nano-conductive adhesives for nano-electronics …...nano-conductive adhesives for nano-electronics...

Nano-conductive Adhesives for Nano-electronicsInterconnection

Yi Li, Kyoung-sik (Jack) Moon, and C.P. Wong

Abstract With the phasing out of lead-bearing solders, electrically conductiveadhesives (ECAs) have been identified as one of the environmentally friendlyalternatives to tin/lead (Sn/Pb) solders in electronics packaging applications. In par-ticular, with the requirements for fine-pitch and high-performance interconnectsin advanced packaging, nanoconductive adhesives are becoming more and moreimportant due to the special electrical, mechanical, optical, magnetic, and chemicalproperties that nano-sized materials can possess. There has been extensive researchfor the last few years on materials and process improvement of ECAs, as well asthe advances of nanoconductive adhesives that contain nano-filler such as nanoparti-cles, nanowires, or carbon nanotubes and nanomonolayer graphenes. In this chapter,recent research trends on electrically conductive adhesives (ECAs) and their relatednanotechnologies are discussed, with the particular emphasis on the emerging nan-otechnology, including materials development and characterizations, processingoptimization, reliability improvement, and future challenges/opportunities identi-fication. The state of the art on nanoisotropic/anisotropic conductive adhesivesincorporated with nanosilver, carbon nanotubes, and nanonickel, and their recentstudies on those for flexible nano/bioelectronics, transparent electrodes, and jettableprocesses are addressed in this chapter. Future studies on nanointerconnect materialsare discussed as well.

Keywords Electrically conductive adhesives (ECAs) · Isotropically conduc-tive adhesives (ICA) · Percolation threshold · Nanotechnology · Carbon nan-otubes (CNT) · Self-assembled monolayer (SAM) · Thermo-compression bonding(TCB) · Sintering · Current-carrying capability · Anisotropic conductive adhe-sives (ACAs) · Fine-pitch interconnection · Nanoconductive fillers · Transparentelectrodes

Y. Li (B)Intel Corporation, Chandler, AZ, USAe-mail: [email protected]

19C.P. Wong et al. (eds.), Nano-Bio- Electronic, Photonic and MEMS Packaging,DOI 10.1007/978-1-4419-0040-1_2, C© Springer Science+Business Media, LLC 2010

20 Y. Li et al.

1 Introduction

Miniaturization of electronic and photonic devices is a demanding trend in cur-rent and future technologies including consumer products, military, biomedical, andspace applications. In order to achieve this miniaturization, new packaging tech-nologies have emerged such as embedded active and passive components and 3Dpackaging, where new packaging designs and materials are required.

Fine-pitch interconnection is one of the most important technologies for thedevice miniaturizations. Metal solders including eutectic tin/lead solder and lead-free alloys are the most common interconnect materials in the electronic/photonicpackaging areas.

Conventional eutectic tin/lead solder or lead-free solders cannot meet the require-ments for fine-pitch assembly due to their stencil printing resolution limit andbridging issues of solders in fine-pitch bonding, which is an intrinsic characteristicof metal solders.

Recently, polymer-based nanomaterials have been intensively studied to replacethe metal-based interconnect materials, and many research efforts to enhancematerial properties of the electrically conductive adhesives (ECAs) are on going.Furthermore, incorporating nanotechnology into the ECA systems not only enablesthe ultra-fine-pitch capability but also enhances the electrical properties such asreducing the percolation threshold and improving the electrical conductivity.

In this chapter, recent research and study trends on ECAs and their relatednanotechnologies are discussed, and the future of the nanomaterial-based polymercomposites for fine-pitch interconnection is reviewed.

Electrically conductive adhesives (ECAs) are composites of polymeric matri-ces and electrically conductive fillers. The polymeric resin, such as an epoxy, asilicone, or a polyimide, provides physical and mechanical properties such as adhe-sion, mechanical strength, impact strength, and the metal filler (such as, silver, gold,nickel, or copper) conducts electricity. Metal-filled thermoset polymers were firstpatented as electrically conductive adhesives in the 1950s [1–3]. Recently, ECAmaterials have been identified as one of the major alternatives for lead-containingsolders for microelectronics packaging applications. ECAs offer numerous advan-tages over conventional solder technology, such as environmental friendliness, mildprocessing conditions (enabling the use of heat-sensitive and low-cost componentsand substrates), fewer processing steps (reducing processing cost), low stress onthe substrates, and fine-pitch interconnect capability (enabling the miniaturizationof electronic devices) [4–9]. Therefore, conductive adhesives have been used in flatpanel displays such as LCD (liquid crystal display), and smart card applications asan interconnect material and in flip-chip assembly, CSP (chip-scale package) andBGA (ball grid array) applications in replacement of the solder. However, no cur-rently commercialized ECAs can replace tin-lead metal solders in all applicationsdue to some challenging issues such as lower electrical conductivity, conductivityfatigue (decreased conductivity at elevated temperature and humidity aging or nor-mal use condition) in reliability testing, limited current-carrying capability, and poorimpact strength. Table 1 gives a general comparison between tin-lead solder and ageneric commercial ECA [6].

Nano-conductive Adhesives for Nano-electronics Interconnection 21

Table 1 Conductive adhesives compared with solder

Characteristic Sn/Pb solder ECA (ICA)

Volume resistivity 0.000015� cm 0.00035 � cmTypical junction R 10–15 mW <25 mWThermal conductivity 30 W/m-deg.K 3.5 W/m-deg.KShear strength 2200 psi 2000 psiFinest pitch 300 μm <150–200 μmMinimum processing temperature 215◦C <150–170◦CEnvironmental impact Negative Very minorThermal fatigue Yes Minimal

Depending on the conductive filler loading level, ECAs are divided into isotrop-ically conductive adhesives (ICAs), anisotropically conductive adhesives (ACAs)and nonconductive adhesives (NCAs). For ICAs, the electrical conductivity in allx-, y-, and z-directions is provided due to high filler content exceeding the percola-tion threshold. For ACAs or NCAs, the electrical conductivity is provided only inz-direction between the electrodes of the assembly. Figure 1 shows the schematicsof the interconnect structures and typical cross-sectional images of flip-chip jointsby ICA, ACA, and NCA materials illustrating the bonding mechanism for all threeadhesives.

Fig. 1 Schematic illustrations and cross-sectional views of (a, b) ICA, (c, d) ACA, and (e, f) NCAflip-chip bondings

Isotropic conductive adhesives, also called as “polymer solder,” are compos-ites of polymer resin and conductive fillers. The adhesive matrix is used to form amechanical bond for the interconnects. Both thermosetting and thermoplastic mate-rials are used as the polymer matrix. Epoxy, cyanate ester, silicone, polyurethane,etc., are widely used thermosets, and phenolic epoxy, maleimide acrylic preimidized

22 Y. Li et al.

polyimide, etc., are the commonly used thermoplastics. An attractive advantageof thermoplastic ICAs is that they are reworkable, i.e., can easily be repaired. Amajor drawback of thermoplastic ICAs, however, is the degradation of adhesionat high temperature. A drawback of polyimide-based ICAs is that they generallycontain solvents. During heating, voids are formed when the solvent evaporates.Most of commercial ICAs are based on thermosetting resins. Thermoset epox-ies are by far the most common binders due to the superior balanced properties,such as excellent adhesive strength, good chemical and corrosion resistances, andlow cost, while thermoplastics are usually added to allow toughening and reworkunder moderate heat. The conductive fillers provide the composite with electricalconductivity through physical contact between the conductive particles. The pos-sible conductive fillers include silver (Ag), gold (Au), nickel (Ni), copper (Cu),and carbon in various forms (graphites, carbon nanotubes, etc.), sizes, and shapes.Among different metal particles, silver flakes are the most commonly used con-ductive fillers for current commercial ICAs because of the high conductivity andthe maximum contact between flakes. In addition, silver is unique among all thecost-effective metals by nature of its conductive oxide. Oxides of most commonmetals are good electrical insulators. Copper powder, for example, becomes a poorconductor after oxidation/aging. Nickel and copper-based conductive adhesives gen-erally do not have good resistance stability. As such, electrical reliability issues ofECA joints, in particular on non-noble metal finishes at elevated temperature andhigh relative humidity are of critical concern for implementing ECAs into the high-performance device interconnects. Intensive studies on this ECA joint failure ledto understanding of the failure mechanisms, that corrosion at the ECA joints wasthe underlying mechanism rather than a simple oxidation process of the non-noblemetal used [10, 11] (Fig. 2). Under the elevated temperature/humidity (for exam-ple, 85◦C/85%RH), the non-noble metal acts as an anode, and is oxidized by losingelectrons, and then turns into metal ion (M–ne– = Mn+). The noble metal acts as acathode, and its reaction generally is 2H2O + O2 + 4e– = 4OH–. Then Mn+ com-bines with OH– to form a metal hydroxide, which further oxidizes to metal oxide.After corrosion, a layer of metal hydroxide or metal oxide is formed at the inter-face. Because this layer is electrically insulating, the contact resistance increasesdramatically.

