Revision as of 11:46, 6 September 2024

Individually packaged MOSFETs

MOSFET (Metal-Oxide-Semiconductor Field-Effect-Transistor), also known as the MOS transistor (metal–oxide–silicon transistor) or IGFET (Insulated-Gate Field-Effect Transistor), is a type of insulated-gate field-effect transistor that is fabricated by the controlled oxidation of a semiconductor (typically silicon) and utilizes an insulator (such as SiO₂) between the gate and the body. Today, the MOSFET is the most common type of transistor for both digital and analog circuits. The voltage of the gate terminal determines the electrical conductivity of the device; this ability to change conductivity with the amount of applied voltage can be used for amplifying or switching electronic signals.

The MOSFET was invented by Mohamed M. Atalla and Dawon Kahng at Bell Labs in 1959, and first presented in 1960. It is the basic building block of modern electronics, and the most frequently manufactured device in history, with an estimated total of 13 sextillion (10²²) MOSFETs manufactured between 1960 and 2018.^[1] It is the dominant semiconductor device in digital and analog integrated circuits (ICs),^[2] and the most common power device.^[3] It is a compact transistor that has been miniaturised and mass-produced for a wide range of applications, revolutionizing the electronics industry and the world economy, and being central to the digital revolution, silicon age and information age. MOSFET scaling and miniaturization has been driving the rapid exponential growth of electronic semiconductor technology since the 1960s, and enables high-density ICs such as memory chips and microprocessors.

A key advantage of a MOSFET is that it requires almost no input current to control the load current, when compared with bipolar junction transistors (BJTs). In an enhancement mode MOSFET, voltage applied to the gate terminal can increase the conductivity from the "normally off" state. In a depletion mode MOSFET, voltage applied at the gate can reduce the conductivity from the "normally on" state.^[4] MOSFETs are also capable of high scalability, with increasing miniaturization, and can be easily scaled down to smaller dimensions. They also have faster switching speed (ideal for digital signals), much smaller size, consume significantly less power, and allow much higher density (ideal for large-scale integration), compared to BJTs. MOSFETs are also cheaper and have relatively simple processing steps, resulting in high manufacturing yield.

MOSFETs can either be manufactured as part of MOS integrated circuit chips or as discrete MOSFET devices (such as a power MOSFET), and can take the form of single-gate or multi-gate transistors. Since MOSFETs can be made with either p-type or n-type semiconductors (PMOS or NMOS logic, respectively), complementary pairs of MOSFETs can be used to make switching circuits with very low power consumption: CMOS (Complementary MOS) logic.

The name "metal–oxide–semiconductor" (MOS) typically refers to a metal gate, oxide insulation, and semiconductor (typically silicon).^[5] However, the "metal" in the name MOSFET is sometimes a misnomer, because the gate material can also be a layer of polysilicon (polycrystalline silicon). Because originally the gate was made from metal, the name metal-oxide semiconductor (MOS) stuck, as today the gates are typically made of polycrystalline sillicon, although in recent years, advancements in technology reintroduced metallic gates in order to solve various performance issues. Along with oxide, different dielectric materials can also be used with the aim of obtaining strong channels with smaller applied voltages. The MOS capacitor is also part of the MOSFET structure.

Overview

Process

Further information: doping, n-type semiconductor, and p-type semiconductor

The basic starting material for most integrated circuits based on MOSFET technology is typically Silicon (Si); though other processes such as silicon-germanium (Si_1−xGe_x) also exist. Silicon is a very brittle metalloid. It has the same structure as a diamond in elemental form (the actual structure is a 3D tetrahedral, just like carbon). It's a Group IV element - each silicon forming single covalent bonds with four adjacent silicon atoms. Because all of its valence electrons are involved in chemical bonds, pure silicon is quite a poor conductor of electricity. It's possible to raise the conductivity of silicon by introducing impurities, known as dopants, into the silicon lattice through a processes known as doping. Similar results can also be achieved by adding group V elements (which have 5 bonding electrons vs 4 for Si) such as phosphorus or arsenic. By inserting those group V elements into the silicon lattice it can still bond to the 4 original silicon atoms neighbors. The 5th valence electron is loosely bound to that group V element. The thermal vibrations is enough make that electron free to move - leaving positive ions and a free electron. It is this free electron that can carry current thereby increasing the conductivity of the lattice. The process forms a new semiconductor called an n-type semiconductor. The processes can be done with a group III element as well. This creates a situation where each atom is now short by an electron. The missing electron (or hole) propagates about the lattice. The hole acts as a positive carrier gaining the name p-type semiconductor.

