pressure at various temperatures-ranging from900 to 1200 �C for 10 min. The densities reported were 8700 kg/m3 at 900 �C to 8880 kg/m3 at 1000–1050 �C and near theoreticaldensity of 8890 kg/m3 at 1100–1200 �C. Rodiger et al. had carriedout sintering of hardmetals through 2.45 GHz microwave heatingand reported sintering temperature of 1300 �C in microwave pro-cess was achieved in �1.5 h with 1 kW power whereas conven-tional process took nearly 5 h to reach 1400 �C with 4.5 kW.Platelets microstructure embedded in a fine-grained hardmetalmatrix with an average size of 0.6 lm was obtained [8]. Prabhuet al. had examined the comparative sinterability of as receivedpowder and activated tungsten powder in microwave. It was ob-served that the activated tungsten powder shows better densifica-tion because of reduced particle size and higher specific surfaceenergy [9]. Mondal et al. have exposed different particle sized elec-trically conductive material like copper having varied initial poros-ity. The reported results indicate that the smaller the particle sizewith higher porosity, the higher will be the microwave absorptionrate and hence heating is rapid [10]. Gupta and Wong reported thetwo-directional microwave assisted rapid sintering of aluminum,magnesium and lead free solder. The results revealed that the den-sity of the microwave sintered and conventionally sintered sam-ples are same whereas the marginal increase in microhardnesswith superior ultimate tensile strength of the microwave sinteredaluminum and magnesium [11].

Microwaves have been efficiently employed for joining of cera-mic materials. It has been reported that sintered alumina and 30%zirconia ceramic composites were successfully joined by micro-wave hybrid heating at 2.45 GHz frequency and power 700 W[12]. The joints were fabricated with and without sodium silicateglass powder as an interlayer. The flexural strength of such joints(with interlayer) was reported in the range of 28 MPa. Microwavejoining of 48%alumina–32%zirconia–20%silica ceramics throughsuitable temperature control has also been reported which yieldedjoint strength in excess of about 107% of the base material [13].

However, application of microwaves in joining of metals is chal-lenging owing to reflection of microwaves by most of the bulk met-als at ordinary conditions. Further, works have been reported onbrazing of selected metals under specific conditions. Bartmatzet al. in the year 2000 have reported in the form of patent on braz-ing of titanium carbide tip to diamond cutter to enhance the prop-erties of the cutter [14]. Braze powder was used as interface layerwith microwave temperature upto 1000 �C. In continuation of theprevious study, Sallom et al. have reported the brazing of GammaTiAl with Ag-based filler metal by microwave heating between925 �C to 1050 �C in 5 min with 1 MPa load [15]. Budinger have re-ported brazing of nickel based superalloys with nickel basedmetallic powders in a multimode microwave cavity [16]. Particlesize used in this work was about 44 lm. Results show the finerparticle attained maximum temperature of 1140 �C, whereas coar-ser particles were heated upto 827 �C. This evidence shows thatmicrowaves have greater susceptibility towards finer particle size.

Successful joining of thin steel sheet in the thickness range of0.1–0.3 mm using microwave was reported by Siores and Rego[17]. The authors showed that the localized arcing was enough tomelt such thin test sheets by using a 2 kW multimode magnetron.Agrawal, on the other hand, has reported joining of regular steeland cast iron in a microwave field within 2–3 min using a braze

powder [18]. However, no details on characterisation of such jointswere reported. However, joining of bulk metallic materials usingmicrowaves has hardly been reported. Sharma et al. have success-fully carried out joining of bulk metallic pieces with fusion of theparent metals in a low cost home microwave system [19]. Joiningof pure metals (copper), similar alloys (stainless steel to stainlesssteel), and dissimilar alloys (stainless steel to mild steel) in thebulk form have been reported. Further, the work has been ex-tended for cladding/coating of metallic powder on bulk metallicmaterials in the same laboratory; the work has been filed for pat-ent [20].

Applications of copper in industries and researches have beenenormous. Joining of bulk copper through conventional joining(welding) techniques, on the other hand, is difficult owing to itshigh heat conduction capability. The present paper reports on theresults of an on-going project on joining and characterisation ofbulk copper using microwave energy in a multimode applicator.