Since this underlying mechanism was reported, many research activities for reli-able ECAs have been intensively conducted via various approaches, such as usinglow moisture absorption resins, incorporating corrosion inhibitors or oxygen scav-engers, using sacrificial anode materials and oxide-penetrating sharp particles [10,12–15]. In addition, many approaches to improve the electrical conductivity ofECAs have been reported, such as increasing the cure shrinkage of the matrix resin,replacing a lubricant of Ag flake with a shorter chain length carboxylic acid, addingan in-situ reducing agent for preventing the Ag flake-surface oxidation and addinglow melting point alloys (LMA) for enhancing conductivity [16–21]. With the helpof those intensive researches on enhancing the ECA performance, recently, ICAs


Non-noble metal

Ag flake

Polymer binder

Electrical conduction

Ag

(Before corrosion)

Non-noble metal

Ag

Ag flake

Polymer binder

Condensed water solution

Metal hydroxide or oxide

Electrical conduction

(After corrosion)

Fig. 2 Metal hydroxide oroxide formation aftergalvanic corrosion

have also been considered as an alternative to tin/lead solders in surface mounttechnology (SMT) [22, 23], flip chip, [24] and other applications.

Anisotropic conductive adhesives (ACAs) or anisotropic conductive films(ACFs) provide uni-directional electrical conductivity in the vertical or Z-axis. Thisdirectional conductivity is achieved by using a relatively low-volume loading ofconductive fillers (5–20 vol.%). The low-volume loading is insufficient for inter-particle contact and prevents conductivity in the X–Y plane of the adhesives. TheZ-axis adhesive, in film or paste form, is interposed between the surfaces to beconnected. Heat and pressure are simultaneously applied to this stack-up until theparticles bridge the two conductor surfaces.

The ACF-bonding process is a thermo-compression bonding as shown inFig. 3. In case of tape-carrier packages (TCP) bonding, the ACF material is attachedon a glass substrate and pre-attached. Final bonding is established by the thermalcure of the ACF resin, typically at 150–220◦C, 3–20 s, and 100–200 MPa, andthe conductive particle deformation between the electrodes of TCP and the glasssubstrate by applied bonding pressure.

Interconnection technologies using ACFs are major packaging methods forflat panel display modules to be of high resolution, light weight, thin profile,

24 Y. Li et al.

Electrode

Substrate

ACF

Circuit

Circuit

Substrate

Substrate

Circuit

Circuit

Substrate

Heat &Pressure

Fig. 3 Thermo-compressionbonding using ACF

and low power consumption [25], and are already successfully implemented inthe forms of outer lead bonding (OLB), flex to PCB bonding (FPCB), reli-able direct chip attach such as chip-on-glass (COG), chip-on-film (COF) for flatpanel display modules [26–29], including liquid crystal display (LCD), plasmadisplay panel (PDP), and organic light-emitting diode display (OLED). As forthe small and fine-pitched bump of driver ICs to be packaged, fine-pitch capa-bility of ACF interconnection is much more desired for COG, COF, and evenOLB assemblies. There have been advances in development works for improvedmaterial system and design rule for ACF materials to meet fine-pitch capa-bility and better adhesion characteristics of ACF interconnection for flat paneldisplays.

In addition to the LCD industry, ACA/ACF is now finding various applica-tions in flex circuits and surface mount technology (SMT) for chip-scale package(CSP), application-specific integrated circuit (ASIC), and flip-chip attachment forcell phones, radios, personal digital assistants (PDAs), sensor chip in digital cam-eras, and memory chip in lap-top computers. In spite of the wide applicationsof ACA/ACF, there are some key issues that hinder their implementations ashigh-power devices. The ACA/ACF joints generally have lower electrical con-ductivity, poor current-carrying capability, and electrical failure during thermalcycling.

To meet the requirements for future fine-pitch and high-performance intercon-nects in advanced packaging, ECAs with nanomaterial or nanotechnology attractmore and more interests due to the specific electrical, mechanical, optical, magnetic,and chemical properties. There has been extensive research on nanoconductiveadhesives, which contain nano-filler such as nanoconductive particles, nanowires,or carbon nano tube (CNT). This chapter will provide a comprehensive review ofmost recent research results on nanoconductive adhesives.


2 Recent Advances on Nanoisotropic Conductive Adhesive(Nano-ICA)

2.1 ICAs with Silver Nanowires

Wire-type nano1D materials (such as silver nanowires and CNT) have beenaddressed for their interconnect applications because of the flexibility and loweringthe percolation threshold in composites. Those advantages bring stress absorbingas well as light-weight interconnections. Silver nanowires are one of the impor-tant 1D materials for interconnect due to their high electrical conductivity. An ICAfilled with nanosilver wires has been reported and compared to two other ICAsfilled with micrometer-sized (roughly 1 μm and 100 nm, respectively) silver par-ticles for the electrical and mechanical properties [30]. It was found that, at a lowfiller loading (e.g., 56 wt%), the bulk resistivity of ICA filled with the Ag nanowireswas significantly lower than the ICAs filled with 1 μm or 100 nm silver particles.The better electrical conductivity of the ICA-filled nanowires was contributed to thefewer number of contact (lower contact resistance) between nanowires, more stableconductive network, and more significant contribution from the tunneling effectsamong the nanowires.

It was also found that, at the same filler loading (e.g., 56 wt%), the ICAs filledwith Ag nanowires showed similar shear strength to those of the ICAs filled with1 μm and 100 nm, respectively. However, to achieve the same level of electricalconductivity, the filler loading must be increased to at least 75 wt% for the ICAfilled with micro-sized Ag particles, and the shear strength of these ICAs is thendecreased (lower than that of the ICA filled with 56 wt% nanowires) due to thehigher filler loading.

Chen et al. reported synthetic routes of Ag nanowires for conductive adhesiveapplications and demonstrated that the electrical conductance of a multicomponentsystem (mixture of different geometry fillers) was better than that of a single-component system (e.g., sphere or flake themselves). A possible mechanism wasdescribed in the report as these nanowires act as bridges to establish perfect linkageamong particles, and the probability of contact and contact area becomes more thanthe cases without wires [31].

2.2 Effect of Nano-sized Silver Particles to the Conductivityof ICAs

Nano-sized fillers are being introduced into the ECA to replace the micro-sizedsilver flakes in recent research. Lee et al. studied the effects of nano-sized fillerto the conductivity of conductive adhesives by substituting micro-sized Ag par-ticles with nano-sized Ag colloids either in part or as a whole to a polymericsystem (polyvinyl acetate - PVAc) [32]. It was found that, when nano-sized sil-ver particles were added into the system at 2.5 wt% each increment, the resistivity

26 Y. Li et al.

increased in almost all the cases, except when the quantity of micro-sized silverwas slightly lower than the threshold value. At that point, the addition of about2.5 wt% brought a significant decrease in resistivity. Near the percolation thresh-old, when the micro-sized silver particles were still not connected, the additionof a small amount of nano-sized silver particles helped to build the conductivenetwork and lowers the resistivity of the composite. However, when the filler load-ing was above the percolation threshold and all the micro-sized particles wereconnected, the addition of nano-particles seemed only to signify the relative con-tribution of contact resistance between the particles. Due to its small size, for afixed amount of addition, the nano-sized silver colloid contained a larger numberof particles when compared with micro-sized particles. This large number of par-ticles should be beneficial to the connection between particles. However, it alsoinevitably increased the contact resistance. As a result, the overall effect was anincrease in resistivity upon the addition of nano-sized silver colloids. They alsostudied the effects of temperature on the conductivity of ICAs. Heating the com-posite to a higher temperature could reduce the resistivity significantly. This islikely due to the high surface activity of nano-sized particles. For micro-sizedpaste, this temperature effect was considered negligible. The interdiffusion of sil-ver atoms among nano-sized particles helped to reduce the contact resistance quitesignificantly and the resistivity reached 5×10–5 � cm after treatment at 190◦C for30 min.

Ye et al. reported that the addition of nanoparticles showed a negative effecton electrical conductivity [33]. They proposed two types of contact resistance,i.e., restriction resistance due to small contact area and tunneling resistance whennanoparticles are included in the system. It was believed that the conductivityof micro-sized Ag particle-filled adhesives could be dominated by constrictionresistance, while the nanoparticle containing conductive adhesives is controlledby tunneling and even thermionic emission. Fan et al. also observed a simi-lar phenomenon (adding nano-size particles reduced both electrical and thermalconductivities) [34].

