Referred Journal Papers

Electromigration (EM) is a major failure mode in electronic interconnections, which is commonly made of Ag, Cu, and their alloys. Herein, a single- and two-phase AgCu alloys were subjected with current density of 4 × 105 A/cm2 for in situ microstructural and crystallographic characterizations. Anomalous grain growth and the formation of cord-like grains with an aspect ratio over 1000 and asymmetrical hillock/voids clusters were found. These abnormal morphological evolutions were resulted from the electron flow-induced lattice strains, which was inversely correlated with the lifespan of interconnection. Multi-phase alloys are less resistant to EM due to the unbalanced electron flow-induced strains.

Package-on-Package (PoP) is a popular technology for fabricating chipsets of accelerated processing units. However, the coefficient of thermal expansion mismatch between Si chips and polymer substrates induces thermal warpage during the reflow process. As such, the reflow temperature and reliability of solder joints are critical aspects of PoP. Although Sn58Bi is a good candidate for low-temperature processes, its brittleness causes other reliability issues. In this study, an in-situ observation was performed on composite solders (CSs) made of an Sn-3.0Ag-0.5Cu (SAC305) solder and an Sn58Bi solder that were mixed during reflow at temperatures of 170 °C (CS-170), 180 °C (CS-180), and 190 °C (CS-190). The volumes of the mixed-solder regions were in the order of CS-190 > CS-180 > CS-170. A calculated Sn58Bi-SAC305 isoplethal section of the Sn-Ag-Cu-Bi quaternary system was employed to elucidate the melting behavior of CS-190. CS-190 with Sn cyclic twin boundary and Bi phase preferred orientation of [0001] was demonstrated by electron backscatter diffraction. Moreover, Cu/CS-190/Cu joints with a shear strength of 46 MPa were formed. CS-190 can reduce the brittleness of SnBi solder owing to the reduced solid-solution hardening with the decrease in the Bi content. These factors can improve the reliability of PoP technology.

Lithium titanate (Li4Ti5O12, LTO) has already occupied its niche as an anode material for high-power and long-lifespan lithium batteries, but some novel directions for basic and applied research are still open. One of the most promising approaches in improving its properties, e.g., electronic conductivity and rate capability, is based on controllable defect engineering. The “defects” may be intentionally introduced into LTO via doping, surface modifications, and the synergy between them. However, the defects, which have significant effects to the electrical and electrochemical properties, are usually extremely dilute. Reliable material characterizations are essential and challenging, but the instrumental tools for revealing dilute defects are still insufficient. Herein, detailed analyses on the surface or subsurface defects of carbon-coated LTO were performed using various material characterization methods. Raman spectroscopy has been identified as a unique tool for the probing of structural defects.

The Li4Ti5O12 (LTO) defect spinel is known for its excellent durability of “10,000” cycle counts and high level of safety as an anode material in lithium-ion batteries, but it shows an intrinsic insulating property and poor electrochemical kinetics. Doping is a direct approach to manipulate the electronic conductivity of LTO. However, doping may induce multiple effects influencing the overall electrochemical kinetics, e.g., changing the size of particles and the ionic and electronic conductivities. Here we systematically investigated the phase stability, electronic conductivity, and electrochemical kinetics of M-doped LTO (M = Na, K, Mg, Ca, Sr, Al, and Ga). With both ab initio calculations and experiments, the mechanism of electron transport within LTO is elucidated, the desired type of dopants for improving electronic conductivity of LTO is clarified, and the role of electronic conductivity in the electrochemical kinetics of LTO is revealed. These results provide an in-depth understanding of metal-doped LTO and would help the development of a variety of electrode materials.

The electronic interconnections in the state-of-the-art integrated circuit manufacturing have been scaled down to the micron or sub-micron scale. This results in a dramatic increase in the current density passing through interconnections, which means that the electromigration (EM) effect plays a significant role in the reliability of products. Although thorough studies and reviews of EM effects have been continuously conducted in the past 60 years, some parts of EM theories lack clear elucidation of the electric current-induced non-directional effects, including electric current-induced phase equilibrium changes. This review article is intended to provide a broad picture of electric current-induced lattice stability changes and to summarize the existing literature on EM-related phenomena, EM-related theoretical models, and relevant effects of the electroplastic (EP) effect in order to lead to a better understanding of electric current-induced effects on materials. This article also posits that EM is either part of the EP effect or shares the intrinsic electric current-induced plastic deformation associated with the EP effect. This concept appears to contribute to the missing parts of the EM theories.

In this study, the interfacial reactions, microstructures, and intermetallic compounds (IMCs) of 68In–32Bi, 50In–50Bi, and 33In–67Bi (in weight percent) low melting alloys reacted with Cu substrates were investigated. CuIn2, Cu11In9, and Cu2In IMCs formed at the interface in different In–Bi/Cu systems and were characterized by grazing incidence X-ray diffraction (GIXRD) and electron probe X-ray microanalyzer (EPMA). The CuIn2 and Cu11In9 IMCs formed at the interface phase transfer to Cu11In9 when annealing temperature was greater than 70 °C. CALPHAD-type thermodynamic calculations were performed to illustrate the corresponding diffusion paths. Moreover, shear tests were conducted and indicated that 50In–50Bi solder joint provide the highest shear strength among the three alloys.

The effective charge of an element is a parameter characterizing the electromigration effect, which can determine the reliability of interconnection in electronic technologies. In this work, machine learning approaches were employed to model the effective charge (z*) as a linear function of physically meaningful elemental properties. Average fivefold (leave-out-alloy-group) cross-validation yielded root-mean-square-error divided by whole data set standard deviation (RMSE/σ) values of 0.37 ± 0.01 (0.22 ± 0.18), respectively, and R2 values of 0.86. Extrapolation to z* of totally new alloys showed limited but potentially useful predictive ability. The model was used in predicting z* for technologically relevant host–impurity pairs.

A new Sn-45Bi-2.6Zn (wt%) alloy was developed to replace the eutectic Sn-58Bi alloy as a low-melting point solder alloy. A 112% increase in the tensile elongation (0.68 vs. 0.32 in strain) was obtained by increasing the Sn-to-Bi volume ratio by reducing the volume content of brittle Bi. A solidus temperature of 133 °C was calculated from the Sn–Bi–Zn ternary system. The calculation of phase diagram (CALPHAD) method was performed to help understand the melting behavior of Sn-45Bi-2.6Zn alloy.

The transformation of pyrite into pyrrhotite above 500 °C was observed in the Chelungpu fault zone, which formed as a result of the 1999 Chi-Chi earthquake in Taiwan. Similarly, pyrite transformation to pyrrhotite at approximately 640 °C was observed during the Tohoku earthquake in Japan. In this study, we investigated the high-temperature phase-transition of iron sulfide minerals (greigite) under anaerobic conditions. We simulated mineral phase transformations during fault movement with the aim of determining the temperature of fault slip. The techniques used in this study included thermogravimetry and differential thermal analysis (TG/DTA) and in situ X-ray diffraction (XRD). We found diversification between 520 °C and 630 °C in the TG/DTA curves that signifies the transformation of pyrite into pyrrhotite. Furthermore, the in situ XRD results confirmed the sequence in which greigite underwent phase transitions to gradually transform into pyrite and pyrrhotite at approximately 320 °C. Greigite completely changed into pyrite and pyrrhotite at 450 °C. Finally, pyrite was completely transformed into pyrrhotite at 580 °C. Our results reveal the temperature and sequence in which the phase transitions of greigite occur, and indicate that this may be used to constrain the temperature of fault-slip. This conclusion is supported by field observations made following the Tohoku and Chi-Chi earthquakes.

