Professional Documents
Culture Documents
IC FABrication Fundamental
IC FABrication Fundamental
Introduction
Integrated circuits were made possible by experimental discoveries which showed that semiconductor devices could perform the functions of vacuum tubes, and by mid20th-century technology advancements in semiconductor device fabrication. The integration of large numbers of tiny transistors into a small chip was an enormous improvement over the manual assembly of circuits using discrete electronic components. The integrated circuit's mass production capability, reliability, and building-block approach to circuit design ensured the rapid adoption of standardized ICs in place of designs using discrete transistors. There are two main advantages of ICs over discrete circuits: cost and performance. Cost is low because the chips, with all their components, are printed as a unit by photolithography and not constructed one transistor at a time. Furthermore, much less material is used to construct a circuit as a packaged IC die than as a discrete circuit. Performance is high since the components switch quickly and consume little power (compared to their discrete counterparts) because the components are small and close together.
Invention
The integrated circuit can be credited as being invented by both Jack Kilby of Texas Instruments and Robert Noyce of Fairchild Semiconductor working independently of each other. Kilby recorded his initial ideas concerning the integrated circuit in July 1958 and successfully demonstrated the first working integrated circuit on September 12, 1958. Kilby won the 2000 Nobel Prize in Physics for his part of the invention of the integrated circuit. Robert Noyce also came up with his own idea of integrated circuit, half a year later than Kilby. Noyce's chip had solved many practical problems that the microchip developed by Kilby had not. Noyce's chip, made at Fairchild, was made of silicon, whereas Kilby's chip was made of germanium. 1
Generations
The first integrated circuits contained only a few transistors. Called "Small-Scale Integration" (SSI), digital circuits containing transistors numbering in the tens provided a few logic gates for example, while early linear ICs such as the Plessey SL201 or the Philips TAA320 had as few as two transistors. The term Large Scale Integration was first used by IBM scientist Rolf Landauer when describing the theoretical concept, from there came the terms for SSI, MSI, VLSI, and ULSI. SSI circuits were crucial to early aerospace projects, and vice-versa. Both the Minuteman missile and Apollo program needed lightweight digital computers for their inertial guidance systems; the Apollo guidance computer led and motivated the integrated-circuit technology, while the Minuteman missile forced it into mass-production. These programs purchased almost all of the available integrated circuits from 1960 through 1963, and almost alone provided the demand that funded the production improvements to get the production costs from $1000/circuit (in 1960 dollars) to merely $25/circuit (in 1963 dollars).They began to appear in consumer products at the turn of the decade, a typical application being FM inter-carrier sound processing in television receivers. The next step in the development of integrated circuits, taken in the late 1960s, introduced devices which contained hundreds of transistors on each chip, called "Medium-Scale Integration" (MSI). They were attractive economically because while they cost little more to produce than SSI devices, they allowed more complex systems to be produced using smaller circuit boards, less assembly work (because of fewer separate components), and a number of other advantages. Further development, driven by the same economic factors, led to "Large-Scale Integration" (LSI) in the mid 1970s, with tens of thousands of transistors per chip. Integrated circuits such as 1K-bit RAMs, calculator chips, and the first microprocessors, that began to be manufactured in moderate quantities in the early 1970s, had under 4000 transistors. True LSI circuits, approaching 10000 transistors, began to be produced around 1974, for computer main memories and second-generation microprocessors. The final step in the development process, starting in the 1980s and continuing through the present, was "Very Large-Scale Integration" (VLSI). The development started with hundreds of thousands of transistors in the early 1980s, and continues beyond several billion transistors as of 2007. There was no single breakthrough that allowed this increase in complexity, though many factors helped. Manufacturing moved to smaller rules and cleaner fabs, allowing them to produce chips with more transistors with adequate yield. Design tools improved enough to make it practical to finish these designs in a reasonable time. The more energy efficient CMOS replaced NMOS and PMOS, avoiding a prohibitive increase in power consumption.
Better texts such as the landmark textbook by Mead and Conway helped schools educate more designers, among other factors. In 1986 the first one megabit RAM chips were introduced, which contained more than one million transistors. Microprocessor chips passed the million transistor mark in 1989 and the billion transistor mark in 2005. The trend continues largely unabated, with chips introduced in 2007 containing tens of billions of memory transistors To reflect further growth of the complexity, the term ULSI that stands for "Ultra-Large Scale Integration" was proposed for chips of complexity of more than 1 million transistors. Wafer-scale integration (WSI) is a system of building very-large integrated circuits that uses an entire silicon wafer to produce a single "super-chip". Through a combination of large size and reduced packaging, WSI could lead to dramatically reduced costs for some systems, notably massively parallel supercomputers. The name is taken from the term Very-Large-Scale Integration, the current state of the art when WSI was being developed. System-on-a-Chip (SoC or SOC) is an integrated circuit in which all the components needed for a computer or other system are included on a single chip. The design of such a device can be complex and costly, and building disparate components on a single piece of silicon may compromise the efficiency of some elements. However, these drawbacks are offset by lower manufacturing and assembly costs and by a greatly reduced power budget, because signals among the components are kept on-die, much less power is required
Active device count 1 - 100 100 - 1000 1000 100,000 105 106 > 106
Note: Active device count means the number of transistors in a single chip (IC).