The overall resistance (Rtotal) of the isotropic conductive adhesive (ICA) for-mulation is the sum of the resistance of fillers (Rfillers), the resistance between fillers(Rbtw fillers), and the resistance between filler and pads (Rfiller to bond pad) (Equation 1).In order to decrease the overall contact resistance, the reduction of the number ofcontact points between the particles may be obviously effective. Incorporation ofnano-fillers increases the contact resistance and reduces the electrical performanceof the ICAs. The number of contacts between the small particles is larger than thatbetween the large particles as shown in Fig. 4a, b. A recent study by Jiang et al.showed that nano-silver particles could exhibit sintering behavior at curing tem-perature of ICAs [35, 36]. If nanoparticles are sintered together, then the contactbetween fillers will be fewer. This will lead to smaller contact resistance (Fig. 4c).By using effective surfactants for the dispersion and effective capping those nano-sized silver fillers in ECAs, obvious sintering behavior of the nano-fillers can beachieved. The sintering of nano-silver fillers improved the interfacial properties ofconductive fillers and polymer matrices and reduced the contact resistance between


Fig. 4 Schematic illustration of particles between the metal pads [35]

fillers. The dispersion and interdiffusion of silver atoms among nano-sized particlescould be facilitated and the resistivity of ICA could be reduced to 5×10–6 � cm.

Rtotal = Rbtw fillers + Rfiller to bond pad + Rfillers (1)

2.3 ICA Filled with Aggregates of Nano-sized Ag Particles

To improve the mechanical properties under thermal cycling conditions while stillmaintaining an acceptably high level of electric conductivity, Kotthaus et al. studiedan ICA material system filled with aggregates of nano-sized Ag particles [37]. Theidea was to develop a new filler material which did not sacrifice the mechanicalproperty of the polymer matrix to such a great extent. A highly porous Ag powderwas attempted to fulfill these requirements. The Ag powder was produced by theinert gas condensation method. The powders consist of sintered networks of ultra-fine particles in the size range 50–150 nm. The mean diameter of these aggregatescould be adjusted down to some microns. The as-sieved powders were characterizedby low level of impurity content, an internal porosity of about 60%, a good abilityfor resin infiltration.

Using Ag IGC instead of Ag flakes is more likely to retain the properties of theresin matrix because of the infiltration of the resin into the pores. Measurements ofthe shear stress–strain behavior indicated that the thermo-mechanical properties ofbonded joints may be improved by a factor up to two, independent of the chosenresin matrix.

Resistance measurements on filled adhesives were performed within a temper-ature range from 10 to 325 K. The specific resistance of a Ag IGC-filled adhesiveis about 10–2 � cm and did not achieve the typical value of commercially avail-able adhesives of about 10–4 � cm. The reason may be that Ag IGC particles aremore or less spherical whereas Ag flakes are flat. So, the decrease of the per-colation threshold because of the porosity of Ag IGC is overcompensated by thedisadvantageous shape and the intrinsically lower specific conductivity. For certainapplications where mechanical stress plays an important role, this conductivity maybe sufficient and therefore the porous Ag could be suitable as a new filler materialfor conductive adhesives.

28 Y. Li et al.

2.4 Nano-Ni Particle-Filled ICA

Sumitomo Electric Industries, Ltd. (SEI) has developed a new liquid-phase deposi-tion process using plating technology. This new nanoparticle fabrication processachieves purity greater than 99.9% and allows easy control of particle diame-ter and shape. The particles’ crystallite size calculated from the result of X-raydiffraction measurement is 1.7 nm, which leads to an assumption that the parti-cle size of primary particles is extremely small. When the particle size of nickeland other magnetic metals becomes smaller than 100 nm, they are changed fromthe multidomain particles to the single-domain particles, and their magnetic prop-erties change. That is, if the diameter of nickel particles is around 50 nm, eachparticle acts like a magnet that has only a pair of magnetism, and magneticallyconnects with each other to form the chain-like clusters. When the chain-like clus-ters are applied to conductive paste, electrical conduction of the paste is expectedto be better than the existing paste. The developed chain-like nickel particles weremixed with pre-defined amount of polyvinylidene difluoride (PVdF) that acts as anadhesive. Then, N-methyl-2-pyrrolidone was added to this mixture to make con-ductive paste. This paste was applied on a polyimide film and then dried to makea conductive sheet. Specific volume resistivity of the fabricated conductive sheetwas measured by the quadruple method. The same measurement was also con-ducted on the conductive sheet that used the paste made of conventional sphericalnickel particles. Measurement of the sheet resistance immediately after paste appli-cation defined that the developed chain-like nickel powders had low resistance ofabout one-eighth of that of the conventionally available spherical nickel particles.This result showed that, when the newly developed chain-like nickel particles wereapplied as conductive paste, high conductivity can be achieved without pressing thesheet.

2.5 Nano-conductive Adhesives for Via-Filling Applicationsin Organic Substrates

The demand of higher density organic substrates is increasing recently.Traditionally, greater wiring densities are achieved by reducing the dimensions ofvias, lines, and spaces, increasing the number of wiring layers, and utilizing blindand buried vias. However, each of these approaches possess inherent limitations,for example, those related to drilling and plating of high aspect ratio vias, reducedconductance of narrow circuit lines, and increased cost of fabrication related to addi-tional wiring layers. One method of extending wiring density beyond the limitsimposed by these approaches is to form more metal-to-metal Z-axis interconnec-tion of sub-composites during lamination to form a composite structure. Conductivejoints can be formed during lamination using an electrically conductive adhesive. Asa result, one is able to fabricate structures with vertically terminated vias of arbitrarydepth. Replacement of conventional plated through holes with vertically terminated


vias opens up additional wiring channels on layers above and below the terminatedvias and eliminates via stubs, which cause reflective signal loss.

Conductive adhesives usually have high filler loading that weaken the overallmechanical strength. Therefore, reliability of the conductive joint formed betweenthe conductive adhesive and the metal surface to which it is mated is of primeimportance. Conductive adhesives can have broad particle size distributions. Largerparticles can be a problem when filling smaller holes (e.g., diameter of 60 μm orless), resulting in voids.

Das et al. developed conductive adhesives using controlled-sized particles, rang-ing from nanometer scale to micrometer scale, and used them to fill small diameterholes to fabricate Z-axis interconnections in laminates for interconnect applications[38]. A variety of metals, including Cu, Ag, and low melting point (LMP) alloyswith particle sizes ranging from 80 nm to 15 μm were used to make the epoxy-based conductive adhesives. Among all, silver-based adhesives showed the lowestvolume resistivity and highest mechanical strength. It was found that with increas-ing curing temperature, the volume resistivity of the silver-filled paste decreaseddue to sintering of metal particles. Sinterability of silver adhesive was further eval-uated using high-temperature/pressure lamination and shows a continuous metallicnetwork when laminated at 365◦C.

As a case study, they used a silver-filled conductive adhesive as a Z-axis inter-connecting construction for a flip-chip plastic ball grid array package with a 150 μmdie pad pitch. High aspect ratio, small diameter (∼55 μm) holes were success-fully filled. Silver-filled adhesives were electrically and mechanically better thanCu and LMP-filled adhesives. All adhesives maintained high tensile strength evenafter 1000 cycles thermal cycling (–55 to 120◦C). Conductive joints were stableafter three times reflowing, 1000 cycles thermal cycling, pressure cooker test (PCT),and solder shock. The adhesive-filled joining cores were laminated with circuitizedsubcomposites to produce a composite structure. High-temperature/pressure lami-nation was used to cure the adhesive in the composite and provide stable, reliableZ-interconnections among the circuitized subcomposites.

2.6 Nano-ICAs Filled with CNT

2.6.1 Electrical and Mechanical Characterization of CNT-Filled ICAs

Carbon nanotubes are a new form of carbon, which was first identified in 1991by Sumio Iijima of NEC, Japan [39]. Nanotubes are sheets of graphite rolled intoseamless cylinders. Besides growing single wall nanotubes (SWNTs), nanotubescan also have multiple walls (MWNTs) – cylinders inside the other cylinders,which act as multichannel transport of electrical and thermal for thermal interfacematerials (TIMs) applications. The carbon nanotube can be 1–50 nm in diame-ter and 10–100 μm or up to a few centimeters in length, with each end “capped”with half of a fullerene dome consisting of five- and six-member rings. Along thesidewalls and cap, additional molecules can be attached to functionalize the

30 Y. Li et al.

nanotube to adjust its properties. CNTs are chiral structures with a degree of twist inthe way that the graphite rings join into cylinders. The SWCNT chirality determineswhether a nanotube will conduct in a metallic or semiconducting manner. Carbonnanotubes possess many unique and remarkable properties. The measured electricalconductivity of metallic carbon nanotubes is in the order of 104 S/cm. The thermalconductivity of carbon nanotubes at room temperature can be as high as 6600 W/mK [40]. The Young’s modulus of carbon nanotube is about 1 TPa. The maximumtensile strength of carbon nanotube is close to 30 GPa, some reported at TPa [41].Since carbon nanotubes have very low density and long aspect ratios, they have thepotential of reaching the percolation threshold at very low weight percent loadingin the polymer matrix.

