Metallurgy

1 UBM

The UBM of a flip chip interconnect serves as a diffusion barrier between the metals of the joint and the silicon and as a wettable layer that forms a metallurgical bond with the solder. The structure of the UBM is designed to adhere to the interconnect metallization on the silicon (aluminum or copper), act as a diffusion barrier between the solder and silicon, and be a solder wettable surface. The most common UBM is the evaporated Cr/Cr–Cu/Cu/Au developed by IBM for use with an evaporated high lead content Pb–Sn alloy joined to a ceramic substrate. This UBM is expensive and new UBM systems with sputtered or plated metallizations have been developed. Plated metallizations are acknowledged to be the lowest-cost UBM.

A generic example of the processes and materials required to manufacture a flip chip package is as follows. Figure 2 is a schematic illustration that shows the structure of a UBM/solder interconnect. A deposition of metal on the aluminum or copper pads on the silicon device is used to provide a solder wettable surface for the flip chip solder interconnects and a diffusion barrier to the silicon. The metals are deposited by physical (evaporating or sputtering) or electrochemical (plating) means. The UBM typically consists of an adhesion layer to the pad (e.g., chromium, titanium, and zincated aluminum) and a diffusion barrier (e.g., nickel). These layers are typically a few thousand angstroms thick and are patterned photolithographically. A solder wettable layer (e.g., nickel or copper) is deposited to form the exposed surface of the UBM and has a thickness that varies from a few thousand angstroms for nickel and gold to tens of micrometers for copper. Gold is used as an oxidation barrier on top of the UBM to enhance solderability. Solder is deposited on the UBM by evaporation (high lead content solders), plating, or printing solder paste. The solder is reflowed to react with the UBM to create a metallurgical bond and form the solder into ball shapes. The wafer is then diced. For assembly into a package, the silicon die is placed on a substrate that is patterned to mirror the configuration of the solder balls on the die. The substrate is a ceramic or organic material. The substrate pads are metallized with copper, Ni/Au, or a Au/Pd film to be wettable by the solder. A die is placed on the substrate and the assembly is reflowed to join the die to the substrate. The surface tension of the molten solder assists in the alignment of the die and substrate pads. After the joints are formed a silica-filled anhydride resin underfill flows under the die and is cured to enhance mechanical adhesion to the substrate.

Figure 2. Schematic illustration of a cross-section of a generic bumped flip chip interconnect.

3.1.4 Scaling-Up and Scaling-Down Operations in Process Metallurgy

Metallurgy used to be an art rather than science until mid of nineteenth century during which period large metallurgical plants were developed (based on empirical relations, trial and error, intuition, etc.) and operated successfully. Since more than one and a half centuries, metallurgy has emerged as a science and more predictive. Therefore, in metallurgy field, often the problems are posed in terms of real problem which plants are facing during operation like less productivity, environmental problems, etc. As these processes were evolved not based on well-established scientific criteria, therefore, to understand these processes and to address the problem, one has to scale down the industrial process (quite opposed to scaling up which is carried out more in aerospace/mechanical/chemical engineering discipline where the science is well developed and plant came up after studying the laboratory scale and pilot models) to laboratory scale. However, the principle involved in either scaling up or scaling down is the same. An excellent example of improving the process by scaling down is the study of ironmaking blast furnace. Blast furnaces used to produce 50 tons a year hot metal in late eighteenth century, and in the beginning of 21st century, its capacity has improved to more than 5.5 million tons/year. This is possible only by studying the blast furnace process in various parts to understand the science behind all the processes which are occurring in a blast furnace. To illustrate this, an example is given in section 3.1.6 on raceway size prediction as it affects the aerodynamics of the blast furnace and thus the heat and mass transfer and hence the productivity.

Metallurgy, ceramics, and pharmacy are ancient chemical technologies, but chemical science developed in the West from the seventeenth century. A. L. Lavoisier in the late eighteenth century, describing his work as revolutionary, introduced new language and new theory; and its exciting links to electricity made chemistry the fundamental science of the early nineteenth century. By mid-century, graduate schools like Justus Liebig's, journals, institutes, and connections with industry had inaugurated professional chemistry. Its first major international conference, settling formulae, was held in 1860; by then crude vitalism had been abandoned, and chemistry was separating into specialisms (inorganic, organic, and physical) and subspecialisms. From about 1900, chemistry has been perceived as reduced to physics, a service science rather than a fundamental one; and its achievements in fertilizers, pesticides, explosives, and drugs are looked at askance. But chemists are ubiquitous, essential in scientific teamwork; and the creation of new molecules unknown in nature has gone on apace. The history of chemistry used to be written by practitioners interested in finding the roots of modern ideas, but has been largely taken over by historians more interested in contexts.

4.1. General

Metallurgy is above all the science of alloying. From the big three: iron, copper and aluminium, and perhaps twenty other elements which are also household names, it is possible to generate more or less the complete range of alloys in common use. Of course, the interaction between pairs, or larger numbers, of elements is described, as a function of temperature and compositions, by a phase diagram. The understanding of these diagrams is as mother's milk to a metallurgist, and their importance cannot be questioned.

Phase diagrams do exist in polymer science, but their significance is not pre-eminent, and there are many involved with polymeric materials who can ply their trade without ever having to encounter one. Why the difference?

