What are semiconductors made of



Under a semiconductor one understands a solid body, which can be regarded as both a conductor and a non-conductor in terms of its electrical conductivity. The conductivity is strongly temperature dependent. Semiconductor insulators are in the vicinity of absolute temperature zero. At room temperature, they are conductive or non-conductive depending on the material-specific distance between the conduction and valence bands. The electrical conductivity of semiconductors increases with increasing temperature, so they belong to the NTC thermistors.

The conductivity can be specifically influenced within wide limits by introducing foreign atoms (doping) from another main chemical group.

Semiconductors are important for electrical engineering and especially for electronics, where their conductivity can be changed by applying a control voltage or a control current (e.g. with a transistor) to suitable structures or they have a direction-dependent conductivity (diode, rectifier).
Applications are NTC thermistors, varistors, radiation sensors (photoconductors, photoresistors, photodiodes or solar cells), thermoelectric generators, Peltier elements and radiation or light sources (laser diode, light-emitting diode).

Semiconductors are used in single crystal, polycrystalline and amorphous forms.

Basics and characteristics

The semiconductor properties of the substances mentioned above are due to their chemical bonds and thus their atomic structure. The semiconductors crystallize in different structures, silicon and germanium crystallize in the diamond structure (purely covalent bond), III-V and II-VI compound semiconductors, on the other hand, mostly in the zincblende structure (mixed covalent-ionic bond).

The basic properties of semiconductors can be explained using the ribbon model: The electrons in solids interact with each other over a large number of atomic distances. In fact, this leads to an expansion of the possible energy values ​​(which are still present in the individual atom as discrete levels) to form extended energy ranges, the so-called energy bands. Since the energy bands are different from each other depending on the expansion and atomic type, bands can overlap or be separated by energy ranges in which, according to quantum mechanics, no permitted states exist (energy or band gap).

In semiconductors, due to their crystal structure, the highest occupied energy band (valence band) and the next higher band (conduction band) are separated by a band gap. The Fermi level is therefore exactly in the band gap. At a temperature close to absolute zero, the valence band is fully occupied and the conduction band is completely empty of charge carriers. Since unoccupied bands do not conduct electrical current due to a lack of mobile charge carriers and charge carriers in fully occupied bands cannot absorb energy due to a lack of free states that can be reached, semiconductors do not conduct electrical current at a temperature close to absolute zero (insulators).

Partially occupied strips are necessary for the conduction process, which can be found in metals and semimetals by overlapping the outer strips at any temperature. As mentioned above, this is not the case with semiconductors and insulators. The band gap ("forbidden band" or "forbidden zone") in semiconductors is, in contrast to insulators (EG > 3 eV) but relatively small (InAs: ~ 0.4 eV, Ge: ~ 0.7 eV, Si: ~ 1.1 eV, GaAs: ~ 1.4 eV, diamond: ~ 5.45 eV), see above that for example electrons can be excited from the fully occupied valence band to the conduction band by the energy of the heat oscillations at room temperature or by the absorption of light. Semiconductors therefore have an intrinsic electrical conductivity that increases with temperature. That is why semiconductors are also called NTC resistors. However, the transition from semiconductors to insulators is fluid. For example, gallium nitride (GaN; used in blue LEDs) with a band gap energy of ~ 3.6 eV is also counted among the semiconductors.

If, as described above, an electron in a semiconductor is excited from the valence band into the conduction band, it leaves a defect electron, also known as a “hole”, at its original location. Bound valence electrons in the vicinity of such holes can “jump” into a hole by changing their place, and the hole moves. It can therefore be viewed as a mobile positive charge.

Electrons from the conduction band can recombine with the defect electrons (electron-hole recombination). This transition between the levels involved can take place with the emission of electromagnetic recombination radiation (photon) and / or with the emission of a pulse to the crystal lattice (phonon).

Both the excited electrons and the defect electrons thus contribute to electrical conduction.

Direct and indirect semiconductors

 

Semiconductors are divided into two groups, the direct and the indirect Semiconductor. Their different properties can only be understood by looking at the band structure in the so-called momentum space: Free charge carriers in the semiconductor can be understood as matter waves with a quasi-momentum, that is, the charge carriers are determined not only by their energy level in the band scheme but also by their "speed" (momentum = mass · Speed).