Epoxy-based conductive adhesives filled with MWNTs have been recently devel-oped [42]. It was found that ultrasonic mixing process helped disperse CNTs inthe epoxy more uniformly and made them contact better, and thus lower elec-trical resistance was achieved. The contact resistance and volume resistivity ofthe conductive adhesive decreased with increasing CNT loading. The percola-tion threshold for the MWNTs used was less than 3 wt%. With 3 wt% loading,the average contact resistance was comparable with solder joints. It was alsofound that the performance of CNTs-filled conductive adhesive was compara-ble with solder joints in high frequency. By replacing metal particle fillers withCNTs in the conductive adhesive, a higher percentage of mechanical strengthwas retained. For example, with 0.8 wt% of CNT content, the 80% of shearstrength of the polymer matrix was retained, while conventional metal-filled con-ductive adhesives only retain less than 28% of shear strength of the polymermatrix.

Experiments conducted by Qian et al. [43] show 36–42% and 25% increases inelastic modulus and tensile strength, respectively, in polystyrene (PS)/CNT com-posites. The TEM observations in their experiments showed that cracks propagatedalong weak CNT–polymer interfaces or relatively low CNT density regions andcaused failure. If the outer layer of MWNTs can be functionalized to form strongchemical bonds with the polymer matrix, the CNT/polymer composites can be fur-ther reinforced in mechanical strength and have controllable thermal and electricalproperties.

2.6.2 Effect of Adding CNT to the Electrical Properties of ICAs

Effect of adding CNT to the electrical conductivity of silver-filled conductive adhe-sive with various CNT loadings were reported [44]. It was found that the CNT couldenhance the electrical conductivity of the conductive adhesives greatly when thesilver filler loading was still below percolation threshold. For example, 66.5 wt%silver-filled conductive adhesive without CNT had a resistivity of 104 � cm, butshowed a resistivity of 10–3� cm after adding 0.27 wt% CNT. Therefore, it is pos-sible to achieve the same level of electrical conductivity by adding a small amountof CNT to replace the silver fillers.


2.6.3 Composites Filled with Surface-Treated CNTs

Although CNTs have exceptional physical properties, incorporating CNTs into othermaterials has been inhibited by the surface chemistry of carbon. Problems suchas phase separation, aggregation, poor dispersion within a matrix, and poor adhe-sion to the host must be overcome. Zyvex claimed that they have overcome theserestrictions by developing a new surface treatment technology that optimizes theinteraction between CNTs and the host matrix [45]. A multifunctional bridge wascreated between the CNT sidewalls and the host material or solvent. The powerof this bridge was demonstrated by comparing the fracture behavior of the com-posites filled with untreated and surface-treated nanotubes. It was observed thatthe untreated nanotubes interacted poorly with the polymer matrix, and thus leftbehind voids in the matrix after fracture. However, for composite filled with treatednanotubes, the nanotube remained in the matrix even after the fracture, indicatingstrong CNT interaction with the matrix. Due to their superior dispersion in the poly-mer matrix, the treated nanotubes achieved the same level of electrical conductivityat much lower loadings than the untreated nanotubes [45].

2.7 Inkjet Printable Nano-ICAs and Inks

Areas for printing very fine-pitch matrix (e.g., very fine-pitch paths, antennas) arevery attractive. But there are special requirements for inkjet printing materials,namely the most important ones are low viscosity and very homogenous structure(such as molecular fluid) to avoid sedimentation and separation during the process.Additionally, for electrical conductivity of printed structure, the liquid has to containconductive particles, with nano-sized dimension to avoid the printing nozzle block-ing and prevent sedimentation phenomenon. The nano-sized silver seems to be oneof the best candidates for this purpose, especially when its particle size dimensionswill be less than 10 nm.

Inkjetting is an accepted technology for dispensing small volumes of material(50–500 pl). Currently, traditional metal-filled conductive adhesives cannot be pro-cessed by inkjetting (due to their relatively high viscosity and the size of fillermaterial particles). Smallest droplet size achievable by traditional dispensing tech-niques is in the range of 150 μm, yielding proportionally larger adhesive dotson the substrate. Electrically conductive inks are available on the market withmetal particles (gold or silver) <20 nm suspended in a solvent at 30–50 wt%.After deposition, the solvent is evaporated and electrical conductivity is enabledby a high metal ratio in the residue. Some applications include a sintering step[46, 47].

There are many requirements for an inkjettable, Ag particle-filled conductiveadhesive. The silver particles must not exceed a maximum size determined by thediameter of the injection needle used. At room temperature, the adhesive shouldresist sedimentation for at least 8 , preferably 24 hours. A further requirement bythe end user on the adhesive’s properties was a two-stage curing mechanism. In the

32 Y. Li et al.

first curing step, the adhesive surface is dried. In this state, the product may be storedfor several weeks. The second curing step involves glueing the components with thepreviously applied adhesive. By heating and applying pressure the adhesive is cured.Thus, the processing operation is similar to that required for soldering. A conduc-tivity in the range of 10–4� cm in the bulk material is required. An adhesive lessprone to sedimentation was formulated by using suitable additives. Furthermore, theformation of filler agglomerations during deflocculation and storage was reduced.This effect was achieved by making the additives adhere to the filler particle sur-faces. This requires a very sensitive balance. If the insulation between individualsilver particles becomes very strong, overall electrical conductivity is significantlyreduced.

Jana Kolbe et al. demonstrated feasibility of an inkjettable, isotropically conduc-tive adhesive in the form of a silver-loaded resin with a two-step curing mechanism[48, 49]. In the first step, the adhesive was dispensed (jetted) and precured leavinga “dry” surface. The second step consisted of assembly (wetting of the second part)and final curing. The attainable droplet sizes were in the range of 130 μm, but couldbe further reduced by using smaller (such as 50 μm) and more advanced nozzleshapes.

Nano-Ag particles for nano-ink applications are coated with capping agentsfor dispersion, which should be removed before the solidification process such assintering. Otherwise, the particles cannot fuse or sinter each other. The cappingagent removal usually requires heating the printed pattern to evaporate solvent andinduce sintering. Wakuda et al. recently demonstrated room temperature sinteringAg nanoparticles that are protected by a dispersant in an air atmosphere, where thedodecylamine dispersant was removed in methanol and the Ag nanoparticles weredensely sintered. The sintered Ag wires showed low resistivity of 7.3×10–5� cm,which can be used for inkjet printing [50].

3 Recent Advances of Nano-ACA/ACF

3.1 Low-Temperature Sintering of Nano-Ag-Filled ACA/ACF

One of the concerns for ACA/ACF is the higher joint resistance since interconnec-tion using ACA/ACF relies on mechanical contact, unlike the eutectic solder, whichforms the metallurgical bonding during reflow. An approach to minimize the jointresistance of ACA/ACF is to make the conductive fillers fuse each other and formmetallic joints such as metal solder joints. However, to fuse metal fillers in polymersdoes not appear feasible, since a typical organic printed circuit board (Tg∼125◦C),on which the metal filled polymer is applied, cannot withstand such a high tempera-ture; the melting temperature (Tm) of Ag, for example, is around 960◦C. Studieshave showed that Tm and sintering temperatures of materials could be dramati-cally reduced by decreasing the size of the materials [51–53]. It has been reportedthat the surface premelting could be a primary mechanism of the Tm depression of


the fine nanoparticles (<50 nm). For nano-sized particles, sintering behavior couldoccur at much lower temperatures, as such, the use of the fine metal particles inACAs would be promising for high electrical performance of ACA joints by elim-inating the interface between metal fillers. The application of nano-sized particlescan also increase the number of conductive fillers on each bond pad and result inmore contact area between fillers and bond pads as illustrated in Fig. 5. For the sin-tering reaction in a certain material system, temperature and duration are the mostimportant parameters, in particular, the sintering temperature.