2. Experimentation

Joining of bulk metallic materials using microwaves is difficult.Most of the metals do not allow microwaves to penetrate insidethe bulk at room temperature owing to the presence of electron-cloud. Heating in microwave processing is, on the other hand,due to microwave-material interaction at the molecular level. Inorder to overcome such processing challenges, a number of trialshave been carried out. The following sections illustrate experimen-tal procedures adopted for fabrication of the joints and differentcharacterisation techniques employed to investigate the jointproperties.

2.1. Material selection

Copper is one of the most widely used metals in the manufac-turing of electrical gadgets and components for aerospace indus-tries. Moreover, copper poses significant processing difficulties inconventional welding practices. Localized heating and consequentmelting of copper in conventional welding is complicated due to itshigh thermal conductivity. Thus, copper was chosen as thecandidate material for joining through microwave heating. Trialswere carried out with commercial grade copper (purity � 99.9%)plates and coins having dimensions 15 mm � 12 mm � 4 mm andø18 mm � 12 mm respectively. Commercially available copperpowder with 99.5% purity and average particle size of �5 lmwas used as a sandwich layer. Morphology of the powders usedin the trials is shown in Fig. 1. Typical XRD spectrum of the copperpowders is illustrated in Fig. 2. Three major peaks of copper with adominating (3 1 1) structure are seen. An epoxy resin (Bisphenol-A,Blumer 1450XX) was mixed with the metal powder to make slurryso as to avoid collapsing of the sandwich layer. This low meltingpoint epoxy comes out of the sandwich matrix much early duringheating. The detailed specifications of the material are as shown inTable 1.

2.2. Development of joints

Interface surfaces of the bulk copper pieces were cleaned ultra-sonically in an acetone bath prior to applying the sandwich layer(slurry). The prepared slurry was uniformly placed over the candi-date surfaces between the two bulk pieces maintaining an averagethickness of five hundred micrometers. Experiments were carriedout in atmospheric condition. It was observed that, microwavecoupling has been significantly improved by the use of epoxy resin.However, as the temperature rises, the resin gets burnt and thusfails to sustain the initial coupling with microwaves.

Fig. 1. Morphology of copper powder used in sandwich layer.

Fig. 2. Typical XRD Spectrum of the copper powder.

Table 1Details of the material used.

A – Base material (copper)Purity 99.9%Micro hardness 70 ± 5 HvUltimate tensile strength 284 MPa

B – Interface powder (copper)Purity 99.5%Particle size �5 lm

C – Epoxy resinSolvent Ketones, Esters, GlycolEthersType Bisphenol-A based

M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694 2687

The depth of microwave penetration into the material (alsocalled ‘skin depth’) is much lesser than the size of the bulk copperpieces used in the present work. This results in reflection of micro-waves from the target material. In order to avoid this problem,appropriate masks were prepared to seal the bulk metallic bodyusing refractory brick material. A solid layer of graphite was used

to separate the charcoal powders and the sandwich powder layer.Principles of microwave hybrid heating were effectively used tojoin the metallic pieces. Charcoal powder was used as a susceptormedium to initiate coupling of microwaves which results in initialheating. The charcoal powders were placed as near as possible tothe joint area so as to induce selective heating in the sandwichzone. The metallic powders get heated through conventionalmodes of heat transfer from the heated charcoal. Subsequently,the metallic materials at an elevated temperature start couplingwith microwaves leading to further rise in temperature. The tem-perature rise is eventually sufficient to cause melting and fusingof the interfaces. Fig. 3 illustrates a schematic view of the experi-mental setup of the microwave hybrid heating used for joining ofmetallic materials.

Microwave irradiation was carried out in a 1 kW multimodemicrowave system. Bulk pieces were exposed up to 900 s at a fre-quency of 2.45 GHz and a power of 900 W. Microwave processingparameters and other details are summarized in Table 2.