A diode is the junction between p-type semiconductors and n-type semiconductors. When the voltage on the p-type semiconductors, known as anode, is raised above the n-type semiconductors, known as a cathode, the diode is said to be forward biased. When that happens, current flows. When the anode voltage is equal to or less than the cathode voltage, the diode is reverse biased - at which point very little current flows. Varying the voltage between the gate and body modulates the conductivity of this layer effectively controlling the current flow between drain and source.

MOS sandwich-like structure is created by superimposing several layers of conducting and insulating together. The actual process involves oxidation of the silicon, doping of the silicon using dopants, and etching of metal wires and contacts. Transistors are built on a pure and flawless single crystals of silicon. Each transistor consists of a body - the silicon wafer. The body is often grounded, often considered the reference node. Each transistor has a stack of the conducting gate that sits on top of an isolating glass (SiO2), and the substrate (also known as the body).

n-type semiconductor.

An nMOS transistor is built with a p-type body with two regions of n-type semiconductor adjacent to the gate called the source and the drain. For all practical purposes they are physically equivalent and can be used interchangeably. A pMOS transistor is built with an n-type body with two regions of p-type semiconductors adjacent to the gate.

Controlling Gate

Both pMOS and nMOS have a controlling gate. The controlling gate, as the name implies, controls the flow of electrons between the source and drain.

nMOS gate behavior

Main article: nMOS transistor

In the nMOS transistor, since the body is grounded, the p-n junctions of the source and drain to body are reverse-biased. If the voltage at the gate is raised, an electric field starts to build up - attracting free electrons to the underside of the Si-SiO2 interface. When the voltage is high enough, the electrons end up filling all the holes and a thin region under the gate called the channel gets inverted to act as an n-type semiconductor - creating a conducting path from the source to the drain, allowing current to flow. When the transistor is at that state, we say the transistor is ON. If the gate is grounded, little to no current flows through the reverse-biased junction. When that happens, we say the transistor is OFF.

pMOS gate behavior

Main article: pMOS transistor

p-type semiconductor.

In the pMOS transistor, the behavior and setup is the complement of the nMOS transistor. The body is held at positive voltage. When the gate is also positive - the source and drain are reverse-biased. When that happens, no current flows and we say the transistor is OFF.

When the voltage at the gate is lowered, positive charges are attracted to the underside of the Si-SiO2 interface. When the voltage gets sufficiently low the channel gets inverted - creating a conducting path from the source to the drain, allowing current to flow. Because the behavior of a pMOS transistor is the opposite of that of an nMOS transistor, the symbol for pMOS transistor is identical to that of nMOS with an additional bubble on the gate. That bubble is known as an inversion bubble.

When dealing with digital logic there are generally only have two distinct values - ON and OFF, 1 and 0, or HIGH and LOW. The positive voltage of the transistor is called VDD (or POWER or PWR). VDD represents the logic 1 value in digital circuits. In TTL logic, the VDD voltage levels were usually around 5 volts. Today's transistors cannot really withstand such high voltages - they are typically in the 1.5V to 3.3V range. The low voltage is often called GROUND (or GND or VSS). VSS represents the logic 0. It is also normally set to 0 volts.

Modes of operation

This section is empty; you can help add the missing info by editing this page.

Symbols

There is no one standard that is commonly accepted across all organizations. Typically individual design groups adopt their own notations. Generally speaking, however, most follow the designs shown below which consists of a lone for the channel with the source and drain leaving it at right angles and then bending back at right angles outwards. Each device has a gate (G), drain (D), and a source (S), A distinction is sometimes made for enhancement mode where the channel is broken down into three smaller lines; it is sometimes drawn as a dotted line instead as well. Depletion mode devices have a solid line instead, this is due to the channel existing prior to power being applied.

Sometimes a fourth terminal for the body (B) is show. In discrete MOSFETs, the body lead is connected internally to the source. When that's the case, the body is also omitted from the symbol. When the bulk is shown, it is sometimes sometimes angled to meet up with the source. In ICs with a common bulk, the bulk is typically not shown and instead an inverting bubble is used to represent a PMOS.