(a)

(b)

Circuit substrate

Polymer binder

Electronic Conductive filler

Fig. 5 Schematic illustrations of ACA/ACF with (a) micro-sized conductive fillers and (b) nano-sized conductive fillers

Current–resistance (I–R) relationship of the nano-Ag-filled ACA joint is shownin Fig. 6. As can be seen from the figure, with increasing curing temperatures, theresistance of the ACA joints decreased significantly, from 10–3 to 5×10–5�. Also,ACAs cured at higher temperature exhibited higher current-carrying capability thanACAs cured at low temperature. This phenomenon suggested that more sintering ofnano-Ag particles and subsequently superior interfacial properties between fillersand metal bond pads were achieved at higher temperatures [53], yet the x–y directionof the ACF maintains an excellent insulation property for electrical insulation.

3.2 Self-Assembled Molecular Wires for Nano-ACA/ACF

To enhance the electrical performance of ACA/ACF materials, self-assembledmolecular wires (SAMW) have been introduced into the interface between metalfillers and metal-finished bond pad of ACAs [54, 55]. These organic moleculesadhere to the metal surface and form physico-chemical bonds, which allow elec-trons to flow, as such, it reduces electrical resistance and enables a high currentflow. The unique electrical properties are due to their tuning of metal work func-tions by those organic monolayers. The metal surfaces can be chemically modified

34 Y. Li et al.

I-R curve of 20nm-Ag filled ACACurrent (mA)

1.00E-03

1.00E-05

1.00E-04

1.00E-020

Res

ista

nce

(Oh

m)

150C

210C

220C

4000300020001000

Fig. 6 Current–resistance (I–R) relationship of nano-Ag-filled ACAs with different curingtemperatures [53]

by the organic monolayer and the reduced work functions can be achieved by usingsuitable organic monolayer coatings. An important consideration when examiningthe advantages of organic monolayers pertains to the affinity of organic compoundsto specific metal surfaces. Table 2 gives the examples of molecules preferred formaximum interactions with specific metal finishes; although only molecules withsymmetrical functionalities for both head and tail groups are shown, molecules andderivatives with different head and tail functional groups are possible for interfacesconcerning different metal surfaces.

Different organic molecular wires, dicarboxylic acid, and dithiol have beenintroduced into ACA/ACF joints. For SAM-incorporated ACA with micron-sizedgold/polymer or gold/nickel fillers, lower joint resistance and higher maximumallowable current (highest current applied without inducing joint failure) wasachieved for low-temperature curable ACA (<100◦C). For high curing tempera-ture ACA (150◦C), however, the improvement was not as significant as low curingtemperature ACAs, due to the partial desorption/degradation of organic monolayercoating at the relatively high temperature [56]. However, when dicarboxylic acidor dithiol was introduced into the interface of nano-silver-filled ACAs, signifi-cantly improved electrical properties could be achieved at high-temperature curableACA/ACF, suggesting the coated molecular wires did not suffer degradation on sil-ver nanoparticles at the curing temperature (Fig. 7). The enhanced bonding couldattribute to the larger surface area and higher surface energy of nanoparticles, whichenabled the monolayers to be more readily coated and relatively thermally stable onthe metal surfaces [57].


Table 2 Potential organic monolayer interfacial modifiers for different metal finishes

Formula Compound Metal finishes

R-S-H Thiols Au, Ag, Cu, Pt, ZnR-COOH Carboxylates Fe/FexOy, Ti/TiO2, Ni, Al, AgR-C≡N Cyanides Au, Ag, PtR-N=C=O Isocyanates Pt, Pd, Rh, Ru

N

NH

Imidazole and triazole derivatives CuR-SiOH Organosilane derivatives SiO2, Al2O3, quartz, glass, mica, GeO2, etc.

R denotes alky and aromatic groups

0 500 1000 1500 2000 2500 3000 3500 4000 4500

10–1

10–2

Current (mA)

conventional ACA nano Ag ACA nano Ag ACA dithiol nano Ag ACA di-acid

Join

t res

ista

nce

(Ω),

100

× 1

00 μ

m2

10–3

10–4

10–5

100

Fig. 7 Electrical properties of nano-Ag-filled ACA with dithol or dicarboxylic acid [57]

3.3 Silver Migration Control in Nano-silver-Filled ACA

Silver is the most widely used conductive fillers in ICAs and exhibits exciting poten-tials in nano-ACA/ACF due to many unique advantages of silver. Silver has thehighest room temperature electrical and thermal conductivity among all the con-ductive metals. Silver is also unique among all the cost-effective metals by natureof its conductive oxide (Ag2O). In addition, silver nanoparticles are relatively easyto be formed into different sizes (a few nanometers to 100 nm) and shapes (suchas spheres, rods, wires, disks, flakes) and well dispersed in a variety of polymericmatrix materials. Also the low-temperature sintering makes silver one of the promis-ing candidate as conductive fillers in nano-ACA/ACF. However, silver migrationhas long been a reliability concern in electronic industry. Metal migration is an

36 Y. Li et al.

electrochemical process, whereby, metal (e.g.,silver), in contact with an insulatingmaterial, in a humid environment and under an applied electric field, leaves its initiallocation in ionic form and deposits at another location [58]. It is considered that athreshold voltage exists above which the migration starts. Such migration may leadto a reduction in electrical spacing or cause a short circuit between interconnections.The migration process begins when a thin continuous film of water forms on an insu-lating material between oppositely charged electrodes. When a potential is appliedacross the electrodes, a chemical reaction takes place at the positively biased elec-trode where positive metal ions are formed. These ions, through ionic conduction,migrate toward the negatively charged cathode and over time, they accumulate toform metallic dendrites. As the dendrite growth increases, a reduction of electricalspacing occurs. Eventually, the dendrite silver growth reaches the anode and createsa metal bridge between the electrodes, resulting in an electrical short circuit [59].

Although other metals may also migrate under specific environment, silver ismore susceptible to migration, mainly due to the high solubility of silver ion, lowactivation energy for silver migration, high tendency to form dendrite shape, andlow possibility to form a stable passivation oxide layer [60–62]. The rate of silvermigration is increased by (1) an increase in the applied potential; (2) an increase inthe time of the applied potentials; (3) an increase in the level of relative humidity; (4)an increase in the presence of ionic and hydroscopic contaminants on the surface ofthe substrate; and (5) a decrease in the distance between electrodes of the oppositepolarity.

To reduce silver migration and improve the reliability, several methods have beenreported. The methods include (1) alloying the silver with an anodically stable metalsuch as palladium [61] or platinum [63] or even tin [64]; (2) using hydrophobic coat-ing over the PWB to shield its surface from humidity and ionic contamination [65],since water and contaminates can act as a transport medium and increase the rateof migration; (3) plating of silver with metals such as tin, nickel, or gold, to protectthe silver filler and reduce migration; (4) coating the substrate with polymer [66];(5) applying benzotriazole (BTA) and its derivatives [67]; (6) employing siloxaneepoxy polymers as diffusion barriers due to the excellent adhesion of siloxane epoxypolymers to conductive metals [68]; (7) chelating silver fillers in ECAs with molec-ular monolayers [69]. As an example shown in Fig. 8 [70], with carboxylic acidsand forming chelating compounds with silver ions, the silver migration behavior(leakage current) could be significantly reduced and controlled.

3.4 ACF with Straight-Chain-Like Nickel Nanoparticles

Sumitomo Electric recently developed a new concept ACF using nickel nanopar-ticles with a straight-chain-like structure as conductive fillers [72]. They appliedthe formulated straight-chain-like nickel nanoparticles and solvent in a mixture ofepoxy resin on a substrate film. Then the particles were made to orient towardthe vertical direction of the film surface and fixed in a resin by evaporating thesolvent, and this approach is similar to the elastomeric conductive polymer inter-connects (ECPI), dispersed Ni-coated glass spheres in a silicone matrix, which has


0 1 2 3 4 5-500

0

500

1000

1500

2000

2500

3000

3500

4000

Leak

age

curr

ent (

nA)

Applied voltage (V)

untreated diacid-treated monoacid-treated linear fit

linear fit

(a)

0 100 200 300 400 500

0

200

400

600

800

1000

Leak

age

curr

ent (

μA)

Applied voltage (V)

untreated diacid-treated monoacid-treated

3-order polynomial fit

linear fit

(b)

(b)

Fig. 8 Leakagecurrent–voltage relationshipof nano-Ag conductiveadhesives at (a) low voltagesand (b) high voltages [71]

been developed by AT&T Bell Laboratory [73]. In the estimation using 30 μm-pitch IC chips and glass substrates (the area of Au bumps was 2000 μm2, thedistance of space between neighboring bumps was 10 μm), the new ACF showedexcellent reliability of electrical connection after high-temperature, high-humidity(60◦C/90%RH) test and thermal cycle test (between –40 and 85◦C). The sampleswere also exposed to high-temperature, high-humidity (60◦C/90%RH) for insu-lation ability estimation. Although the distance between two electrodes was only10 μm, ion migration did not occur and insulation resistance has been maintained atover 1 G� for 500 h. This result showed that the new ACF has superior insulationreliability and has potential to be applied in very fine interconnections.