2.3. Characterisation of the joints

Joining of copper in bulk form was successfully carried out asdiscussed in the previous section. The copper joints were madeinto sections across the joint using a 200 lm thick diamond cutter,followed by mechanical polishing and etching. The joints werecharacterized through XRD, field emission scanning electronmicroscope (FESEM), microhardness, porosity measurement andstrength test.

The XRD patterns were obtained at room temperature in a Bru-ker AXS instrument with Cu-Ka X-ray. The scan rate was main-tained at 1� min�1 and the scan range was from 5� to 100�. Theanalysis of the metallic joint microstructure was carried out usinga field emission scanning electron microscope at an accelerationvoltage of 20 kV equipped with an energy dispersive X-ray detector(FEI Quanta 200 FEG-SEM, Czech Republic). The microhardness atdifferent positions of the joints was evaluated by Vickers microh-ardness tester (Mini load, Leitz, Germany) at the load of 10 g ap-plied for a duration of 30 s. The porosity of the joint area wasmeasured using linear count method. The joined specimens werealso subjected to standard tensile test using a universal testing ma-chine. The specimens were prepared according to the ASTM Desig-nation: E8/E8 M–09 [21] standard having gauge length of 18 mmand width 3 mm. The schematic of a standard tensile specimen isshown in Fig.4. Specimens were subjected to uniaxial tension atambient condition on a Hounsfield Monsanto (H25KS/05) machineat a strain rate of 8.3 � 10�3 mm/s.

3. Mechanism of joint formation

It has been already discussed in the previous section that, atroom temperature, bulk copper reflects microwaves owing to verylow skin depth. The skin depth of copper can be obtained by usingthe relation (1) [22]. Accordingly, the calculated skin depth forcopper is obtained in the order of �1.3 lm.

d ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

qpflrlo

rð1Þ

where d is the skin depth in lm, q is the resistivity of copper(=167.3 lX mm), f is frequency of microwaves (=2.45 GHz, in thepresent study), l is the magnetic permeability = lrl0, l0 is theabsolute permeability (=4p � 10�7 Henries/m), lr is the relativepermeability (=1, for copper).

This computed depth is less than the size of bulk copper piecesused in the present work resulting in reflection of microwaves. Inorder to avoid this problem, masks made of refractory brick

Fig. 3. Schematic of microwave hybrid heating process for joining of metallic copper.

Table 2Detailed specification and parametersused in microwave joining process.

Applicator Multimode

Microwavefrequency

2.45 GHz

Susceptormaterial

Charcoalpowder

Exposure time 300–900 sExposure power 900 W

Fig. 4. Tensile strength test specimen of microwave processed copper joint.

2688 M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694

material were prepared to accommodate the metallic pieces insideit, so that they are not directly exposed to microwaves. For initialcoupling of microwaves with metallic materials, charcoal powderwas used as a susceptor medium to facilitate microwave hybridheating. The metallic body was suitably sealed from microwaveexposure so as to avoid microwaves getting reflected. A solid layerof graphite was used to separate the charcoal powders and thesandwich powder layer.

Fig. 5 illustrates the interaction of microwaves with copper pow-der particles. Microwaves starts coupling with the highly absorbingmaterial charcoal (susceptor) as soon as the irradiation is initiated.However, the bulk copper pieces remain largely unaffected by themicrowave radiation owing to insignificant penetration of micro-waves (skin depth � 1.3 lm) relative to their sizes. The penetrationof microwaves into the metallic powders, on the other hand, will berelatively better owing the small size of the particles. Considering aspherical morphology as shown in the Fig. 5a with diameters 5 lm(as used in the present study), the microwaves will penetratethrough a significant volume of �58.2 cubic lm, which is approxi-mately 88.94% of the original volume of each individual particle.Thus, an interaction in the powder layer is likely to take place withmicrowaves. Further, the epoxy resin present in the sandwich layerabsorbs some amount of microwave energy and gets heated up.Simultaneously, heat generated in the charcoal layer due to themicrowave absorption gets transferred to the bulk metallic piecesand the metal powders through conventional modes of heat transfer.