If one now looks at the band model in the momentum space, one finds that the line and valence band edges are not the same for every pulse, but that both band edges have at least one extreme. If an electron is now excited from the valence band into the conduction band, it is energetically most favorable (and therefore most likely) when it is excited from the maximum of the valence band to the minimum of the conduction band.

If these extremes are now (almost) at the same quasi-impulse, an excitation z. B. easily possible by a photon, since the electron only has to change its energy, but not its momentum. One speaks of one direct semiconductor. However, if the extremes are at different quasi-impulses, the electron must change its momentum in addition to its energy in order to be excited into the conduction band. This impulse cannot come from a photon (which has a very small impulse), but has to be contributed by a lattice oscillation (also called phonon).

In principle, the same applies to the recombination of electron-hole pairs. In a direct semiconductor, a light quantum can be emitted during recombination. In the case of an indirect semiconductor, on the other hand, the energy released during recombination is emitted as a lattice oscillation. It follows from this that only direct semiconductors can be used to effectively generate radiation.

Direct and indirect semiconductors can be distinguished from one another by means of an absorption test.

As a rule, element semiconductors (Si, Ge) and compound semiconductors from main group IV are indirect and compound semiconductors from various main groups (III / V: GaAs, InP, GaN) are direct.

It can happen that a semiconductor conducts worse than before after the excitation of the electron in the conduction band, i.e. the current falls despite the increasing voltage. This effect is called the Gunn effect.

Intrinsic semiconductors and impurity semiconductors

The density of free electrons and holes in pure, i.e. H. undoped semiconductors, are called intrinsic charge carrier density or intrinsic conduction density - an intrinsic semiconductor is therefore also called an intrinsic semiconductor. If, on the other hand, the concentration of electrons in the conduction band is mainly determined by the dopant, one speaks of an impurity semiconductor or extrinsic semiconductor.

history

Stephen Gray discovered the difference between conductors and non-conductors in 1727. After Georg Simon Ohm established Ohm's law in 1821, which describes the proportionality between current and voltage in an electrical conductor, the conductivity of an object could also be determined.

The Nobel laureate Ferdinand Braun discovered the rectifying effect of semiconductors in 1874. He wrote: “... with a large number of natural and artificial sulfur metals [...] the resistance of the same was different with the direction, intensity and duration of the current. The differences amount to up to 30% of the total value. ”He thus described for the first time that the ohmic resistance can be variable.

Greenleaf Whittier Pickard received a patent in 1906 (U.S. Patent 836,531[1]) for a silicon-based tip diode for demodulating the carrier signal in a detector receiver[2]. Initially, in the receiver of the same name ("Pickard Crystal Radio Kit"), galena was mostly used as a semiconductor, whereby more robust and more powerful diodes based on copper sulfide-copper contacts were created in the 1920s. The functioning of the rectifier effect based on a semiconductor-metal transition remained unexplained for decades, despite its technical application. It was not until 1939 that Walter Schottky was able to lay the theoretical foundations for describing the Schottky diode named after him.

When scientists John Bardeen, William Bradford Shockley and Walter Houser Brattain at Bell Laboratories in 1947 stuck two metal wire tips onto the germanium plate and were thus able to control the p-conductive zone with the second wire tip with an electrical voltage, the transistor was invented. This earned them the 1956 Nobel Prize in Physics and founded microelectronics.

Alan Heeger, Alan MacDiarmid and Hideki Shirakawa showed in 1976 that when polyacetylene - a polymer that is an insulator in the undoped state - is doped with oxidizing agents, the specific electrical resistance is up to 10−5 Ω m (silver: ~ 10−8 Ωm) can decrease. In 2000 they received the Nobel Prize in Chemistry for this (see section organic semiconductors).

Different semiconductors

Semiconductors used in microelectronics can be classified into two groups, the Element semiconductors and the Compound semiconductors. Element semiconductors include elements with four valence electrons, for example silicon (Si) and germanium (Ge). The group of compound semiconductors includes chemical compounds that have an average of four valence electrons. These include elements of III. and V. main group of the periodic table (III-V semiconductors), such as gallium arsenide (GaAs) or indium antimonide (InSb), and the II. and VI .. main group (II-VI semiconductors), such as zinc selenide (ZnSe) or cadmium sulfide (CdS).