38 Y. Li et al.

3.5 Nanowire ACF for Ultra-fine-pitch Flip-Chip Interconnection

To satisfy the reduced I/O pitch and avoid electric shorting, a possible solutionis to use high-aspect-ratio metal post. Nanowires exhibit high possibilities due tothe small size and extremely high aspect ratio. In literature, nanowires could beapplied in FET (field effective transistor) sensor for gas detection, magnetic hard-disk, nanoelectrodes for electrochemical sensor, thermal-electric device for thermaldissipation and temperature control, etc. [71,74,75]. To prepare nanowires, it isimportant to define nanostructures on the photoresist. Many expensive methodssuch as e-beam, X-ray, or scanning probe lithography have been used but the lengthof nanowires cannot be achieved to micro-meter size. A less expensive alternativeis electrodeposition of metal into nano-porous template such as anodic aluminumoxide (AAO) [76] or block-copolymer self-assembly template. The disadvantagesof block-copolymer template include thin thickness (that means short nanowires),nonuniform distribution and poor parallelism of nano-pores. However, AAO hasbenefits of higher thickness (>10 μm), uniform pore size and density, larger size,and very parallel pores. Lin et al. [77] developed a new ACF with nanowires. Theyused AAO templates to obtain silver and cobalt nanowires array by electrodepo-sition. And then low viscous polyimide (PI) was spread over and filled into thegaps of nanowires array after surface treatment. The bi-metallic Ag/Co nanowirescould remain parallel during fabrication by magnetic interaction between cobaltand applied magnetic field. The silver and cobalt nanowires/polyimide compositefilms could be obtained with a diameter of about 200 nm for nanowire and maxi-mum film thickness up to 50 μm. The X–Y insulation resistance is about 4–6 G�and Z-direction resistance including the trace resistance (3 mm length) is less than0.2�. They also demonstrated the evaluation of this nanowire composite film bystress simulation. They found that the most important factor for designing nanowireACF was the volume ratio of nanowires. However, actually the ratio of nanowirescannot be too small to influence the electric conductance. They concluded that itis important to get a balance between electric conductance and thermal-mechanicalperformance by increasing film thickness or decreasing modulus of polymer matrix.

3.6 An In Situ Formation of Nano-conductive Fillers in ACA/ACF

One of the challenging issues in the formation of nano-filler ACA/ACF is thedispersion of nano-conductive fillers in ACA/ACF. A lot of research has beengoing on in recent years to address the dispersion issue of nano-composite becausenano-fillers due to their high surface reactivity and high electrostatic force tendto agglomerate. For the fine-pitch electronic interconnects using nano-ACA/ACF,the dispersion issues need to be solved. The efforts usually include the physi-cal approaches such as sonication and chemical approaches such as surfactants.Recently, a novel ACA/ACF incorporated with in-situ formed nano-conductiveparticles is proposed for the next generation high-performance fine-pitch elec-tronic packaging applications [78, 79]. This novel interconnect adhesive combines


the electrical conduction along the z-direction (ACA-like) and the ultra fine-pitch (<100 nm) capability. Instead of adding the nano-conductive fillers in theresin, the nanoparticles can be in situ formed during the curing/assembly pro-cess. By using in situ formation via chemical reduction of nanoparticles, duringthe polymer curing process, the filler concentration and dispersion could be con-trolled and the drawback of surface oxidation of the nano-fillers could be easilyovercome [80].

3.7 CNT-Based Conductive Nanocomposites for Transparent,Conductive, and Flexible Electronics

The electrically conductive and optically transparent coating is an upcoming poten-tial for various applications such as transistors [81–84], diodes [85], sensors [86,87], optical modulators [88] or as conductive backbone for electrochemical poly-mer coating [89], smart windows [90], photovoltaics [91], solar cells [92,93] andorganic light-emitting diode (OLED) [94, 95].

With the popularity of flex circuits or substrates, in particular for large-scaleflexible flat panel displays, there are demands for flexible interconnect materi-als, especially under highly mechanical bending conditions of printed electronics.ITO (indium tin oxide), fluorine tin oxide (FTO), and conductive polymers suchas polyaniline (PANI) have been de facto options as transparent and conductivecoating materials. However, those materials have no flexibility in nature, which iscritical for flexible devices such as bendable displays and sensors. Thus, materi-als with mechanical flexibility as well as electrical conductivity and transparencycan find various applications. The CNT incorporated polymer composite is oneof the candidates because its tremendously high aspect ratio can provide a sig-nificantly low percolation threshold in electrical conduction. Its tiny size (<10 nmin diameter (SWCNT), smaller than 1/4λ of visible light) lets light pass throughand the polymer matrix renders flexibility. Therefore, intensive efforts to material-ize transparent, conductive, and flexible CNT-based polymer composite have beendedicated.

Kaempgen et al. found that single wall CNTs (SWCNTs) are more suitable fortransparent conductive coatings than multi wall CNTs (MWCNTs) [96]. This islikely due to a significantly higher diameter of each MWCNT, which increases thelight absorption (Fig. 9). The longer CNTs increase the conductivity for the sametransparency with the decreased number of contacts because the electrical trans-port is dominated by contact resistances between the CNTs [97–101]. The uniformdispersion is a key factor to manufacture reproducible coatings or composite films.

Some CNT-based polymer composites already met the requirements for vari-ous applications. Figure 10 compares the resistivity (1 k�/sq) of the CNT coatingwith 90% light transmittance to those of various target applications [96], where itcan be seen that currently 90% transparent CNT coating can meet the requirementsfor many applications such as touch screen and electrostatic discharge, although

40 Y. Li et al.

0 15 30 45 60 75 90

0

1

2

3

4

5

6 CVD III ~50μm CVD III ~100μm CVD I HiPco SWCNT ITO

Res

istiv

ity [k

Ω]

Transmission [%]

Fig. 9 Transmittance vs.resistivity plot for sprayedCNT layers and ITO [96]

EMA FPD TS BR ESA1

10

1000

100

100000

Ele

ctro

stat

ic D

issi

patio

n

Cat

hode

Tub

es

Tou

ch S

cree

n

EM

I Shi

eldi

ng

Req

uire

d R

esiti

vity

[Ω/s

q]

Applications forTransparent Conductive Coatings

Dis

play

s

1000000

onPET

Fig. 10 Required surface resistivity in typical applications for transparent conductive coatingscompared to the performance of thin conductive CNT composites [96]

only shielding of electromagnetic interference (EMI) is out of reach at least fortransparent coatings. There are trade-off values between transparency and surfaceresistance of the CNT composites with regard to their applications. Recently, a CNTfilm on polymer or glass with 80% light transmittance exhibited a resistance of20�/sq [102]. With extensive research and development of CNT nanocomposite,more reliable conductive and transparent adhesives will find versatile applicationsin microelectronics.


4 Concluding Remarks

In this chapter, recent advances in nano-electrically conductive adhesives (ECAs)were reviewed and in particular, innovative approaches by using cutting-edge nan-otechnologies were presented. The nanotechnology-based ECAs will be playinga very important role in future nanoelectronics, nanophotonics, nano-biomedicaldevices etc.

Although Ag filler is known as one of best conductive materials for ECA, itshigh cost and intrinsic propensity for silver migration are of concern for low-costconsumer products or high-performance device-packaging interconnects. As such,copper will be a silver alternative as ECA filler, as copper is the second highest elec-trically conductive element after silver and has less migration tendency. However,copper is very vulnerable to oxygen and corrosion, and unlike Ag, its oxide is apoor electrical conductor. Therefore, in order to achieve all the benefits from cop-per, much research effort on preventing copper from oxidation/corrosion is neededand the use of SAM/surfactant as mentioned in Table 2 may provide some solutions.

Tremendous evolution in display technologies has been demonstrated overthe past decades, and future display technologies need very flexible and high-performance interconnect materials for nano- or bio-medical device packaging.In general, bending or physical deformation of ECAs can degrade the electricalconductivity. Therefore, novel flexible electrically conductive materials to respondto repeated bending or high elongations will be required for future flexible dis-play technologies. In addition, transparent conductive adhesive is needed for someapplications such as organic LED (OLED) application. Ouyang et al. [103] havedemonstrated a transparent conducting polymer glue and the feasibility of fab-ricating organic electronic devices through a lamination process by using thePEDOT:PSS electric glue. By combining this technology with the large-area contin-uous coating process of organic films, it is possible to fabricate large-area organicelectronic devices through a continuous roll-to-roll coating process. This electricglue made from a conducting polymer provides new applications for conductingpolymers; it will revolutionize the fabrication process of flexible electronic devices.