The skin depth of the heated copper particles increases furtherowing to increase in resistivity and decrease in permeability withcontinued heating as obtained from the relation (1). This leads to en-hanced coupling of the metallic particles with microwaves resultingin high temperature and subsequent melting of the particles. Theskin depth on the bulk pieces, on the other hand, does not changeappreciably owing to the bulk effect (overall dimension is too large).Even the minimum dimension of the plate pieces to be joined (4 mm)is 3077 times larger than the skin depth of copper at room tempera-ture (�1.3 lm). At the same time, the change in temperature of thebulk pieces due to the conventional modes of heat transfer will notbe significant owing to unfavorable conditions. The rise in tempera-ture of the bulk copper pieces will thus be limited to a very thin layerwhich will get metallurgically bonded (fused) with the molten par-ticles in the sandwich layer. Further, since the microwaves penetratethroughout the sandwich layer, volumetric heating takes placeresulting in complete to near complete melting of the particles andsubsequent wetting of the candidate surfaces. On cooling, the mol-ten sandwich layer becomes the ‘weld bead’ of the conventionalmetallic welding as shown in Fig. 5b.

4. Results and discussion

Metallurgical bonding of thermally conductive materialsthrough conventional welding processes is extremely difficult. Inthe present work, trials were conducted for joining of copper inthe bulk form (coin and plate) using microwave hybrid heating.Metallurgical bonding has been achieved in this work. Joining pro-cess has been successfully carried out with controlled microwaveirradiation as discussed. A few typical joints of copper plates anda scanning electron microscope micrograph of a typical section ofa butt joint produced with an irradiation of 300 s is presented inFig. 6a and Fig. 6b respectively. Good metallurgical bonding be-tween the bulk metallic plates has been seen. The sandwich layerof the metallic powder has been observed to be completely fused.The area beyond the fused zone appears least affected by the heatof fusion as the rapid heating is initiated through the microwave-metal interaction. Results are discussed with suitable illustrationsin the following sections.

4.1. Analysis on XRD spectrum

A typical XRD spectrum of MHH induced copper joint is pre-sented in Fig. 7. The spectrum indicates five dominant copperpeaks of FCC lattice. Some oxides of copper were also detected inthe diffraction pattern. The joining was carried out in the atmo-spheric condition; consequently, the copper powder got partiallyoxidized and formed oxide of copper (CuO). It is well known thatthe metal oxides have significantly different properties from thoseof the pure metals. Thus, the presence of copper oxide during

Fig. 5. Schematic of (a) interaction of microwaves with powder particles and the bulk, (b) fusion of molten particles and the bulk pieces at the interface.

M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694 2689

microwave processing further facilitates enhanced coupling ofmicrowaves with the powder resulting in faster heating and cre-ates better bonding with oxide of copper (Cu2O) and bulk copper.Earlier, the importance of the presence of copper oxides during sin-tering of copper powders through microwave heating was also re-ported by Takayama et al. [23]. The peak corresponding to 2h � 18�(Fig. 7) indicates significant transformation of copper into CuO dur-ing microwave irradiation too. Further, the major Cu-peak (corre-sponding to 2h � 43�) with (1 1 1) orientation shows themaximum intensity; while the other minor peaks get significantlyattenuated. It is observed that the dominant (3 1 1) orientation inthe starting copper powder got transformed into a preferential ori-entation (1 1 1) apart from some got transformed into (2 2 2) and(2 0 0) structures during microwave irradiation with lattice strain-ing. This indicates the effect of microwave induced heating onchange in structural orientation of copper. Straining of lattices dur-ing microwave irradiation has also been reported earlier [5].

The phases were also analyzed using relative peak intensities ofthe respective phases. Peak intensities of different phases and nor-malized intensity ratio (NIR) are shown in Table 3. The normalizedintensity ratios were calculated by using the relation (2) [24].

NIRx ¼Ix � Iback

Ix þ Iy � 2Ibackð2Þ

where Ix and Iy are the intensities of the x and y phases respectivelyand Iback is the background intensity (�69 counts, in this study).