Steels with good machinability are required for the ever-increasing demand for machining efficiency in industry. The manganese sulfide (MnS) inclusions in lead-free resulfurized free-cutting steels act as stress raisers, lowering the shear strength of steel such that the cutting resistance is reduced. Because the morphology of MnS critically determines the machinability of steels, the reactions involving the formation of MnS during solidification need to be carefully assessed, especially with regard to whether they are eutectic or monotectic reactions. In this paper, we established the relationships between alloying elements and solidified microstructures by utilizing both the CALPHAD-type thermodynamic modeling and high-temperature experiments at 1600 °C. The effects of sulfur content on the MnS microstructure were evaluated based on the solidification paths of the ideal iron-manganese-sulfur ternary system as well as the realistic iron-manganese-sulfur-silicon-carbon quinary system. Moreover, we systematically evaluated the effects of various alloying elements, namely hydrogen, boron, carbon, nitrogen, oxygen, aluminum, silicon, phosphor, argon, vanadium, chromium, cobalt, nickel, copper, arsenic, zirconium, niobium, molybdenum, tin, tantalum, tungsten, on the microstructure of MnS. Among those, oxygen is identified as a super-strong monotectic-stabilizer, and the addition of oxygen can drastically enhance the monotectic-type MnS, which is desirable for free-cutting steels. The thermodynamic predictions agree closely with experiments. With the combined efforts of thermodynamic calculations and high-temperature experiments, the morphology, size, and uniformity of MnS inclusions can be optimized for the development of better free-cutting steels.

By using knowledge of the fundamental laws of physics, we can determine multiple structural, and electronic property relationships that can be used as complementary guides in obtaining improved experimental information. Computational materials design allows us to fabricate optimized materials based on the understanding of key atomistic processes. In this review, we present an up to date summary of the computational approaches using first principles (ab initio) aimed at designing better lithium titanate oxide Li4Ti5O12 as anode material for lithium-ion batteries, and some key challenges and opportunities that lie ahead.

Low-temperature lead (Pb)-free solders are demanding in the electronic packaging industry, because it would open the door for various economic choices of polymeric materials as substrates and also revives the lower cost processes. Here, we proposed a tin–bismuth–indium–gallium (Sn-52.5Bi-2.68In-1Ga, SBIG (in wt.%)) quaternary low-temperature solder, designed based on systematic CALPHAD (CALculation of PHAse Diagram)-type thermodynamic calculations and corresponding key experiments. The solidification behavior of SBIG was carefully elaborated based on the computations using the lever rule and the Scheil model, and the experiments in terms of thermal analyses and microstructures of sample produced with step-quenching and various cooling rates. The mechanical properties of as-cast and 80 °C-annealed SBIG as well as their microstructures and fracture surfaces were evaluated in the tensile tests. The proposed SBIG solder is with a low liquidus temperature of 141.9 °C and is typically composed of the primary (Sn) phase, the (Sn) + (Bi) eutectic structure and a small amount of (Ga) phase. Air cooling has been identified as a satisfactory cooling rate, which would not lead to the formation of the brittle BiIn intermetallic compound. The as-cast SBIG solder exhibited high yield strength (YS) of 43.7 MPa, high ultimate tensile strength (UTS) of 53.3 MPa and an extremely large elongation of 97.3% as comparing to the conventional eutectic Sn-58Bi solder (YS: 43.1 MPa, UTS: 49.5 MPa, and elongation: 37.5%). However, the proposed SBIG solder does not possess qualified thermal stability, that significant degradation in both strength and elongation were observed after being subjected to extensive thermal ageing at 80 °C for 504 h.

Titanium (Ti)-added tin-58 wt% bismuth (Sn58Bi) solder alloys were synthesized in this study. The microstructural evolution and mechanical property change of the eutectic Sn58Bi and Ti added Sn58Bi alloys were studied before and after thermal aging. Ti added Sn58Bi alloys exhibit a considerably refined microstructure compared with the eutectic Sn58Bi alloy because of the presence of the Ti2Sn3 and Ti6Sn5 intermetallic compounds (IMCs). The formation history of these Ti-Sn IMCs was studied based on a thermodynamic calculation. After thermal aging, the yield strength (YS) and ultimate tensile strength (UTS) of the eutectic Sn58Bi decreased, while the Ti added Sn58Bi alloys maintained the highest YS and UTS. Notably, the superior elongation of 0.5 wt% Ti added sample compared with that of the eutectic Sn58Bi alloy was obtained after 1008 h aging. The fracture morphology of the Ti added Sn58Bi alloys almost unchanged during aging due to the stable microstructure. In addition, the electrical resistivities of the eutectic Sn58Bi were tested, while the resistivities decreased in the Ti added Sn58Bi alloys at room temperature.

Copper (Cu)-to-Cu interconnection is crucial in electronic packaging for applications ranging from three-dimensional integrated circuits to die-attachment in wide band-gap devices. Solid-solution phases are innate with better ductility and electrical and thermal conductivity over intermetallic compounds, which are brittle, electrical resistant, and unstable; yet are commonly seen in electronic joints. The desired solid-solution Cu-to-Cu joints have recently been demonstrated using transient-liquid-phase bonding with gallium (Ga) as the filler material to bond Cu substrates with electroplated nickel (Ni) as under-bump-metallurgy. The Cu/Ni/Ga/Ni/Cu couples can fully transform into the Cu/face-centered cubic (fcc)-(Ni,Cu,Ga)/Cu solid-solution joints. However, why and how the solid-solution joints were formed remains unclear. Here we proposed the formation mechanism of the fcc-(Ni,Cu,Ga) solid-solution joints based on five sets of Cu-Ni-Ga couples, namely the Ni/Ga, Ni/Cu/Ga, Ni/Ni/Ga, (Cu,Ni)/Ga, and Cu/Ni/Ni3Ga7/Ni3Ga7/Ni/Cu reactions. The roles of each element in the Cu-Ga-Ni interactions as well as the reaction progressions are elaborated.

The Li4Ti5O12 is applied in lithium ion batteries as anode material, which can be synthesized by various synthesis techniques. In this study, the molten salt synthesis technique at low temperatures, i.e. 350 °C, was applied to synthesize Li4Ti5O12. Surprisingly, the Li4Ti5O12 was not formed according to XRD analysis, which raised question about the stability range of Li4Ti5O12. To investigate the stability of Li4Ti5O12 at low temperatures, the high-temperature calcined Li4Ti5O12 powder was equilibrated in the LiCl-KCl eutectic salt at 350 °C. The result of experiment revealed that the Li4Ti5O12 is not decomposed. Results of ab initio calculations also indicated that the Li4Ti5O12 phase is a stable phase at 0 K. The products of molten salt synthesis technique were then annealed at 900 °C, which resulted in the Li4Ti5O12 formation. It was concluded that the Li4Ti5O12 is a stable phase at low temperatures and the reasons for not forming the Li4Ti5O12 by molten salt technique at low temperature are possibly related to activation energy and kinetic barriers. The Li4Ti5O12 formation energy is also very small, due to the results of ab initio calculations.