In addition to these frequently used semiconductors, there are also those I-VII semiconductorssuch as copper (II) chloride. Materials that do not have four valence electrons on average can also be called semiconductors if they have a specific resistance in the range of greater than 10−4 Ω m and less than 106 Have Ω m. A group of promising new semiconductors are, for example, organic semiconductors, which have already been used in organic solar cells or in organic field effect transistors.

Chemical classification

Elementary semiconductors Compound semiconductors Organic semiconductors
Ge, Si, α-Sn, C (fullerenes), B, Se, TeIII-V: GaP, GaAs, InP, InSb, InAs, GaSb, GaN, AlN, InN, AlxGa1-xAsTetracene, pentacene, phthalocyanines, polythiophenes, PTCDA, MePTCDI, quinacridone, acridone, indanthrone, flavanthrone, perinone
Under pressure: Bi, Ca, Sr, Ba, Yb, P, S, III-VI: ZnO, ZnS, ZnSe, ZnTe, CdS, CdSe, CdTe, Hg (1-x) Cd (x) Te, BeSe, BeTe, HgS Mixed systems: polyvinyl carbazole, TCNQ complexes
  III-VI: GaS, GaSe, GaTe, InS, InSe, InTe ....  
  I-III-VI: CuInSe2, CuInGaSe2, CuInS2, CuInGaS2....  

Semimagnetic semiconductors

Semimagnetic semiconductors belong to the important group of Compound semiconductors or Compound Semiconductors. These are compounds in which an ion is replaced by e.g. B. Manganese was replaced. A characteristic property of these semi-magnetic semiconductors is the large Zeeman effect. Actually one calls semimagnetic semiconductors diluted magnetic semiconductorsbecause they are magnetically thinned.

Organic semiconductors

In general, organic materials are electrically insulating. If molecules or polymers have a conjugated bond system consisting of double bonds, triple bonds and aromatic rings, these can also become electrically conductive. This was first observed in 1976 with polyacetylene[3]. Polyacetylene is a linear polymer with an alternating double bond and a single bond (... C = C − C = C − C ...). If a donor such as chlorine, bromine or iodine is added to this plastic (Oxidative doping), there are additional electrons. By adding an atom such as sodium (Reductive doping) the plastic receives an acceptor. As a result of this chemical change, the double bonds break and a continuous conduction band is created: the originally non-conductive polymer becomes electrically conductive. If molecules or polymers have semiconducting properties even in the undoped state, they are referred to as intrinsic conductivity (e.g. pentacene or poly (3-hexylthiophene), as with inorganic semiconductors). See also category Organic semiconductor).

When the plastic is made in the form of a thin layer (about 5 nm – 1 µm thick), it is ordered enough to form an electrically uninterrupted layer.

Semiconductor technology

Semiconductor technology deals with the technical production of microelectronic components and assemblies. The prerequisite is the knowledge of how the semiconductor must be processed in order to show the desired electrical behavior. This includes doping the semiconductor and designing the interface between the semiconductor and another material.

Doping and impurity conduction

When foreign atoms are built into the crystal lattice of a semiconductor material, this is called doping. Doping increases the conductivity of the pure semiconductors. In addition, different components such. B. a bipolar transistor can be produced.

Doping generates mobile charge carriers. If foreign atoms, which have one electron more in the valence band than the pure semiconductor, are introduced into a semiconductor (n-doped semiconductor), each of these foreign atoms brings with it an electron that is not required for the bond and can be easily detached. In the band scheme, such an electron is located close to the edge of the conduction band. A foreign atom that releases an electron is called a donor (Latin donare = to give). Similarly, foreign atoms that have one electron less in the valence band bring an additional defect electron (hole) with (p-doping), which can easily be occupied by valence band electrons. In the band scheme, such a hole is close to the valence band edge. A foreign atom that "releases" a hole, that is, accepts an electron, is called an acceptor (Latin: accipere = to accept).

In the case of doping with donors, it is mainly the electrons in the conduction band that ensure electrical conductivity, in the case of doping with acceptors the imaginary, positively charged holes in the valence band. In the first case one speaks of electron conduction or n-conduction (n-> negative), in the other case of hole conduction or p-conduction (p-> positive). Semiconductor areas with excess electrons are referred to as n-doped, those with deficiency, ie with "excess holes", as p-doped.

Has the temperature been increased to such an extent that all doping atoms are ionized, i. H. contribute to the line, one speaks of impurity exhaustion. The self-conduction temperature represents the transition between the impurity depletion and the self-conduction of the semiconductor.