A similar calculation was carried out to evaluate Iy. Dominatingpeaks in the joint spectrum were considered in these calculations,which include (1 1 1) peak for copper and (1 0 1) peak for copperoxide. However, the NIR values do not indicate the exact amountof phases present in the joint, yet, provides relative representationof amount of phases. Thus, during microwave processing in theatmospheric condition, approximately 26% of copper powder gets

transformed into copper oxide (Table 3), while the remaining metal-lic copper (�74%) exists in various structures (111, 200, 220, 222and 311).

4.2. Observation on joint microstructure

Microstructures of typical copper joints developed by micro-wave hybrid heating are illustrated through scanning electronmicrographs in Fig. 8. The microstructures shown in Fig. 8a and bindicate complete melting of copper powder particles at the sand-wich zone and good metallurgical bond with the bulk pieces.Homogeneous and dense joint interface of fused copper metal withnegligible pores could be seen. The observed homogeneity of thejoint is attributed to the uniform microwave heating throughoutthe sandwich layer, also called the volumetric heating. Initially,the susceptor material (charcoal), couples with microwaves thatleads to rapid heating. This heat is transferred to the metallic pow-der particles in the sandwich layer within a short duration. Contin-ued microwave exposure at elevated temperature enhancesfurther microwave coupling and causes the powder particles tomelt and subsequently wetting of the bulk interface. In the meanwhile, the temperature of a thin layer on both the bulk interfaces(adjacent to the sandwich layer) also increases owing to the pres-ence of the heated powder layer. This facilitates better metallurgi-cal bonding; and on cooling, a uniform joint is obtained asillustrated in Fig.8. A fully fused weld interface (WI) can be seenclearly (Fig. 8a). Fusing of both the base metal surfaces with thesandwich layer is clearly evident in the Fig. 8b. A well-bondedmicrostructure on the sandwich powder layer and powder-bulkinterface shown in Fig. 8b is an indication of good joint efficiency.It is observed that no cracks are visible even at 2000 �magnifica-tion. The elemental distribution of the joint (Fig. 8) was analyzedthrough energy dispersive spectroscopy (EDS).

Fig. 6. (a) Typical view of a copper joint; (b) SEM micrograph of the joint.

Fig. 7. Typical XRD spectrum of the copper joint developed through microwaveenergy.

Table 3Relative phase intensities in copper joint.

SI. No. Phases Ix Iy Iback NIR (%)

1 Copper 3400 – 69 73.862 Copper oxide – 1250 69 26.14

2690 M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694

A typical EDS spectrum of the copper joint is presented in Fig. 9.The figure shows the relative presence of oxygen and copper mate-rial. The peaks in the spectrum indicate the weight percentages ofthe elements present in the joint which confirms the presence of

copper oxide as observed through XRD (Fig. 7). Further, traces ofcarbon appeared in the spectra is attributed to negligible mixing/diffusion of carbon from the separator sheet (graphite) at highertemperature during microwave irradiation. It is also possible thatthe epoxy used for making the slurry for the sandwich layer getsburnt at elevated temperature leaving some traces of carbon.

4.3. Observation on microhardness and joint porosity

Measurement of microhardness of the microwave induced jointwas also carried out on joint and on the base metal by using a loadof 10 g for 30 s. The mean microhardness at the joint was observedto be 78 ± 7 Hv as shown in Fig. 10. A narrow distribution ofmicrohardness in the joint area is an indication of uniform fusionof the particles. A relatively wider band of microhardness alongthe joint interface accounts for presence of possible porosity andbulk material effect on the measurement. The base material, onthe other hand, shows marginally higher microhardness owing toits higher purity level with less porosity. Average Vickers’ microh-ardness of the copper joints was found within 84% of the base me-tal (93 ± 12 Hv), which is significantly close. This can be partiallydue to the presence of copper oxide inclusions and a high degreeof fusion indicating good bonding between the particles.