Red-light phosphor materials are crucial components in solid-state lighting (SSL) for simulating natural sunlight. Mn-doped Mg2TiO4 is a promising fluoride-free red-emitting phosphor; however, a sensitizer is necessary to enhance its brightness. In this work, we perform ab initio calculations based on the density functional theory (DFT) to systematically examine the electronic-band coupling between the luminescent center, Mn, and several possible sensitizers, Zn, Nb, Mo, In, Sn, and Ta. Nb was identified as the optimal sensitizer. Well-crystallized 0.1 at. % Mn and 0.0–0.7 at. % Nb-codoped Mg2TiO4 were synthesized at 1450 °C. Synchrotron-radiation-based X-ray absorption spectroscopy (XAS) experimentally validated the proposed atomistic structure, indicating that the Nb5+ dopant substitutes Ti4+ at the 16d sites, leading to the formation of Ti vacancies and of a parasitic MgTiO3 phase. Effective sensitization, resulting in a 243% enhancement of the photoluminescence intensity, was achieved. The 0.1 at. % Mn and 0.5 at. % Nb-codoped Mg2TiO4 were obtained as an ultrabright “rare-earth-free” (RE-free) and “fluoride-free” red-light phosphor.

The electromigration (EM) effect involves atomic diffusion of metals under current stressing. Recent theories of EM are based on the unbalanced electrostatic and electron-wind forces exerted on metal ions. However, none of these models have coupled the EM effect and lattice stability. Here, we performed in situ current-stressing experiments for pure Cu strips using synchrotron X-ray diffractometry and scanning electron microscopy and ab initio calculations based on density functional theory. An intrinsic and non-uniform lattice expansion – larger at the cathode and smaller at the anode, is identified induced by the flow of electrons. If this electron flow-induced strain is small, it causes an elastic deformation; while if it is larger than the yield point, diffusion as local stress relaxation will cause the formation of hillocks and voids as well as EM-induced failure. The fundamental driving force for the electromigration effect is elucidated and validated with experiments.

The lithium titanate defect spinel, Li4Ti5O12 (LTO), is a promising “zero‐strain” anode material for lithium‐ion batteries in cycling‐demanding applications. However, the low‐rate capability limits its range of applications. Surface modifications, for example, coating and defect engineering, play an intriguing role in interfacial electrochemical processes. Herein, a novel synthesis of highly oxygen‐deficient “defective‐LTO” anode material with high‐rate performance is reported. It is synthesized using conventional precursors via a one‐pot thermal reduction process. A high level of oxygen vacancies of ≈6.5 at% and conformal amorphous carbon coating are achieved simultaneously, resulting in a compounding effect for a high discharge capacity of 123 mAh g−1 with Coulombic efficiency of 99.8% at 10 C‐rate. Ab initio calculations foresight that oxygen vacancies increase the electron donor density of the neighboring titanium atoms, while conformal amorphous carbon significantly lowers interfacial charge‐transfer resistance. The formation mechanism, as well as the origin of enhanced electrochemical properties, is elaborated in this paper.

Copper (Cu)-to-Cu bonding is an essential process for most advanced electronic devices, such as Cu-pillar joining in flip-chip packaging, through-silicon-via interconnections in three-dimensional integrated circuits, and die attachments in wide band-gap electronics. Here we propose an approach for fabricating face-centered cubic solid-solution joints without formation of intermetallic compounds by using a trace amount of gallium (Ga) and nickel (Ni) under-bump-metallurgy (UBM) for the reactive diffusion bonding at 300 °C. The formation of solid-solution joints involves interactions with the NiCuGa mixture at the bonding interface. A shear-strength of 43.5 MPa was achieved in the as-bonded joint due to the formation of grains across the bonding interface, while it does not degrade after prolonged post-annealing at 300 °C for 200 h. Instead, an even stronger joint with a shear-strength of 48.2 MPa was obtained. We demonstrate the fabrication of strong, thermally stable Cu-to-Cu joints using transient molten Ga and Ni UBM.

The sodiation and desodiation of sodium (Na) into the Super-P carbon anode material were investigated using electrochemical analyses, high-resolution transmission electron microscopy (HRTEM), and neutron powder diffraction (NPD). In the sodiated Super-P carbon, sodium is stored both in the graphite interlayer space of carbon nano-particles and pores between the particles. Sodium metal clusters found in micro-pores between the carbon particles are responsible for the large irreversible capacity of the Super-P electrode. The graphite interlayer distance increases on sodiation from 3.57 Å to two distinct values of ∼3.84 and 4.41 Å. The mechanism of the process is discussed.

Silver (Ag) is one of the seven metals of antiquity and an important engineering material in the electronic, medical, and chemical industries because of its unique noble and catalytic properties. Ag thin films are extensively used in modern electronics primarily because of their oxidation-resistance. Here we report a novel phenomenon of Ag nano-volcanic eruption that is caused by interactions between Ag and oxygen (O). It involves grain boundary liquation, the ejection of transient Ag-O fluids through grain boundaries, and the decomposition of Ag-O fluids into O2 gas and suspended Ag and Ag2O clusters. Subsequent coating with re-deposited Ag-O and the de-alloying of O yield a conformal amorphous Ag coating. Patterned Ag hillock arrays and direct Ag-to-Ag bonding can be formed by the homogenous crystallization of amorphous coatings. The Ag “nano-volcanic eruption” mechanism is elaborated, shedding light on a new mechanism of hillock formation and new applications of amorphous Ag coatings.

Thin-film electroluminescent devices are promising solid-state lighting devices. Red light-emitting phosphor is the key component to be integrated with the well-established blue light-emitting diode chips for stimulating natural sunlight. However, environmentally hazardous rare-earth (RE) dopants, e.g. Eu2+ and Ce2+, are commonly used for red-emitting phosphors. Mg2TiO4 inverse spinel has been reported as a promising matrix material for “RE-free” red light luminescent material. In this paper, Mg2TiO4 inverse spinel is investigated using both experimental and theoretical approaches. The Mg2TiO4 thin films were deposited on Si (100) substrates using either spin-coating with the sol–gel process, or radio frequency sputtering, and annealed at various temperatures ranging from 600°C to 900°C. The crystallinity, microstructures, and photoluminescent properties of the Mg2TiO4 thin films were characterized. In addition, the atomistic model of the Mg2TiO4 inverse spinel was constructed, and the electronic band structure of Mg2TiO4 was calculated based on density functional theory. Essential physical and optoelectronic properties of the Mg2TiO4 luminance material as well as its optimal thin-film processing conditions were comprehensively reported.

Manipulation or detection of domain patterns in ferroelectrics is of great interest due to wide applications for such materials. Among all the simulation methods, q-state Potts model is a common method used to simulate the domain pattern of ferroelectrics. The method, however, expresses the effect of temperature and coupling energy between interacting cells in an implicit way. In this study, we developed a new scheme that can explicitly study the temperature and coupling energy effect on the domain pattern of ferroelectrics. The method combined molecular dynamics (MD) simulations with a modified q-state Potts model. In the traditional q-state Potts model, each state is represented by a constant. Here we propose that the states are a function of temperature. We tested our method with a standard Jp in various temperatures. The results are consistent with those by phase-field method. The effect of Jp was also studied. By adjusting the Jp, effect of external stress/strain or the buffer layer on the domain pattern of PbTiO3 may be simulated. The correlation between the hysteresis loop and changing of domain pattern was also investigated. This new method provides insights in evolution of domain patterns in terms of temperature, the coupling energy and external electric voltage.

Graphite, a predominantly chosen anode material for commercial lithium ion batteries (LIBs), has been reported to have negligible intercalation capacity as an anode for sodium ion batteries (NIBs). Disordered carbon exhibits high Na intercalation capacity and emerges as a leading candidate for NIB applications. However, the mechanism of Na+ ion insertion into disordered carbon is still controversial. Here, we propose an ab initio model for disordered carbon and investigate the intercalation mechanism of Na into the layered domains. Our ab initio calculations reveal that a larger interlayer distance and the presence of defects can effectively overcome the van der Waals interaction between graphene sheets and help Na intercalation to form NaC8. The calculation results clarify the mechanism of the Na intercalation and account for the presence of sloping and flat regions of charge–discharge curves in disordered carbon reported in numerous experiments. This reveals new prospects for helping Na intercalation into graphite.