In the n-conductor, the electrons are called majority carriers (charge carriers present in the majority), the holes are called minority carriers, in the p-conductor the corresponding reversal applies.

By cleverly combining n- and p-doped areas (see p-n junction), it is possible to build individual, so-called discrete semiconductor components such as diodes and transistors and complex integrated circuits or microchips made up of many components in a single crystal.

Usual doping levels

  • Normal doping:
    • n-conductive: 1 donor on 107 Si atoms
    • p-conducting: 1 acceptor on 106 Si atoms
  • Heavy doping:
    • n-conductive: 1 donor on 104 Si atoms (n+)
    • p-conducting: 1 acceptor on 104 Si atoms (p+)

Interfaces

The combination of a p-doped and an n-doped semiconductor creates a p-n junction at the interface. The combination of a doped semiconductor with a metal (e.g. Schottky diode) or a dielectric is also of interest, and if two semiconductors, e.g. gallium arsenide and aluminum gallium arsenide, are superposed, then a heterojunction results. Not only p-n junctions are important, but also p-p junctions and n-n junctions, the so-called isotypic heterojunctionsthat are used in a quantum well, for example.

Efforts have recently been made to combine semiconductors, superconductors, and silicon and III-V semiconductors on one chip. Since the crystal structures are not compatible, breaks and lattice defects occur in the interface if it is not possible to find suitable materials for an intermediate layer that is a few atomic layers thick and in which the lattice spacing can be adjusted.

economy

The main producers of silicon (about 10,000 tons) are Hemlock Semiconductor and MEMC Electronic Materials from the USA, Renewable Energy Corporation AS (REC) from Norway, Tokuyama from Japan and Brave from Germany (as of 2005).

The Japanese company is the world's largest manufacturer of wafers Shin-etsu Handotai (SEH) with 2004 sales of $ 2.2 billion. The world's second largest Japanese manufacturer Sumitomo Mitsubishin Silicon Corp. (Sumco) had sales of $ 1.6 billion in the same year. This is followed by the American company MEMC Electronic Materials and the German Siltronic AG with $ 1 billion and $ 900 million. These four companies share about 75% of the total $ 7.3 billion wafer market.

See also

literature

  • Robert F. Pierret: "Semiconductor Device Fundamentals", Addison-Wesley 1996, ISBN 0131784595
  • Peter Y. Yu, Manuel Cardona: Fundamentals of Semiconductors: Physics and Materials Properties, Springer 2004, 3rd edition, ISBN 3540413235
  • Marius Grundmann: The Physics of Semiconductors. An Introduction Including Device and Nanophysics, Springer 2006, 1st edition, ISBN 354025370X
  • Simon M. Sze: Physics of Semiconductor Devices. John Wiley and Sons (WIE) 1981, 2nd edition, ISBN 0471056618
  • S.M. Sze, Kwok K. Ng: Physics of Semiconductor DevicesJohn Wiley & Sons 2006, 3rd edition, ISBN 0-471-14323-5
  • Michael Reisch: Semiconductor components, Springer 2004, ISBN 3540213848
  • Ulrich Tietze, Christoph Schenk, Eberhard Gamm: Semiconductor circuit technology, Springer 2002, 12th edition, ISBN 3540428496
  • Ulrich Hilleringmann: Silicon semiconductor technology, Teubner 2004, ISBN 3519301490
  • Bernhard Hoppe: Microelectronics 1, Vogel Fachbuch Kamprath series, 1997. 154 p., ISBN 3802315189
  • Bernhard Hoppe: Microelectronics 2, Vogel Fachbuch Kamprath series, 1997. 320 p. With 225 figs., ISBN 380231588X
  • Werner Gans: The art of electrifying plastics. Nobel Prize in Chemistry 2000. In: Spectrum of science No. 12, 2000, pp. 16-19.

swell

  1. Pickard, Greenleaf Whittier: Means For Receiving Intelligence Communicated By Electric Waves. 11/20/1906. Publication No. US 836531
  2. Margolin, Jed: The Road to the Transistor. URL: [1]. - Date updated: 2005-10-09. - Review date 2007-02-22
  3. C. K. Chiang et al .: Electrical Conductivity in Doped Polyacetylenes. In: Physical Review Letters 39, 1977, pp. 1098-1101.

Categories: Semiconductors | Electrotechnical material