Further, typical morphology of the indentations while measur-ing microhardness at different locations is illustrated throughSEMs in the insets of Fig. 10. Largely well defined indentationgeometry in the weld zone is indicator of well fused joint. How-ever, occasional collapsing of the indentation geometry in thesandwich layer as seen in the figure is attributed to the presenceof localized porous pockets and possible oxides and/or impurity.It is also an indication that the molten copper particles do not re-gain the ductility of bulk copper upon solidification following theend of microwave irradiation.

The indentation geometry at the base metal region within150 lm from the joint interface indicates the possibility of loss ofductility by the base metal copper to some extent. The measuredmicrohardness value of the joint as well as base metal was ob-served to be higher than the starting copper pieces. The loss of duc-tility might affect the strength of the copper joints.

Porosity of the copper joints was measured using linear countmethod. It was observed that the porosity in the joint region was1.92%. An important observation in the joint area is the formationof round pores as observed in Fig.8. Round pores, in general, con-tribute towards higher ductility of the zone and hence higherstrength than the materials having pores with sharp edges whichare normally observed in conventional welding techniques [18].The formation of round pores reduces the possibility of high stressconcentration region, and the material tends to exhibit higher duc-tility and strength. Ductility is associated with shear stress; how-ever, at the tip of sharp edges shear stress component are zeroand tri-axial stress states are setup leading to brittle failure ofthe joint area. Significantly sharp-edged pores are absent in themicrowave induced metallic joints as observed in the microstruc-tures presented in Fig. 8.

4.4. Observation on tensile strength

In order to evaluate tensile properties of the joints, the micro-wave hybrid heating induced copper joints were subjected to

Fig. 8. (a) SEM micrograph of the copper joint (WI: weld interface, BM: base metal, WZ: weld zone); (b) microstructure of the fused zone.

Fig. 9. Typical EDS spectrum of copper joint with relative elemental distribution at the joint.

M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694 2691

standard tensile test as discussed earlier. A uniform strain rate of8.3 � 10�3 mm/s was applied. The proof stress, ultimate tensilestrength (UTS) and percentage elongation or ductility of the jointwere recorded and presented in the Table 3. Fig. 11 shows thestress–strain diagram of microwave irradiated copper joints. Ini-tially, upto the approximate stress of 70 MPa (Segment OA,Fig. 11), the work material exhibits the characteristics of a mono-lithic material and undergoes uniform strain as depicted by thelinear relationship in the Fig. 11. Further increase in stress induceswork hardening effect in the joint region and continues till theloading corresponds to the ultimate tensile strength of164.4 MPa. The exponentially varying stress–strain characteristic(segment AB in Fig. 11) is attributed to strain hardening effect.However, the joints can no longer withstand further tensile loads.Loading beyond this limit results in sharp decrease in inducedstress as a precursor to failure and eventually fails corresponding

to a stress of approximately 20 MPa (Point C, Fig. 11). It is clearfrom the stress–strain characteristics that the deformation of thejoints is plastic mode dominated. The elongation of the joints(Table 4) subjected to tensile loading is obtained to be significant(29.21%, Table 4), which is attributed to reasonably good fusionin the powder particles and metallurgical bonding.

The fractured specimen during tensile testing was further char-acterized through scanning electron microscope. Typical SEMmicrograph of a fractured joint is illustrated in Fig. 12. It is ob-served from the fractured topography that the failure of the jointswas through mixed mode of failure. Evidence of both ductile andbrittle modes of failure in the joint zone is clearly seen. Sharphoneycomb-like fracture morphology is the signature of brittlefracture (Fig 12). Occurrence of ductile failure is, however, charac-terized by the presence of microscopic concave depressions on thefracture surface as shown in Fig. 12. It is important to note that the

Fig. 10. Vickers microhardness profile at various zones.

Fig. 11. Typical stress–strain behavior of the microwave induced copper joints.

Table 4Observed average values of the joints during tensile tests.