Three-dimensional (3D) integrated circuits (ICs) are the most important packaging technology for next-generation semiconductors. Cu-to-Cu throughsilicon via interconnections with micro-bumps are key components in the fabrication of 3D ICs. However, significant reliability concerns have been raised due to the formation of brittle intermetallic compounds in the entire 3D IC joints. This study proposes a Ga-based Cu-to-Cu bonding technology with Pt under bump metallurgy (UBM). A systematic analysis of reactive wetting between Ga solders and polycrystalline, single-crystalline, and Ptcoated Cu substrates was conducted. Pt UBM as a wetting layer was identified to be a key component for Ga-based Cu-to-Cu bonding. Pt-coated Cu substrates were bonded using Ga solders with various Ga-to-Pt ratios (n) at 300℃. When n ≥ 4, the Cu/Pt/Ga/Pt/Cu interface evolves to Cu/facecentered cubic (fcc)/γ1-Cu9Ga4/fcc/Cu, Cu/fcc/γ1-Cu9Ga4 + Ga7Pt3/fcc/Cu, and finally Cu/fcc + Ga7Pt3/Cu structures. The desired ductile solid solution joint formed with discrete Ga7Pt3 precipitates. When n ≤ 1, a Cu/Ga7Pt3/Cu joint formed without Cu actively participating in the reactions. The reaction mechanism and microstructure evolution were elaborated with the aid of CALPHAD thermodynamic modeling.

The substantial heat generated in three-dimensional integrated circuits and high-power electronics has made thermal management a critical challenge for reliability in the electronics industry. Pure indium solder has been used as a thermal interface material to minimize the contact thermal resistance between a chip and its heat sink. Indium and indium-based alloys are potential lead-free solder for low-temperature applications. Heat sinks in the heat dissipation system as well as substrates of electronic joints are usually made of copper, with nickel being the most commonly used diffusion barrier on the chip side. Therefore, the Cu/In/Ni sandwich structure would be encountered in electronic devices. The soldering process for forming the Cu/In/Ni structure crucially determines the reliability of devices. In this study, Cu/In/Ni interfacial reactions at 280 °C were investigated. Intermetallic compounds were identified and the microstructural evolution was observed. A strong coupling effect between Cu and Ni was found, which caused several peculiar phenomena: (1) the formation of a Cu–In compound (the Cu11In9 phase) at the In/Ni interface; (2) the formation of two sub-layers of the Cu11In9 phase at the Cu/In interface; (3) the formation of faceted rod-like Cu11In9 grains; and (4) the formation of a half-Cu11In9, half-Ni3In7 microstructure after prolonged reactions. The mechanism of phase transformations is elucidated based on the calculated Cu–In–Ni ternary phase diagram using CALPHAD thermodynamic modeling.

The Sn-20 wt.%In (Sn–20In) alloy is a promising base material for low-temperature Pb-free solders. Zn is usually added in solders to reduce the extent of undercooling during reflow, while Ag and Ni are commonly seen under bump metallurgy in contact with solder in electronic products. In this study, solid-state reactions at 150 °C between Zn-doped Sn–20In solders and Ag and Ni substrates are investigated. In Sn–20In–xZn/Ag couples, when the Zn-doping level is low (x ≤ 1.0), the reaction path is γ-InSn 4 /ζ-AgZn/ζ-(Ag,In)/ζ-AgZn/Ag, that the ζ-AgZn layer near the solder has non-uniform composition, the ζ-(Ag,In) layer has a porous microstructure, and the ζ-AgZn phase near the substrate is composed of small grains. When the Zn doping level is high (x ≥ 2.0), the reaction path becomes γ-InSn 4 /ε-AgZn 3 /γ-Ag 5 Zn 8 /ζ-AgZn/Ag, that all intermetallic compounds (IMCs) are planar and neither Sn nor In participate in the reactions. Although the reactions are very sensitive to Zn contents, the overall thicknesses of IMCs do not vary much with different Zn-doping levels. In Sn–20In–xZn/Ni couples, the planar Ni 5 Zn 21 phase is the only reaction product with a very slow growth rate. The interfacial liquation in Sn–20In/Ni contacts can be fully mitigated with a minor Zn addition of 0.5 wt.%.

In this study we have used atomistic simulations to investigate the role of surface on the size-dependent mechanical properties of nano-wires. In particular, we have performed computational investigation on single crystal face-centered cubic copper nano-wires with diameters ranging from 2 to 20 nm. The wire axis for all the nano-wires are considered along the [0 0 1] direction. Characterization of the initial optimized structures revealed clear differences in interatomic spacing, stress, and potential energy in all the nano-wires. The mechanical properties with respect to wire diameter are evaluated by applying tension along the [0 0 1] direction until yielding. We have discussed the stress–strain relationships, Young’s modulus, and the variation in potential energy from surface to the center of the wire for all the cases. Our results indicate that the mechanical response (including yield strain, Young’s modulus, and resilience) is directly related to the proportion of surface to bulk type atoms present in each nano-wire. Thus the size-dependent mechanical properties of single crystal copper nano-wire within elastic region are attributed to the surface to volume ratio (surface effect). Using the calculated response, we have formulated a mathematical relationship, which predicts the nonlinear correlation between the mechanical properties and the diameter of the wire.

Edge-decorated graphene nanoribbons are investigated with the density functional theory; they reveal three stable geometric structures. The first type is a tubular structure formed by the covalent bonds of decorating boron or nitrogen atoms. The second one consists of curved nanoribbons created by the dipole-dipole interactions between two edges when decorated with Be, Mg, or Al atoms. The final structure is a flat nanoribbon produced due to the repulsive force between two edges; most decorated structures belong to this type. Various decorating atoms, different curvature angles and the zigzag edge structure are reflected in the electronic properties, magnetic properties and bonding configurations. Most of the resulting structures are conductors with relatively high free carrier densities, whereas a few are semiconductors due to the zigzag-edge-induced anti-ferromagnetism.

Sn-20.0 wt%In (Sn-20.0In) alloy is a promising base material in Pb-free solders for low-temperature applications. Zn is often used as an additive to Pb-free solders to reduce the extent of undercooling during reflow. Cu is the most commonly used substrate in electronics industry. Interfacial stability at Sn–In–Zn/Cu joints is crucial to reliability of electronic products. In this study, interfacial reactions between Sn-20.0 wt%In-x wt%Zn (Sn-20.0In-xZn) solders and Cu where x = 0.5, 0.7, 1.0, 2.0, 3.0, and 5.0 at 150, 230, and 260 °C were experimentally examined. It is found that the reaction phase formation and interfacial morphologies are strongly influenced by Zn concentrations. The reaction phases evolve from the Cu6Sn5 phase, CuZn and Cu5Zn8 phase, to Cu5Zn8 phase with higher Zn doping in the solders. The Cu5Zn8 phase acted as a diffusion barrier and suppressed the growth of the Cu6Sn5 phase. The results indicate that 2.0 wt%Zn addition resulted in the gentlest reactions during both soldering and solid-state ageing in Sn-20.0In-xZn/Cu couples.