Channel	Depletion MOSFET		Enhancement MOSFT
	W/ bulk	W/O bulk	W/ bulk	W/O bulk
N-type
P-type

Early history

Background

Main article: Field-effect transistor

The basic principle of the field-effect transistor (FET) was first proposed by Austrian physicist Julius Edgar Lilienfeld in 1926, when he filed the first patent for an insulated-gate field-effect transistor.^[6] Over the course of next two years he described various FET structures. In his configuration, aluminum formed the metal and aluminum oxide the oxide, while copper sulfide was used as a semiconductor. However, he was unable to build a practical working device.^[7] The FET concept was later also theorized by German engineer Oskar Heil in the 1930s and American physicist William Shockley in the 1940s.^[8] There was no working practical FET built at the time, and none of these early FET proposals involved thermally oxidized silicon.^[7]

Semiconductor companies initially focused on bipolar junction transistors (BJTs) in the early years of the semiconductor industry. However, the junction transistor was a relatively bulky device that was difficult to manufacture on a mass-production basis, which limited it to a number of specialised applications. FETs were theorized as potential alternatives to junction transistors, but researchers were unable to build practical FETs, largely due to the troublesome surface state barrier that prevented the external electric field from penetrating into the material.^[9] In the 1950s, researchers had largely given up on the FET concept, and instead focused on BJT technology.^[10]

Invention

Mohamed M. Atalla at Bell Labs was dealing with the problem of surface states in the late 1950s. He attempted to passivate the surface of silicon through the formation of oxide layer over it. He thought that growing a very thin high quality thermally grown Template:chem2 on top of a clean silicon wafer would neutralize surface states enough to make a practical working field-effect transistor. He wrote his findings in his BTL memos in 1957, before presenting his work at an Electrochemical Society meeting in 1958.^{[11][12][13][14][8]} This was an important development that enabled MOS technology and silicon integrated circuit (IC) chips.^[15]

The MOSFET was invented when Mohamed Atalla and Dawon Kahng^[12][11] successfully fabricated the first working MOSFET device in November 1959.^[16] The device is covered by two patents, each filed separately by Atalla and Kahng in March 1960.^{[17][18][19][20]} They published their results in June 1960,^[21] at the Solid-State Device Conference held at Carnegie Mellon University.^[22] The same year, Atalla proposed the use of MOSFETs to build MOS integrated circuit (MOS IC) chips, noting the MOSFET's ease of fabrication.^[9]

Commercialization

The advantage of the MOSFET was that it was relatively compact and easy to mass-produce compared to the competing planar junction transistor,^[23] but the MOSFET represented a radically new technology, the adoption of which would have required spurning the progress that Bell had made with the bipolar junction transistor (BJT). The MOSFET was also initially slower and less reliable than the BJT.^[24]

In the early 1960s, MOS technology research programs were established by Fairchild Semiconductor, RCA Laboratories, General Microelectronics (led by former Fairchild engineer Frank Wanlass) and IBM.^[25] In 1962, Steve R. Hofstein and Fred P. Heiman at RCA built the first MOS integrated circuit chip. The following year, they collected all previous works on FETs and gave a theory of operation of the MOSFET.^[26] CMOS was developed by Chih-Tang Sah and Frank Wanlass at Fairchild in 1963.^[27] The first CMOS integrated circuit was later built in 1968 by Albert Medwin.^[28]

The first formal public announcement of the MOSFET's existence as a potential technology was made in 1963. It was then first commercialized by General Microelectronics in May 1964, followed Fairchild in October 1964. GMe's first MOS contract was with NASA, which used MOSFETs for spacecraft and satellites in the Interplanetary Monitoring Platform (IMP) program and Explorers Program.^[25] The early MOSFETs commercialized by General Microelectronics and Fairchild were p-channel (PMOS) devices for logic and switching applications.^[8] By the mid-1960s, RCA were using MOSFETs in their consumer products, including FM radio, television and amplifiers.^[29] In 1967, Bell Labs researchers Robert Kerwin, Donald Klein and John Sarace developed the self-aligned gate (silicon-gate) MOS transistor, which Fairchild researchers Federico Faggin and Tom Klein adapted for integrated circuits in 1968.^[30]

MOS revolution

Main article: MOS revolution

The development of the MOSFET led to a revolution in electronics technology, called the MOS revolution^[31] or MOSFET revolution,^[32] fuelling the technological and economic growth of the early semiconductor industry.

The impact of the MOSFET became commercially significant from the late 1960s onwards.^[33] This led to a revolution in the electronics industry, which has since impacted daily life in almost every way.^[34] The invention of the MOSFET has been cited as the birth of modern electronics^[35] and was central to the microcomputer revolution.^[36]

See also

• Field-effect transistor
• CMOS
• pMOS logic
• nMOS logic

References