Sample Proof stress(MPa)

Ultimate tensile strength(MPa)

%Elongation

Microwaveinduced

71.8 164.4 29.21

Copper joints

2692 M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694

brittle failure is dominant in the zone where total melting of thepowder particles in the sandwich layer and subsequent resoilidifi-cation took place during microwave irradiation. The more elon-gated structures associated with ductile failure as seen in Fig. 12

are, on the other hand, largely located in the proximity to the par-ticles remained semi-molten during irradiation. Thus, it is evidentthat, those few semi-molten (during irradiation) particles may con-tribute towards better tensile strength of the joints. The resolidi-fied particles, it is observed, lose their ductility associated withmetallic structures to some extent. The fact was also observedwhile studying indentation geometry as illustrated in Fig. 10. Thechange in tensile properties can further be attributed to the changein the orientations of the copper atomic structures as observed inthe Fig. 2 and Fig. 7. As discussed earlier, the starting copper pow-der has the orientation of (3 1 1), while the copper in the fusedjoint layer (post microwave irradiation) records a dominating(1 1 1) orientation which primarily affect the strength and microh-ardness of the joint. The changes of orientation plane from (3 1 1)to (1 1 1) might cause lattice straining, which, in turn, inhibit crack

Fig. 12. Microstructure of fractured area.

M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694 2693

growth in the joint region. This results in increase in microhard-ness as well as tensile strength of the joint. However, as the appliedload increases continually, the necking in the joint zone increasesresulting in observed higher elongation. The induced brittlenesson the sandwich layer is also explained by the presence of copperoxide and carbon in the fused zone as observed in the Fig. 7 andFig. 9. This induced brittleness associated with microwave joining,could however, affect a loss of 40% of the original strength of thecopper. It is worth mentioning here that the preliminary resultsshow that the tensile strength of similar grade copper joints madethrough TIG welding was obtained as only159 MPa (a loss of 44% ofthe original strength), which inferior to the strength is obtainedwith microwave induced joints. Further investigations in thisdirection are being carried out.

5. Conclusions

Joining of metallic materials in bulk form using microwave en-ergy is challenging owing to reflection of microwaves by bulk met-als at room temperature. The present paper reports on the initialresults of development of copper (99.9% pure) joint using micro-wave hybrid heating technique. Metallurgical joining of copper inbulk form has been achieved in this work. Metallurgical fusion inthe joint area is near complete. Uniform microstructure in the jointregion confirms the uniform volumetric heating throughout thejoint. Bulk Copper joint with Vickers microhardness around 78 ±7 Hv has been obtained. Presence of oxygen leads to formation ofcopper oxide which facilitates better coupling of microwaves withthe metallic materials. The work establishes the basis for potentialmetallic material joining in bulk form using microwave energy,which is eco friendly and fast. The major conclusions of the presentwork can be summarized as-

(1) It is possible to join bulk metallic materials (for example,copper), using microwave energy.

(2) Microwave joining mechanism using microwave hybrid heat-ing has been explained and demonstrated for copper, one ofthe difficult materials to weld. Smaller metallic particle (pow-der) size is favorable for better interaction with microwaves.

(3) Homogeneous heating in the joint through metallic powderin the sandwich layer results in fusion of the powder parti-cles and wetting of the bulk interfaces during microwaveirradiation.

(4) Change of atomic structure of copper powder does takeplace during microwave irradiation; for example, Cu (3 1 1)gets transformed into Cu (1 1 1). Nearly 26% of metallic cop-per powder gets transformed into copper oxide duringmicrowave irradiation at atmospheric condition. Presenceof oxides in copper improves coupling of microwaves.

(5) Microhardness of the microwave induced joint area is nearer(within 84%) to the bulk metallic material.

(6) The microstructure in the joint is dense (porosity �1.92%).Predominantly round pores are formed at the joint zone pro-duced using microwave energy.

(7) The microwave processed copper joints exhibits significanttensile strength with significantly high elongation. This isdue to near complete melting of powder particles followedby good bonding between the interface surfaces. Failure ofthe microwave induced copper joints is due to both ductileas well brittle modes of failure.

Acknowledgement

Authors gratefully acknowledge the financial support receivedfrom the Indian Institute of Technology Roorkee, India, under theProject Grant No. FIG–100445. Inputs from research student Dhe-eraj Gupta and undergraduate student Vipul have been dulyacknowledged.