In the first instance, the number of polymers which can be envisaged and synthesised is semi-infinite. There is a huge number of ways in which carbon, oxygen, nitrogen and hydrogen atoms can be put together to make different chains or networks. So it could be argued that instead of trying to mix different types of chains to make materials with different properties, the polymer chemist merely dreams up and synthesises another molecule. There may be something in this view; however the central reason why polymer alloys are not centre stage is that they are reluctant to form solutions or compounds with each other. They simply do not alloy very well.

The word ‘alloy’ is not used by the polymer scientist (this chapter excepted). The closest equivalent is blend. In general, solid solutions of polymers are referred to as miscible blends while two-phase mixtures where there is effectively no terminal or liquid–liquid solubility, such as in the copper-lead system, are referred to as immiscible blends.

1.1 Introduction

Metallurgy is a subject evolved over a few millennia, at least the past 3000 years, probably since the dawn of known human history. Old Greek, Hebrew and Hindu scriptures allude to metals like gold, silver, copper, iron, lead, and tin as well as nonmetals like sulfur and carbon [1]. Ancient Greek philosopher Heraclitus among others believed that all substances had a single component. The concept of chemical elements itself has gone through immense changes, from ancient times through Alchemy and Middle Ages to Modern Times. Both Greeks (Empedocles) and Hindus (the Vedas) were in agreement that the basic elements were earth, water, air, and fire, while the Hindus had a fifth component, ether or vacuum. Even the Greek philosopher, Demokritos, had vacuum as the matrix in which particles which were indivisible moved about. Thus, the world of materials had at least four components as described by the illustration given in Figure 1.1, leaving aside the nonmaterial, the ether (Figure 1.1).

Figure 1.1. The four elements occurring in nature [2].

The imagination of the modern scientist leads to drawing parallels to ancient thoughts and modern science, obviously a wishful thinking, albeit very fascinating (Figure 1.2).

Figure 1.2. An imaginary parallelism between ancient concept of elements and the states of matter according to modern science.

As we know today, the ancient basic elements are in fact more complex, as for example, air consisting of nitrogen and oxygen apart from other minor gases and earth being most complex of all with mixtures of various compounds.

The ancients had even thought of combinations of these basic elements to form the substances on the Earth, the predecessors to modern chemical equations. The adherents of “Al-Chemie” (the word comes from “Chemia” prefixed by the Arabic definite form “Al”) starting from about 300 A.D. were convinced that a base metal such as copper could be converted to a noble metal, gold. Despite the fruitlessness of these efforts, some interesting concepts seemed to have evolved. The combinations of the aforesaid four elements, according to the alchemists, can lead to two entirely different “principles”: mercury and sulfur, the former standing for the metallic principle while the latter for “destruction of metallic principle” [2]. These probably were the forerunners of the modern concept of chemical reactions as represented in Figure 1.3.

Figure 1.3. Ancient concept of chemical reactions leading to formation of materials [2].

Interestingly, Paracelsus from Switzerland (1493–1541) introduced, apart from metallic and nonmetallic principles, the third one, namely Salt.

Abstract

Secondary metallurgy is a very central part of modern steelmaking process. It means a variety of different unit processes via which the final composition and even properties of steels are determined and adjusted. Typical unit processes in secondary steelmaking are deoxidation, desulfurization, degassing, decarburization, alloying and trimming additions as well as heating and temperature adjustment.

In this chapter, the fundamentals of thermodynamics and kinetics of unit processes are described. Further, the development and principles of most common technological methods to carry out processes are discussed including gas stirring, vacuum facilities, ladle furnace and chemical heating, and techniques for alloying and trimming additions.

Melt flow technique

Advanced thixotropic metallurgy (ATM) technique is a runner design concept developed by a group at CSIRO, Australia²². The technique employs a flow restriction section, a so-called melt pre-conditioner (MPC), in the runner. The MPC is aimed to shear and accelerate the flowing material before it reaches the gate. In HPDC, there are some equiaxed crystals, so called externally solidified crystals (ESCs), solidified in the shot sleeve prior to being injected into the die cavity. The shearing of partially solid alloy containing ESCs is believed to make ESCs smaller and more globular, and disperse porosity in the final microstructure. Due to smaller ESC size and more disperse porosity, ATM castings often show better mechanical properties than conventional ones.

Smaller and more globular ESCs decrease viscosity of the suspension and therefore reduce the tendency for cold shuts. This allows a lower gate speed than that in conventional HPDC to be used with the ATM design. Die erosion problem associated with high gate speeds can be reduced. Additional time gained from lower filling velocity enhances gas venting from the cavity. The ATM runner system is typically smaller than the conventional design, leading to better material yield and energy saving.

Manufacturing process of MMCs

A very important consideration in choosing the method is the availability of the desired equipment and the type of end-product. The selection of suitable process engineering is based on the desired type of quantity and distribution of reinforcement, the matrix alloy and the application. By altering the manufacturing method, the processing and the finishing, as well as by the form of the reinforcement components it is possible to obtain different material properties, although the same composition and amounts of the components are involved. Metal matrix composites can be made by liquid, solid, or gaseous state processes (Surappa, 2003).

In this research the selection of manufacturing process is based on production of MMCs by liquid process. In this process, the particulates are integrated into a molten metallic matrix by means of various techniques.