Classification
Integrated circuits can be classified into analog, digital and mixed signal (both analog and digital on the same chip). Digital ICs can contain anything from one to millions of logic gates, flip flops multiplexers, and other circuits in a few square millimeters. The small size of these circuits allows high speed, low power dissipation, and reduced manufacturing cost compared with board-level integration. These digital ICs, typically microprocessors, DSPs, and micro controllers work using binary mathematics to process "one" and "zero" signals. Analog ICs, such as sensors, power management circuits, and operational amplifiers, work by processing continuous signals. They perform functions like amplification, active filtering, demodulation, mixing, etc. Analog ICs ease the burden on circuit designers by having expertly designed analog circuits available instead of designing a difficult analog circuit from scratch. Mixed Signal ICs combine analog and digital circuits on a single chip to create functions such as A/D converters and D/A converters. Such circuits offer smaller size and lower cost, but must carefully account for signal interference.
Notable IC Developments
In the 1980s programmable integrated circuits were developed. These devices contain circuits whose logical function and connectivity can be programmed by the user, rather than being fixed by the integrated circuit manufacturer. This allows a single chip to be programmed to implement different LSI-type functions such as logic gates, adders and registers. Current devices named FPGAs (Field Programmable Gate Arrays) can now implement tens of thousands of LSI circuits in parallel and operate up to 550 MHz. The techniques perfected by the integrated circuits industry over the last three decades have been used to create microscopic machines, known as MEMS. These devices are used in a variety of commercial and military applications. Example commercial applications include DLP projectors, inkjet printers, and accelerometers used to deploy automobile airbags. In the past, radios could not be fabricated in the same low-cost processes as microprocessors. But since 1998, a large number of radio chips have been developed using CMOS processes. Examples include Intel's DECT cordless phone, or Atheros's 802.11 card. Future developments seem to follow the multi-core multi-microprocessor paradigm, already used by the Intel and AMD dual-core processors. Intel recently unveiled a prototype, "not for commercial sale" chip that bears a staggering 80 microprocessors. Each core is capable of handling its own task independently of the others. This is in response to the heat-versus-speed limit that is about to be reached using existing transistor technology. This design provides a new challenge to chip programming. X10 is the new open-source programming language designed to assist with this task.
IC Terminology
An integrated circuit (IC) is a combination of interconnected circuit elements inseparably associated on a continuous substrate. The substrate is the supporting material upon which an IC is fabricated. A monolithic IC is an IC, all of whose elements are formed on the same semiconductor substrate. A wafer is a thin disk of silicon crystalline material on which identical ICs are built. The chip is one of the repeated ICs on the wafer. A test plug, or process control monitor (PCM), is a special chip that is repeated only a few times on each wafer. It is used to monitor the process parameters of the technology. If the measurements of key parameters at the test plug level are not acceptable, the wafer is discarded. A test cell is a special chip repeated only a few times on each wafer. It differs from the test plug in that the circuit designer includes this cell specifically to monitor the performance of elementary sub-circuits. A wafer yield is the percentage of chips on a wafer that meet the performance specifications.
Design Specification
Behavioral Description
Functional Verification
No
OK?
Gate-Level Netlist
Logic Verification
No
OK?
Layout Verification
No
OK?
Design Specification
Schematic Entry
Simulation
Layout Verification
Fabrication
Measurements 10
Covalent Bonds
A silicon atom residing in isolation contains four valence electrons [Figure 6(a)], requiring another four to complete its outermost shell. If processed properly, the silicon material can form a crystal wherein each atom is surrounded by exactly four others [Figure 6(b)]. As a result, each atom shares one valence electron with its neighbors, thereby completing its own shell and those of the neighbors. The bond thus formed between atoms is called a covalent bond to emphasize the sharing of valence electrons. The uniform crystal depicted in Figure 6(b) plays a crucial role in semiconductor devices. But, does it carry current in response to a voltage?