References

[1] Osepchuk John M. A history of microwave heating applications. IEEE TransMicrowave Theory Technol, MIT 1984;32(9):1200–24.

[2] Osepchuk John M. Microwave Power applications. IEEE Trans MicrowaveTheory Technol 2002;50(3):975–85.

[3] Sutton WH. Microwave processing of ceramic materials. Am Ceram Soc Bull1989;l68:376–86.

[4] Clark DE, Sutton WH, Lewis D. Microwave processing of materials. Annu RevMater Sci 1996;26:299–301.

[5] Sharma AK, Aravindan S, Krishnamurthy R. Microwave glazing of alumina–titania ceramic composite coatings. Matt Let 2001;50:295–301.

[6] Brosnan Kristen H, Messing Gary L, Agrawal DK. Microwave Sintering ofAlumina at 2.45 GHz. J Am Cer Soc 2003;86(8):1307–12.

[7] Rodiger K, Dreyer K, Gerdes T, Porada Willert M. Microwave Sintering ofHardmetals. Int J Refract Met Hards Mater 1998;16:409–16.

[8] Roy R, Agrawal DK, Cheng J, Gedevanishvili S. Full sintering of powdered metalparts in microwaves. Nature 1999;399:668–70.

[9] Prabhu G, Chakraborty Amitava, Sarma Bijoy. Microwave sintering of tungsten.Int. J Met Hard Mater 2009;27:545–8.

2694 M.S. Srinath et al. / Materials and Design 32 (2011) 2685–2694

[10] Mondal Avijit, Agrawal DK, Upadhyaya Anish. Microwave heating of purecopper powder with varying particle size and porosity. J Microwave PowerElectromag Energy 2009;43(1):5–10.

[11] Gupta M, Wong WLE. Enhancing overall mechanical performance of metallicmaterials using two-directional microwave assisted rapid sintering. ScriptaMater 2005;52:479–83.

[12] Aravindan S, Krishnamurthy R. Joining of ceramic composites by microwaveheating. Matt Let 1999;38:245–9.

[13] Ahmed Ammar, Siores Elias. Microwave joining of 48%alumina–32%zirconia–20%silica ceramics. J Mater Process Technol 2001;118:88–95.

[14] Barmatz Martin, Jackson W, Radtke P Robert. Microwave Technique for BrazingMaterials. US patent 2000; 6054693.

[15] Sallom ZK, Akselsen OM, Zhang J. Brazing of gamma TiAl with Ag based Fillermetal. Euro mater 2005:1–30.

[16] Budinger David Edwin. Microwave brazing process. US Patent 2008;0290137.

[17] Siores E, Rego D. Microwave applications in materials joining. J Mater ProcessTechnol 1995;48:619–25.

[18] Agrawal DK. Microwave sintering, Brazing and melting of metallic materials ofmetallic materials. Sohn Int Symp, Adv Process Met Mater 2006;4:183–92.

[19] Sharma AK, Srinath MS, Kumar Pradeep. Microwave joining of metallicmaterials. Indian Patent 2009; Application No.1994/Del/2009.

[20] Sharma AK, Gupta Dheeraj. A method of cladding/coating of metallic and non-metallic powders on metallic substrates by microwave irradiation. IndianPatent 2010; Application No.527/Del/2010.

[21] ASTM Designation: E8/E8M–09, Standard test methods for tension testing ofmetallic materials. American association state highway and transportationofficials standard AASHTO No. T68. An American National Standard 2009; 1–3.

[22] Metaxes AC, Meredith RJ. Engineer’s handbook of industrial microwaveheating. London: Peter Peregrinus; 1988.

[23] Sadatsugu Takayama, Yasushi Saito, Motoyasu Sato, Takuya Nagasaka, TakeoMuroga, Yoshihiko Ninomiya. Sintering behavior of metal powders involvingmicrowave-enhanced chemical reaction. Jpn J Appl Phys 2006;45(3A):1816–22.

[24] Peelamedu Ramesh D, Rustum Roy, Agrawal DK. Microwave-induced reactionsintering of NiAl2O4. Matt Lett 2002;55:234–40.