Au-Ge alloys are promising materials for high-power and high-frequency packaging and Ni is frequently used as diffusion barriers. This study investigates interfacial reactions in Au-12Ge/Ni joints at 300°C and 400°C. For the reactions at 300°C, typical interfacial morphology was observed and the diffusion path was (Au) + (Ge)/NiGe/Ni5Ge3/Ni. However, an interesting phenomenon – the formation of (Au,Ni,Ge)/NiGe alternating layers – was observed for the reactions at 400°C. The diffusion path across the interface was liquid/(Au,Ni,Ge)/NiGe/···/(Au,Ni,Ge)/NiGe/Ni2Ge/Ni. The periodic thermodynamic instability at the NiGe/Ni2Ge interface caused the subsequent nucleation of new (Au,Ni,Ge)/NiGe pairs. The thermodynamic foundation and mechanism of formation of the alternating layers are elaborated in this paper.

Overgrowth of the intermetallic compound (IMC) layer between Sn–58 wt.% Bi (Sn–58Bi) solders and Cu substrates significantly degrades the reliability of electronic products. Doping active elements into Sn–58Bi solders is a common approach for improving joint properties. In this study, minor amounts of Ga, ranging from 0.25 to 3.0 wt.%, were doped into Sn–58Bi solders. The interfacial reactions between the Ga-doped Sn–58Bi solders and Cu substrates at 200 °C were investigated using electron probe microanalysis (EPMA) and CALPHAD thermodynamic modeling. Four IMCs, namely θ-CuGa2, γ1-Cu9Ga4, η-Cu6Sn5, and ε-Cu3Sn phases, were observed. For longer reaction time or lower doping levels of Ga in the solders, the diffusion paths across the Ga-doped Sn–58Bi/Cu couples changed from L/θ/γ1/Cu to L/γ1/ε/Cu and L/γ1/η/ε/Cu. The dense γ1-Cu9Ga4 layer acted as a native diffusion barrier. The growth of the IMC layer was effectively suppressed with minor Ga addition.

Cu-to-Cu bonding to connect through-silicon vias in three-dimensional integrated-circuit packaging is the most important interconnection technology in the next-generation semiconductor industry. Soldering is an economic and fast process in comparison with diffusion bonding methods. Ga has high solubility of up to 20 at.% in the Cu-rich face-centered cubic (FCC) phase and high mobility at moderate temperatures. In this work, an attempt has been made to evaluate Ga-based Cu-to-Cu interconnection by transient liquid-phase (TLP) bonding. The Cu/Ga interfacial reactions at temperatures ranging from 160°C to 300°C were examined. For reactions at temperatures lower than 240°C, the reaction path is Cu/γ 3-Cu9Ga4/θ-CuGa2/liquid, where the γ 3-Cu9Ga4 and θ-CuGa2 phases are thin planar and thick scalloped layers, respectively, while for the reactions at 280°C and 300°C, the scalloped γ 3-Cu9Ga4 phase is the only reaction product. The phase transformation kinetics, reaction mechanisms, and microstructural evolution in the Cu/Ga couples are elaborated. In addition, reactions of Cu/Ga/Cu sandwich couples at 160°C were investigated. The original Cu/liquid/Cu couples isothermally transformed to Cu/γ 3-Cu9Ga4/ θ-CuGa2/γ 3-Cu9Ga4/Cu couples as the reaction progressed. However, cracks were observed in the θ-CuGa2 phase regions after metallographic processing. The brittle θ-CuGa2 phase is undesirable for Ga-based TLP bonding.

Nano-indentation is a sophisticated method to characterize mechanical properties of materials. This method samples a very small amount of material during each indentation. Therefore, this method is extremely useful to measure mechanical properties of nano-materials. The measurements using nanoindentation is very sensitive to the surface topology of the indenter and the indenting surfaces. The mechanisms involved in the entire process of nanoindentation require an atomic level understanding of the interplay between the indenter and the substrate. In this paper, we have used atomistic simulation methods with empirical potentials to investigate the effect of various types of pristine indenter on the defect nucleation and growth. Using molecular dynamics simulations, we have predicted the load-depth curve for conical, vickers, and sperical tip. The results are analyzed based on the coherency between the indenter tip and substrate surface for a fixed depth of 20 Å. The depth of defect nucleation and growth is observed to be dependent on the tip geometry. A tip with larger apex angle nucleates defects at a shallower depth. However, the type of defect generated is dependent on the crystalline orientation of the tip and substrate. For coherent systems, prismatic loops were generated, which released into the substrate along the close-packed directions with continued indentation. For incoherent systems, pyramidal shaped dislocation junctions formed in the FCC systems and disordered atomic clusters formed in the BCC systems. These defect nucleation and growth process provide the atomistic mechanisms responsible for the observed load-depth response during nanoindentation.

Molecular dynamic simulations were performed to investigate the effects of tensile and compressive loading on the mechanical properties of face-centered cubic single-crystal copper [001] nano-wires with diameters ranging from 2 to 10 nm. Characterization of the initial optimized structures revealed large variations in interatomic spacing, stress, and potential energy in all nano-wires, which resulted in tensile stress for surface atoms and compressive stress for internal atoms. This phenomenon is more apparent in thin nano-wires (<6 nm) than thick nano-wires (≥10 nm). These variations are the origins of asymmetric yielding and asymmetric Poisson ratio in [001] copper nano-wires during tension and compression. For example, the Poisson’s ratio exceeds 0.5 as the compressive strain approaches yield, indicating that the mechanical properties of single-crystal [001] nano-wire show strong directionality. The finding provides a fundamental understanding of the influence of the wire diameter on the mechanical properties of [001] nano-wires.

Sintering is an important step in densification of materials. Surfaces play a critical role in sintering, which are magnified for nano-systems (nanowire or nanoparticle). This work focuses on the investigation of the initial stages of nanoparticle sintering by atomistic simulation methods. We have characterized two different distinct systems, Argon clusters and mixed oxide systems, for analysis. Due to better particulate mixing, a combination of nanowire and nanoparticle was easier to sinter than a combination of nanoparticle and nanoparticle for Argon clusters. For a mixed oxide system (rutile-TiO2 and MgO), our results indicate that size, dimensions and structural stability of oxide nano-structures play an important role during sintering. For TiO2 nanowire and MgO nanoparticle system, Ti atoms diffuse faster than Mg atoms, while for TiO2 nanosphere and MgO nanoparticle combination, Mg atoms diffuse faster than Ti atoms during sintering. This diffusion is guided by the relative size and cohesive energies of the nano-systems.

Interfacial reactions at 100 and 150 °C in the Sn–20.48 at.% In–3.05 at.% Ag (Sn–20.0 wt% In–2.8 wt% Ag)/Ni couples are studied. Three unusual phenomena are observed. First, liquation is found in Sn–20.48 at.% In–3.05 at.% Ag (Sn–In–Ag)/Ni couples that are reacted at 150 °C, which is lower than the melting points of both the solder and the Ni substrate. In addition to the Ni3Sn4 phase, liquid phase is formed in the reaction layer. Second, the liquid phase disappears and isothermal solidification occurs when there is prolonged isothermal heat treatment at 150 °C. The results are similar to those for transient liquid phase bonding. Third, the thickness of the reaction layer in Sn–In–Ag/Ni couples that are reacted for 1440 h at 150 °C is 40 times thicker than that of those reacted at 100 °C. The reaction mechanisms for these three unusual phenomena: liquation, isothermal solidification, and an extraordinary increase in the reaction rate for only 50 °C difference in temperature are elaborated and are related to each other.