Figure 6: (a) Silicon atom, (b) covalent bond between atoms, (c) free electron. 11
At temperatures near absolute zero, the valence electrons are confined to their respective covalent bonds, refusing to move freely. In other words, the silicon crystal behaves as an insulator for T 0K. However, at higher temperatures, electrons gain thermal energy, occasionally breaking away from the bonds and acting as free charge carriers [Figure 6(c)] until they fall into another incomplete bond. We will hereafter use the term electrons to refer to free electrons.
Holes
When freed from a covalent bond, an electron leaves a void behind because the bond is now incomplete. Called a hole, such a void can readily absorb a free electron if one becomes available. Thus, we say an electron-hole pair is generated when an electron is freed, and an electron-hole recombination occurs when an electron falls into a hole. Why do we bother with the concept of the hole? After all, it is the free electron that actually moves in the crystal.
Figure 7: Movement of electrons through crystal. To appreciate the usefulness of holes, consider the time evolution illustrated in Figure 7. Suppose covalent bond number 1 contains a hole after losing an electron some time before t = t1. At t = t2, an electron breaks away from bond number 2 and recombines with the hole in bond number 1. Similarly, at t = t3, an electron leaves bond number 3 and falls into the hole in bond number 2. Looking at the three snapshots, we can say one electron has traveled from right to left, or, alternatively, one hole has moved from left to right. This view of current flow by holes proves extremely useful in the analysis of semiconductor devices.
Bandgap Energy
We must now answer an important question. Does any thermal energy create free electrons (and holes) in silicon? No, in fact, a minimum energy is required to dislodge an electron from a covalent bond. Called the bandgap energy and denoted by Eg, this minimum is a fundamental property of the material. For silicon, Eg = 1.12 eV.
12
Recall from Figure 5 that phosphorus (P) contains five valence electrons. What happens if some P atoms are introduced in a silicon crystal? As illustrated in Figure 8, each P atom shares four electrons with the neighboring silicon atoms, leaving the fifth electron unattached. This electron is free to move, serving as a charge carrier. Thus, if N phosphorus atoms are uniformly introduced in each cubic centimeter of a silicon crystal, then the density of free electrons rises by the same amount.
Figure 8: Loosely-attached electron with phosphorus doping. The controlled addition of an impurity such as phosphorus to an intrinsic semiconductor is called doping, and phosphorus itself a dopant. Providing many more free electrons than in the intrinsic state, the doped silicon crystal is now called extrinsic, more specifically, an n-type semiconductor to emphasize the abundance of free electrons. As remarked earlier, the electron and hole densities in an intrinsic semiconductor are equal. But, how about these densities in a doped material? It can be proved that even in this case,
where n and p respectively denote the electron and hole densities in the extrinsic semiconductor. The quantity ni represents the densities in the intrinsic semiconductor (hence the subscript i) and is therefore independent of the doping level. It is justified to call electrons the majority carriers and holes the minority carriers in an n-type semiconductor. We may naturally wonder if it is possible to construct a ptype semiconductor, thereby exchanging the roles of electrons and holes. Indeed, if we can dope silicon with an atom that provides an insufficient number of electrons, then we may obtain many incomplete covalent bonds. For example, the table in Figure 5 suggests that a boron (B) atomwith three valence electronscan form only three complete covalent bonds in a silicon crystal (Figure 9). 13
Figure 9: Available hole with boron doping. As a result, the fourth bond contains a hole, ready to absorb a free electron. In other words, N boron atoms contribute N boron holes to the conduction of current in silicon. The structure in Figure 9 therefore exemplifies a p-type semiconductor, providing holes as majority carriers. The boron atom is called an acceptor dopant. If an intrinsic semiconductor is dope with a density of ND ( >> ni ) donor atoms per cubic centimeter, then the mobile charge densities are given by,
Figure 10 summarizes the concepts introduced in this section, illustrating the types of charge carriers and their densities in semiconductors.
Exercises
Formula for calculating the number of devices that can be fabricated on a silicon wafer is,
r2 N= Area
Where N = Number of devices
Solution:
15
Question 2: Assume an operational amplifier (opamp) requires an area 100 mils x 100 mils and a microprocessor requires an area 1cm x 1cm; a. How many of each type of ICs can be fabricated on a 5 inch wafer? b. If the yield for the opamp is 98% and that for the microprocessor is 30%, compare the average number of good chips per wafer for both ICs. c. If the fabrication cost per wafer is $400, what is the effective cost per good chip for each type of IC. Conversions 1 mil = 0.001 in 1 cm = 10-2 m 1 in = 2.54 x 10 -2 m Solution: Symbols mil = milli-inch m = meter in = inch cm = centi-meter
16