Although some researches on enhancing current-carrying capability of ECAshave been carried out, there is no commercial ECA material to replace metal-lic solders yet. Thus, increasing current-delivering capability of the ECA will bea continuing research effort. Employing molecular wires in the ECA joints orconductive elements and manipulating conductive joint work functions with highthermal stability, are some of the promising approaches for the high-current ECAchallenges.

With particle size of ECAs becoming smaller in particular for nano-ACAs, thereis serious implication from the point of view of the manufacturing process. WhenACAs are used to attach the metal pads on two mating surfaces, the warpage andcoplanarity of these surfaces need to be carefully controlled to ensure all the metalpads on the mating surfaces are making contact to the conductive spheres in the ACAsimultaneously. Large and compliant (deformable) conductive particles will ease themanufacturing process because these particles can accommodate the warpage and

42 Y. Li et al.

noncoplanarity. With conductive particles becoming even smaller in nano-ACAs,the manufacturing will become more challenging.

The ECAs will be continuing to evolve to meet future technological needs, assuch, much attention is needed on the breakthrough technology developments byengineers and scientists including materials scientists, chemists and physicists, etc.We all should bear in mind that high-performance interconnection materials foradvanced nanoelectronics, photonics, bio-medical devices, and systems packagingare the key for future microelectronics.

References

1. Wolfson H. and Elliot G., Electrically conducting cements containing epoxy resins andsilver. US Patent 2,774,747, 1956.

2. Matz K.R., Electrically conductive cement and brush shunt containing the same. US Patent2,849,631, 1958.

3. Beck D.P., Printed electrical resistors. US Patent 2,866,057, 1958.4. Li Y., Moon K., and Wong C.P., Science 2005; 308: 1419–1420.5. Li Y. and Wong C.P., Mater. Sci. Eng. R 2006; 51: 1–35.6. Liu J., (Ed.), Conductive adhesives for Electronics Packaging. Electrochemical Publications

Ltd., Port Erin, Isle of Man, 1999; Chapter 1.7. Hwang J.S. (Ed.), Environment-Friendly Electronics: Lead-free Technology.

Electrochemical Publications Ltd., Port Erin, Isle of Man, 2001; Chapter 1, pp. 4–10.8. Murray C.T., Rudman R.L., Sabade M.B., and Pocius A.V., Mater. Res. Bull. 2003; 28:

449–454.9. Lau J., Wong C.P., Lee N.C., and Lee S.W.R., In: Electronics Manufacturing: with

Lead-Free, Halogen-Free, and Conductive-Adhesive Materials. McGraw-Hill, New York,2002.

10. Lu D., Tong Q.K., and Wong C.P., IEEE Trans. Compon. Packag. Manuf. Technol., Part C1999; 22(3): 228–232.

11. Persson K., Nylund A., Liu J., and Olefjord I., Proceedings of the 7th European Conferenceon Applications of Surface and Interface Analysis, Gothenburg, June 1997.

12. Lu D. and Wong C.P., J. Appl. Polym. Sci. 1999; 74: 399–406.13. Li Y., Moon K., and Wong C.P., J. Adhes. Sci. Technol. 2005; 19(16): 1427–1444.14. Matienzo L.J., Egitto F.D., and Logan P.E., J. Mater. Sci. 2003; 38: 4831–4842.15. Li H., Moon K., and Wong C.P., J. Electron. Mater. 2004; 33: 106–113.16. Lu D., Tong Q.K., and Wong C.P., IEEE Trans. Compon. Packag. Manuf. Technol. Part C

1999; 22: 223–227.17. Li Y., Moon K., and Wong C.P., IEEE Trans. Compon. Packag. Manuf. Technol. 2006; 29(1):

173–178.18. Li Y., Moon K., Whitman A., and Wong C.P., IEEE Trans. Compon. Packag. Manuf.

Technol. 2006; 29(4): 758–763.19. Gallagher C., Matijasevic G., and Maguire J.F., Proceedings of 47th IEEE Electronic

Components and Technology Conference 1997; 554–560.20. Roman J.W. and Eagar T.W., Proceedings of the International Society for Hybrids and

Microelectronics Society, San Francisco, CA, 1992; 52.21. Li Y., Moon K., Li H., and Wong C.P., Proceedings of 54th IEEE Electronic Components

and Technology Conference, Las Vegas, Nevada, June 1–4, 2004; 1959–1964.22. Cavasin D., Brice-Heams K., and Arab A. Proceedings of 53rd Electronic Components and

Technology Conference 2003; 1404–1407.23. Kisiel R., J. Electron. Packag. 2002; 124: 367.


24. de Vries H., van Delft J., and Slob K., IEEE Trans. Compon. Packag. Manuf. Technol. 2005;28: 499.

25. Watanabe I., Gotoh Y., and Kobayashi K. Proc. Asia Display/IDW, Nagoya, Japan 2001;553–556.

26. Nishida H., Sakamoto K., and Ogawa H., IBM J. Res. Dev. 1998; 42: 517.27. Williams D.J., Whalley D.C., Boyle O.A., and Ogunjimi A.O., Solder. Surf. Mount Tech.

1993; 5: 4.28. Liu J., Tolvgard A., Malmodin J., and Lai Z., IEEE Trans. Compon. Packag. Manuf. Technol.

1999; 22: 186.29. Clot P., Zeberli J.F., Chenuz J.M., Ferrando F., and Styblo D., Proceedings of the

International Electronics Manufacturing Technology Symposium 24th IEEE/CPMT, Austin,TX, 1999; 36.

30. Wu H., Wu X., Liu J., Zhang G., Wang Y., Zeng Y., and Jing J., Development of a novelisotropic conductive adhesive filled with silver nanowires. J. Compos. Mater. 2006; 40(21):1961–1968.

31. Chen C., Wang L., Li R., Jiang G., Yu H., and Chen T., J. Mater. Sci. 2007; 42(9): 3172.32. Lee H.S., Chou K.S., and Shih Z.W., Effect of nano-sized silver particles on the resistivity

of polymeric conductive adhesives. Int. J. Adhes. Adhes. 2005; 25: 437–441.33. Ye L., Lai Z., Liu J., and Tholen A., Effect of Ag particle size on electrical conductivity

of isotropically conductive adhesives. IEEE Trans. Electron. Packag. Manuf. 1999; 22(4):299–302.

34. Fan L., Su B., Qu J., and Wong C.P., Electrical and thermal conductivities of polymercomposites containing nano-sized particles. Proceedings of Electronic Components andTechnology Conference. IEEE, NJ, 2004; 148–154.

35. Jiang H.J., Moon K., Lu J., and Wong C.P., J. Electron. Mater. 2005; 34: 1432–1439.36. Jiang H.J., Moon K., Li Y., and Wong C.P., Surface functionalized silver nanoparticles for

ultrahigh conductive polymer composites. Chem. Mater. 2006; 18(13): 2969–2973.37. Kotthaus S., Günther B.H., Haug R., and Schafer H., Study of isotropically conductive bond-

ings filled with aggregates of nano-sized Ag-particles. IEEE Trans. Compon. Packag. Manuf.Technol. Part A 1997; 20(1): 15–20.

38. Das R.N., Lauffer J.M., and Egitto F.D., Proceedings of Electronic Components andTechnology Conference. IEEE, 2006; 112–118.

39. Iijima S., Nature 1991; 354: 56.40. Berber S., Kwon Y.K., and Tomànek D., Unusually high thermal conductivity of carbon

nanotubes. Phys. Rev. Lett. 2000; 84(20): 4613–4616.41. Yu M.F., Files B.S., Arepalli S., and Ruoff R.S., Tensile loading of ropes of single

wall carbon nanotubes and their mechanical properties. Phys. Rev. Lett. 2000; 84(24):5552–5555.

42. Li J. and Lumpp J.K., Electrical and mechanical characterization of carbon nanotube filledconductive adhesive. In: Proceedings of Aerospace Conference. IEEE, NJ, 2006; 1–6.

43. Qian D., Dickey E.C., Andrews R., and Rantell T., Load transfer and deformation mecha-nisms in carbon nanotube polystyrene composites. Appl. Phys. Lett. 2000; 76: 2868.

44. Lin X.C. and Lin F., Improvement on the properties of silver-containing conductive adhe-sives by the addition of carbon nanotube. In: Proceedings of High Density MicrosystemDesign and Packaging. IEEE, NJ, 2004; 382–384.

45. Rutkofsky M., Banash M., Rajagopal R., and Chen J., Using a carbon nanotube addi-tive to make electrically conductive commercial polymer composites. Zyvex CorporationApplication Note 9709: http://www.zyvex.com/Documents/9709.PDF 28; 2006.