Interfacial reactions between Sn–20 wt.%In–2.8 wt.%Ag (Sn–20In–2.8Ag) Pb-free solder and Cu substrate at 250, 150, and 100 °C were investigated. A scallop-type η-Cu6Sn5 phase layer and a planar ε-Cu3Sn phase layer formed at the interface at 250 °C. The indium content in the molten solder near the interface was increased with the formation of the η-Cu6Sn5 phase; and the η-Cu6Sn5, Ag2In, Cu2In3Sn, and γ-InSn4 phases formed from the solidification of the remaining solder. At 100 and 150 °C, only the η-Cu6Sn5 phase was found at the interface. However, unusual liquid/solid reaction-like interfacial morphologies, such as irregular elongated intermetallic layers and isolated intermetallic grains, were observed in the solid-state reactions. These η phase layers had less Sn content than the Sn–20In–2.8Ag alloy, resulting in an excess Sn-rich γ-InSn4 phase accumulating at the interface and forming porous η layers on top of the initially formed dense η layers at 150 °C. At 100 °C, large elongated η grains were formed, whereas the interfacial layers remained almost unchanged after prolonged reaction. Based on the experimental evidence, the growth of the η phase was proposed to follow a diffusion-controlled mechanism at 250, 150 and 100 °C, while that of the ε phase was probably controlled by the reaction.

Soldering is an ancient process, having been developed 5000 years ago. It remains a crucial process with many modern applications. In electronic devices, electric currents pass through solder joints. A new physical phenomenon – the supersaturation of solders under high electric currents – has recently been observed. It involves (1) un-expected supersaturation of the solder matrix phase and (2) the formation of unusual “ring-shaped” grains. However, the origin of these phenomena is not yet understood. Here we provide a plausible explanation of these phenomena based on the changes in the phase stability of Pb-Sn solders. Ab initio-aided CALPHAD modeling is utilized to translate the electric current-induced effect into the excess Gibbs free energies of the phases. Hence, the phase equilibrium can be shifted by current stressing. The Pb-Sn phase diagrams with and without current stressing clearly demonstrate the change in the phase stabilities of Pb-Sn solders under current stressing.

Density functional theory (DFT) calculations using supercells have proven quite successful in predicting defect properties. Although forming defect groups, for example, the Schottky pair in MgO, are usually energetically favorable in many ionic systems, it is useful to obtain the defect energies of such systems without any defect-defect interactions as reference energies. However, determining non-interacting energies through multi-defect supercell calculations is challenging due to interactions between the defects that are difficult to quantify and can only be avoided by using very large supercells. One solution to this problem is to build an effective multi-defect cell through separate isolated defect calculations, with each defect in their own supercell. However, this isolated defect approach requires that the charge compensation be introduced through charged supercells, and a careful treatment of the cell energetics and electron reference energy is required. In this paper we assess the use of an isolated defect approach for modeling charge-compensating defect groups using the test case of MgO. The appropriate asymptotic condition for the electron reference energy shift is formulated and a method to meet the condition is given. We also demonstrate the strong coupling effect between residual strain energy and electrostatic energy in charged cells, demonstrating that these two effects cannot generally be separated and treated in isolation. The key steps in an approach that yields accurate defect group energies from the isolated defect calculations are presented. The non-interacting Schottky defect formation energy in MgO is determined to be 6.1 eV through calculation of separated isolated charged cells containing and, respectively, while the binding energy between the charged defects and is determined to be 1.5 eV. This approach may also be of value for accurate modeling of general isolated defects. The formation energy of an isolated neutral Mg vacancy is found to be lower than that of a Schottky pair, suggesting that it is possible to have significant thermal cation defect formation in MgO without forming charge compensating Schottky pairs.

Ternary Sn–Sb–Cu alloys are prepared. The primary solidification phases, the phase transformation temperatures, and the equilibrium phases at 250 °C are experimentally determined. The liquidus projection and the 250 °C phase equilibria isothermal section of the Sn–Sb–Cu system are proposed based on the experimental results and the phase diagrams of the three constituent binary systems. Using the CALPHAD approach, thermodynamic modeling of the Sn–Sb–Cu ternary system is carried out based on the experimental information determined in this study and those in the literatures, together with the developed thermodynamic models of the three constituent binary systems. The liquidus projection and the isothermal sections are then calculated using the models developed in this study and the results are in good agreement with experimental determinations.

The interfacial reactions in Sn-0.7wt%Cu/ENIG SUS304 couples at 240, 255, and 270 °C are examined in this study. The Ni-containing ternary Cu6Sn5 phase is formed at the Ni/liquid interface in the early reaction stage then it detaches massively from the SUS304 substrate and splits into two layers in the molten solder as the reaction time increases. This phase finally disintegrates and disappears. The square pillar-shaped FeSn2 phase is found on top of the SUS304 substrate when the Cu6Sn5 layer detaches. The reaction phase formation, detachment, and split mechanisms are proposed. The spalling phenomenon is reviewed and discussed. The growth mechanism of the FeSn2 phase obeys the parabolic law, and the activation energy is determined to be 112.5 KJ/mol.

Classical molecular dynamics (MD) simulations are used to examine the deposition of polyatomic fluorocarbon (FC) beams on polypropylene (PP) and polystyrene (PS) surfaces. The goal is to investigate the ways in which different FC ions, in this case CF3+ and C3F5+, chemically modify the two different polymer surfaces. The simulations predict that the chemical reactions that occur upon impact are highly localized. As a result, with the same incident energy, CF3+ ions generate more PP chain fragments and facilitate more etching of the surface than C3F5+ ions. In contrast, C3F5+ ions promote more cross-linking between PP chains and the growth of FC films on the PP surface. In PS, there is more penetration of the ions than in PP as well as increased formation of CF2 particles, which indicates that deposition on PS yields fluorocarbon films more easily than deposition on PP. The simulations thus provide important insights into the complex mechanisms associated with the processes used to engineer polymer thin films in FC ion beams and plasmas and illustrate how differences in polymer structure ultimately influence such properties as sputtering, chemical modification, and thin film growth.

Plant oils have been used as environmentally benign lubricants since they present high viscosity index and flash points and low evaporation loss. Triacylglycerols (TAG) are the major components of naturally occurring oils and fats and are able to produce high strength lubricant films. One of the main concerns that hinders the usage of triacylglycerols as lubricants, however, is the thermal stability of these molecules. In this paper, we report on the effect of chain structure on density, viscosity, and thermal stability of triacylglycerols using molecular dynamics simulations. The selected triacylglycerols are trilauroylglycerol (LLL-TAG), tristearoylglycerol (SSS-TAG), trans-trioleoylglycerol (trans-OOO-TAG), and trans-trilinolenoylglycerol (trans-LeLeLe-TAG). The first two TAGs are saturated molecules with a different number of carbons in the chain, and the second two TAGs are monounsaturated and polyunsaturated molecules, respectively. The computed results demonstrate that the length of the aliphatic chain influences the physical properties of triacylglycerols. TAGs with short chain (LLL-TAG) show higher density than TAGs with longer chains. Viscosity is determined by the degree of recoil of the aliphatic chains and by the number and location of unsaturated bonds. Thermal stability, as represented by the ability of triacylglycerols to stay in a disordered phase during the cooling process, is related to the order−disorder phase transition temperature. Since the phase transition temperature can be correlated to the thermal stability during the cooling process, LeLeLe-TAG shows the highest thermal stability among the systems considered. These results can aid in the design of molecules with specific lubrication properties.

Ternary Sn-In-Cu alloys are prepared and equilibrated at 250 °C for 2 to 20 weeks. The phases formed in these alloys are experimentally determined. The 250 °C Sn-In-Cu isothermal section is established according to the phase equilibrium information obtained in this study and that of the three constituent binary systems. It has eight single-phase regions, namely liquid, δ1-Cu41Sn11, ε-Cu3Sn, δ2-Cu7In3, η-(Cu6Sn5, Cu2In), Cu11In9, Cu2In3Sn, and α-(Cu) phases, 14 two-phase regions, and seven three-phase regions. In the Sn-In-Cu system at 250 °C, the η-Cu6Sn5 and η-Cu2In phases form a continuous solid solution and the ternary Cu2In3Sn compound is observed. The δ1-Cu41Sn11 phase is stabilized at 250 °C with the introduction of indium although it transforms into α-(Cu) and ε-Cu3Sn phases via a eutectoid reaction around 350 °C in the binary Sn-Cu system. Except for the Cu11In9 phase and the Cu2In3Sn ternary compound, the other binary compounds all have significant indium and tin mutual solubilities.

Low-resistance V/Al/V/Ag contacts have recently been reported to Al0.27Ga0.73N/GaN heterostructures with a thin GaN cap. These contacts had smooth surfaces and good edge definition. In this study, the V/Al/V/Ag metallization was adapted to other compositions of n-AlxGa1-xN, and it was found to provide low specific contact resistances as well as smooth surfaces on both n-GaN and n-Al0.58Ga0.42N. Another advantage of these contacts is that limited changes in specific contact resistance and morphology were observed when changing the metal layer thicknesses and processing conditions of the V/Al/V/Ag metallization on these semiconductors. The V (20)/Al (95)/V (20)/Ag (120 nm) contact provided a specific contact resistance of (2.1±0.9)×10-6 Ω cm2 when annealed at 825 °C for 30 s on n-GaN and a V (20)/Al (95)/V (5)/Ag (120 nm) contact provided a value of (2.4±0.3)×10-4 Ω cm2 when annealed at 875 °C for 60 s on n-Al0.58Ga0.42N. In each case, transmission electron microscopy revealed that the reaction between the semiconductor and metallization was limited and the majority of the interface was contacted by a composite of primarily Ag-bearing grains. Measurements of the specific contact resistance as a function of temperature revealed that field emission is the dominant current transport mechanism in low resistanc- e Ohmic contacts to n-GaN and n-Al0.58Ga0.42N.

Sn-In alloys are promising low-melting-point Pb-free solders. Knowledge of the ternary Sn-In-Cu liquidus projection is important for Sn-In solder applications. Sn-In-Cu ternary alloys were prepared and their primary solidification phases and phase-transformation temperatures during heating were determined. The liquidus projection of the Sn-In-Cu ternary system was determined based on the primary solidification phase at different compositional regimes, the phase-transformation temperatures of the ternary alloys, the phase boundaries and reaction temperatures of the constituent binary systems, and the available ternary Sn-In-Cu data in the literature. No ternary compound was found in the as-cast alloys. The Sn-In-Cu liquidus projection has 11 primary solidification phase regions and seven ternary invariant reactions with the liquid phase, and η-(Cu6Sn5,Cu2In) has a very large compositional regime as the primary solidification phase. A very interesting phenomenon that was also observed is that the solidification paths of some Sn-In-Cu alloys surpass the liquidus trough after their intersections.

Polythiophene is a conductive polymer that has attracted much interest in recent years because its properties are desirable for applications that include light-emitting diodes, field-effect transistors, and photovoltaics. Optimization of the performance of polythiophene in these devices requires the development of processing methods that can simultaneously control its chemistry and morphology on the nanometer scale. One such method is surface polymerization by ion-assisted deposition (SPIAD) in which conducting polymer thin films are grown on substrates by the simultaneous deposition of hyperthermal polyatomic ions and thermal neutrals in vacuum. Here, mass-selected beams of thiophene ions are deposited on α-terthiophene oligomers in experiments, and density functional theory−molecular dynamics (DFT−MD) simulations are carried out to determine the dominant mechanisms responsible for the SPIAD process. Both neutral and positively charged systems are considered in the simulations in order to assess the effect of charge on the results. The experimental results show that polymerization occurs preferentially under a narrow set of ion energy and ion/neutral ratio conditions. The DFT−MD simulations illustrate the manner in which ion energies affect polymerization and reveal how secondary chemical reactions can substantially modify both the thin film and the substrate.

Classical molecular dynamics simulations are used to study the effects of continuous fluorocarbon ion beam deposition on a poly(vinylidene fluoride-trifluoroethylene) [P(VDF-trFE)] surface, a polymer with electromechanical properties. Fluorocarbon plasma processing is widely used to chemically modify surfaces and deposit thin films. It is well accepted that polyatomic ions and neutrals within low-energy plasmas have a significant effect on the surface chemistry induced by the plasma. The deposition of mass selected fluorocarbon ions is useful to isolate the effects specific to polyatomic ions. Here, the differences in the chemical interactions of C3F+5C3F5+ and CF+3CF3+ ions with the P(VDF-trFE) surface are examined. The incident energy of the ions in both beams is 50eV. The CF+3CF3+ ions are predicted to be more effective at fluorinating the P(VDF-trFE) surface than C3F+5C3F5+ ions. At the same time, the C3F+5C3F5+ ions are predicted to be more effective at growing fluorocarbon thin films. The simulations also reveal how the deposition process might ultimately modify the electromechanical properties of this polymer surface.

The preferential structures of small copper clusters Cun (n = 2−9) and the adsorption of methanol molecules on these clusters are examined with first principles, molecular dynamics simulations. The results show that the copper clusters undergo systematic changes in bond length and bond order associated with altering their preferential structures from one-dimensional structures, to two-dimensional and three-dimensional structures. The results also indicate that low coordination number sites on the copper clusters are both the most favorable for methanol adsorption and have the greatest localization of electronic charge. The simulations predict that charge transfer between the neutral copper clusters and the incident methanol molecules is a key process by which adsorption is stabilized. Importantly, the changes in the dimensionality of the copper clusters do not significantly influence methanol adsorption.

The development of the microstructure of mechanical-deformation-induced Sn whiskers on electroplated films has been examined using a focused ion beam system (FIB). The 6-μm-thick matte Sn films were compressed by using a ZrO2 ball indenter under ambient conditions. After compression, tin whiskers and small nodules were found adjacent to, and several grains further away from, the indents. The cross-sectional microstructures of the indents and whiskers indicate that the lateral boundaries of the newly created grains caused by recrystallization are the main routes for stress relaxation.

The Cu–Sn binary system is important for various applications, especially for recent developments in the electronics packaging industry. The ϵ-Cu3Sn and η-Cu6Sn5 (η′ phases) phases are frequently encountered in electronics products. However, the two phases have been described as line compounds in previous thermodynamic modeling, and their compositional homogeneities were not considered. In this study, the thermodynamic properties of the Cu–Sn binary system are modeled and the phase diagram is calculated by the CALPHAD method, using experimental information reported in the literature. The ϵ and η (η′) phases are described using compound energy models with two and three sublattices, respectively, so that their compositional homogeneities could be calculated. Good agreement was observed between the calculated result and the existing experimental data.