46. Kamyshny A., Ben-Moshe M., Aviezer S., and Magdassi S., Ink-jet printing of metallicnanoparticles and microemulsions. Macromol. Rapid Commun. 2005; 26: 281–288.

47. Cibis D. and Currle U., Inkjet printing of conductive silver paths. In: 2nd InternationalWorkshop on Inkjet Printing of Functional Polymers and Materials. Eindhoven, TheNetherlands, 2005.

44 Y. Li et al.

48. Kolbe J., Arp A., Calderone F., Meyer E.M., Meyer W., Schaefer H., and Stuve M.,Inkjettable conductive adhesive for use in microelectronics and microsystems technology.In: Proceedings of IEEE Polytronic Conference. IEEE, NJ, 2005; 1–4.

49. Moscicki A., Felba J., Sobierajski T., Kudzia J., Arp A., and Meyer W., Electrically conduc-tive formulations filled nano size silver filler for ink-jet technology. In: Proceedings of IEEEPolytronic Conference. IEEE, NJ, 2005; 40–44.

50. Wakuda D., Hatamura M., and Suganuma K., Chem. Phys. Lett. 2007; 441: 305–308.51. Moon K., Dong H., Maric R., Pothukuchi S., Hunt A., Li Y., and Wong C.P., J. Electron.

Mater. 2005; 34: 132–139.52. Matsuba Y., Erekutoronikusu Jisso Gakkaishi 2003; 6(2): 130–135.53. Li Y., Moon K. and Wong C.P., J. Appl. Polym. Sci. 2006; 99(4): 1665–1673.54. Li Y., Moon K., and Wong C.P. In: Proceedings of 54th IEEE Electronic Components and

Technology Conference. IEEE, NJ, 2004; 1968–1974.55. Li Y. and Wong C.P. In: Proceedings of 55th IEEE Electronic Components and Technology

Conference. IEEE, NJ, 2005; 1147–1154.56. Li Y., Moon K., and Wong C.P., J. Electron. Mater. 2005; 34(3): 266–271.57. Li Y., Moon K., and Wong C.P., J. Electron. Mater. 2006; 34(12): 1573–1578.58. Davies G. and Sandstrom J. Circuits Manuf. 1976; 56–62.59. Harsanyi G. and Ripka G., Electrocomp. Sci. Technol. 1985; 11: 281–290.60. Giacomo G.A., In: J. McHardy and F. Ludwig (Eds.) Electrochemistry of Semiconductors

and Electronics: Processes and Devices. Noyes Publications, Park Ridge, NJ, 1992;pp. 255–295.

61. Manepalli R., Stepniak F., Bidstrup-Allen S.A., and Kohl P.A., IEEE Trans. Adv. Packag.1999; 22: 4–8.

62. Giacomo D., In: Reliability of Electronic Packages and Semiconductor Devices. McGraw-Hill, New York, 1997; Chapter 9.

63. Wassink R., Hybrid. Circ. 1987; 13: 9–13.64. Shirai Y., Komagata M., and Suzuki K. In: 1st International IEEE Conference on Polymers

and Adhesives in Microelectronics and Photonics. IEEE, NJ, 2001; 79–83.65. Der Marderosian A. Ratheon Co. Equipment Division, Equipment Development

Laboratories 1978; 134–141.66. Schonhorn H. and Sharpe L.H., Prevention of surface mass migration by a polymeric surface

coating. US Patent 4377619, 1983.67. Brusic V., Frankel G.S., Roldan J., and Saraf R., J. Electrochem. Soc. 1995; 142: 2591–2594.68. Wang P.I., Lu T.M., Murarka S.P., and Ghoshal R., US Pending Patent 20050236711, 2005.69. Li Y. and Wong C.P. US Patent Pending, 2006.70. Li Y. and Wong C.P., Monolayer-protection for eletrochemical migration control in silver

nanocomposite. Appl. Phys. Lett. 2006; 81: 112112.71. Prinz G.A., Science 1998; 282: 1660.72. Toshioka H., Kobayashi M., Koyama K., Nakatsugi K., Kuwabara T., Yamamoto M., and

Kashihara H., SEI Tech. Rev. 2006; 62: 58–61.73. Chang D.C. and Smith E.L., US patent #5206585, 1993.74. Lieber C.M., Nanowire nanosensors for high sensitive and selective detection of biological

and chemical species. Science 2001; 293: 1289–1292.75. Martin C.R. and Menon V.P., Fabrication and evaluation of nanoelectrode ensembles. Anal.

Chem 1995; 67: 1920–1928.76. Xu J.M., Appl. Phys. Lett. 2001; 79: 1039–1041.77. Russell T.P., Ultra-high density nanowire array grown in self-assembled di-block copolymer

template. Science 2000; 290: 2126–2129.78. Li Y., Moon K., and Wong C.P., In: Proceedings of 56th IEEE Electronic Components and

Technology Conference. IEEE, NJ, 2006; 1239–1245.79. Li Y., Zhang Z., Moon K., and Wong C.P., Ultra-fine pitch wafer level ACF (anisotropic

conductive film) interconnect by in-situ formation of nano fillers with high current carryingcapability. US Patent Pending, 2006.


80. Moon K., Jiang H., Zhang R., and Wong C., In-situ formed nanoparticles in polymer matrixfor low pressure bonding, GTRC Invention Disclosure No. 4443, 2008.

81. Snow E.S., Novak J.P., Lay M.D., Houser E.H., Perkins F.K., and Campbell P.M., J. Vac.Sci. Technol. B 2004; 22(4): 1990.

82. Lay M.D., Novak J.P., and Snow E.S., Nano Lett. 2004; 4(4): 603.83. Shiraishi M., Takenobu T., Iwai T., Iwasa Y., Kataura H., and Ata M., Chem. Phys. Lett.

2004; 394: 110.84. Meitl M.A., Zhou Y., Gaur A., Jeon S., Usrey M.L., Strano M.S., and Rogers J.A., Nano

Lett. 2004; 4(9): 1643.85. Zhou Y., Gaur A., Hur S., Kocabas C., Meitl M.A., Shim M., and Rogers J.A., Nano Lett.

2004; 4(10): 2031.86. Li Z., Dharap P., Nagarajaiah S., Barrera E.V., and Kim J.D., J. Adv. Mat. 2004; 16: 640.87. Abraham J.K., Philip B., Witchurch A., Varadan V.K., and Reddy C.C., Smart Mater. Struct.

2004; 1045: 13.88. Wu Z., Chen Z., Du X., Logan J.M., Sippel J., Nikolou M., Kamaras K., Reynolds J.R.,

Tanner D.B., Hebard A.F., and Rinzler A.G., Science 2004; 305: 1273.89. Ferrer-Anglada N., Kaempgen M., Skákalová V., Dettlaf-Weglikowska U., and Roth S.,

Diam. Relat. Mater. 2004; 13(2): 256.90. Hu L., Zhoa Y-L., Ryu K., Zhou C., Stoddart J.F., and Grüner G. Advanced Materials 2008;

20: 5939–5946.91. Gruner G., J. Mater. Chem. 2006; 16: 3533–3539.92. Ago H., Petritsch K., Shaffer M.S.P., Windle A.H., and Friend R.H., Adv. Mater. 1999; 11:

1281.93. Du Pasquier A., Unalan H.E., Kanwal A., Miller S., and Chhowalla M., Appl. Phys. Lett.

2005; 87: 203511.94. Shan B. and Cho K., Phys. Rev. Lett. 2005; 94: 236602.95. Mitschke U. and Bauerle P., J. Mater. Chem. 2000; 10: 1471.96. Kaempgen M., Duesberg G.S., and Roth S., Appl. Surf. Sci. 2005; 252: 425–429.97. Kaiser A.B., Dusberg G., and Roth S., Phys. Rev. B 1998; 57: 1418.98. Liu K., Roth S., Duesberg G.S., Kim G.-T., and Schmid M., Progress in Molecular

Nanostructures, H. Kuzmany, J. Fink, M. Mehring, S. Roth (Eds.). American Institute ofPhysics, New York, 1998; AIP Conference Proceedings 442, 61–64.

99. Shirashi M., Synth. Met. 2002; 9198: 1.100. Duesberg G.S., Graham A.P., Kreupl F., Liebau M., Seidel R., Under E., and Hönlein W.,

Diam. Rel. Mater. 2004; 13: 354.101. Lu K.L., Lago R.M., Chen Y.K., Green M.L.H., Harris P.J.E., and Tsang S.C., Carbon 1996;

34: 814.102. Trancik J.E., Barton S.C., and Hone J., Nano Lett. 2008; 8(4): 982–987.103. Ouyang H.Y. and Yang Y., Adv. Mater. 2006; 18: 2141–2144.

http://www.springer.com/978-1-4419-0039-5