In this study, we investigated mechanical deformation-induced Sn whisker growth, which is frequently encountered in advanced flexible substrate packaging. Concentrated compressive stresses are introduced around the leads and solder surface finish joints connected by compression fixing. Six types of pure Sn thin films were electroplated on Ni-protected Cu substrates. These were 2- and 6-μm-thick Sn films electroplated with three different current densities: 2, 10, and 20 A/dm2. These films were compressed at room temperature and ambient humidity. The surface and cross-sectional grain morphologies of the films were examined by scanning electron microscopy and focused ion beam spectroscopy, respectively. The grain orientations of the electroplated Sn films were analyzed by x-ray diffraction and electron backscatter diffraction. After compression, nodule hillocks and whiskers were found around the indents. Beneath the indents, the original columnar Sn grains were deformed, and recovery and recrystallization processes occurred. Rapid whisker formation was observed. The whiskers induced by mechanical deformation are closely related to the grain microstructures, and the initial compression stresses are critical to the types and distribution of whiskers as well.

Soldering is the most important of joining technologies and there are numerous solder joints in modern electronic products. The phases and microstructures of solder joints are critical to their properties. Various remarkable phenomena caused by phase transformation and microstructural evolution in solder joints have been reported. The phenomena include ripening, layer detachment, liquation, cruciform pattern formation, solid-state amorphization, alternating layer formation, shift of reaction paths, and the effects of electromigration.

Soldering is the most important of joining technologies and there are numerous solder joints in modern electronic products. The phases and microstructures of solder joints are critical to their properties. Various remarkable phenomena caused by phase transformation and microstructural evolution in solder joints have been reported. The phenomena include ripening, layer detachment, liquation, cruciform pattern formation, solid-state amorphization, alternating layer formation, shift of reaction paths, and the effects of electromigration.

Classical molecular dynamics (MD) simulations are used to study the effect of continuous hydrocarbon (HC) and fluorocarbon (FC) ion beam deposition on a polystyrene (PS) surface. Plasma processing is widely used to chemically modify surfaces and deposit thin films, and it is well-accepted that polyatomic ions and neutrals within low-energy plasmas have a significant effect on the surface chemistry. Here a comparison is made of the manner in which polyatomic FC ions and similarly structured HC ions react with PS and produce new structures. Specifically, the deposition of beams of C3H5+, CH3+, C3F5+, and CF3+ on PS surfaces at experimental fluences is considered. The simulations predict that the backbone chains are modified significantly more than the phenyl groups and that larger ions with lower velocities and larger collision cross sections modify the substrate to a shallower depth than smaller ions with higher velocities, even though all their incident kinetic energies are the same. Additionally, HC ions dissociate more readily than FC ions during deposition. Consequently, smaller HC ions are predicted to chemically modify the polystyrene to a greater extent than larger HC ions or FC ions.

The electromigration effect upon the γ-InSn4/Cu interfacial reactions have been studied by examining the γ-InSn4/Cu/γ-InSn4 couples annealed at 160 °C with and without current stressing. Scallop-type η-Cu6(Sn,In)5 phase layers are formed in the couples without current stressing and at the γ-InSn4/Cu interface where electrons are flowing from the γ-InSn4 to the Cu. The reaction path is Cu/η-Cu6(Sn,In)5/γ-InSn4. However, very large η-Cu6(Sn,In)5 compounds are found at the Cu/γ-InSn4 interface where electrons are from Cu to the γ-InSn4. Although the melting points of both γ-InSn4 and Cu are higher than 160 °C, the liquid phase is formed at 160 °C in the electrified couple at the downstream γ-InSn4 phase near the Cu/γ-InSn4 interface. The reaction path is Cu/η-Cu6(Sn,In)5/liquid/γ-InSn4. The liquid phase propagates along the grain boundaries of the γ-InSn4 matrix. The very large η-Cu6(Sn,In)5 compounds are the coupling results of the liquid phase penetration and the Cu transport enhancement due to electromigration.

Sn–In alloys are promising low-melting-point Pb-free solders. Cu and Ni are common substrates in the electronic products. This study examines the interfacial reactions in the Sn–20 at.% In(γ–InSn4)/Cu and Sn–20 at.% In/Ni couples at 160 °C. Only the η–Cu6Sn5 phase layer is formed in the Sn–20 at.% In/Cu couple, and the layer grows thicker with longer reaction time. The reaction path is γ–InSn4/η–Cu6Sn5/Cu. A peculiar phenomenon with the bulging of the couple near the Ni substrate is found in the Sn–20 at.% In/Ni couple. A liquid phase is formed by interfacial reaction in the solid/solid Sn–20 at.% In/Ni couple at 160 °C, and the reaction path is γ–InSn4/liquid/δ–Ni3Sn4 + liquid/(δ–Ni3Sn4)/Ni. Usually Ni has a slower reaction rate with solders; however, the consumption rates of Ni substrate are much higher than those of Cu substrate in this study when they are in contact with the Sn–20 at.% In alloy at 160 °C due to the formation of the liquid phase in the Sn–20 at.% In/Ni couple.

Interfacial reactions in γ–InSn4(Sn–20 at.% In)/Ni couples at 130, 140, 150, and 160 °C were investigated. Ni3Sn4 phase with significant indium solubility was formed in the couple reacted at 130 and 140 °C, and the reaction path was γ–InSn4/Ni3Sn4/Ni. For the couples reacted at 150 and 160 °C, even though both γ–InSn4 and Ni were solid phases, the liquid phase was formed in the couples. A distinguished feature was the nickel substrates becoming nonplanar with spikes at various locations and the Ni3Sn4 phase layer on top of the nickel spikes. Except at regions near the nickel spikes, the reaction layer consisted of precipitates and was not a homogeneous phase. The reaction path is γ–InSn4/Ni3Sn4/Ni at the location with Ni3Sn4 phase growing on Ni. However, if the Ni3Sn4 phase does not nucleate, the liquid phase forms at the interface with accumulation of indium atoms, and the reaction path is γ–InSn4/ liquid/liquid + Ni3Sn4/Ni.

A Pb-free composite solder is prepared with a Pb-free solder substrate and a plated-indium layer. The indium layer melts during the soldering process, wets the substrates, and forms a sound solder joint. Since the melting temperature of indium is 156.6°C, lower than that of the eutectic Sn-Pb, which is at 183°C, the soldering process can be carried out at a temperature lower than that of the conventional soldering process. Composite solder joints with three different Pb-free solders, Sn, Sn-3.5 wt.% Ag, and Sn-3.5 wt.% Ag-0.5 wt.% Cu, and two substrates, Ni and Cu, are prepared. The interfaces between the indium layer, Pb-free solder, and Ni and Cu substrate are examined. A good solder joint is formed after a 2-min reflow at 170°C. A very thick reaction zone at the indium/Pb-free solder interface and a thin reaction layer at the indium/substrate interface are observed.

Lead-free flip-chip joining processes have attracted the most research interests recently, and inhomogeneous interfacial reactions were observed in the flip chip solder joints. The Cu6Sn5 phase with high nickel solubility, (Cu,Ni)6Sn5 phase, was formed at the substrate side adjacent to the Ni layer in a flip-chip joint with a substrate/Ni/solder/Cu/(Ni,V)/die structure. A Sn/Cu/Sn/Ni/Sn/Cu/Sn couple was prepared to simulate the flip-chip joint. At 200°C, the Cu6Sn5 phase was formed on both ends of the Sn phase at the Cu/Sn and Sn/Ni interfaces for the couple with an electric current stressing or a longer reaction time, but the nickel contents of the two Cu6Sn5 phases are different. The Cu6Sn5 phase at the Ni side has high nickel content and it has almost no nickel at the Cu side. It is concluded that the Cu at the chip-side diffused through the solder phase, reacted with the Ni layer at the substrate side, and the Cu6Sn5 phase with high nickel solubility, i.e. the (Cu,Ni)6Sn5 phase, was thus formed. Electromigration effects significantly enhance the diffusion rate of Cu, but do not alter the phase formation sequence.