Fundamentals of microelectronics (course of lectures). Technological foundations of microelectronics

Useful tips

Home > Lecture course

BASICSMICROELECTRONICS

(lecture course)

Lecture 1 (bring literature and exhibits to the lecture)

Section 1. BASIC PROVISIONS

Integrated circuits (ICs) are the elemental base of a computer. Complex computing systems contain several million elements. Therefore, the problems of increasing reliability And microminiaturization EVA.

If we tried to implement a modern computer using ordinary discrete components, as was done about 30 years ago (remember our domestic televisions), then the following would happen: one transistor in the case occupies a volume of about 1 cm 3 . Only the processor itself in this case would have a volume of 5–10 million cubic meters. cm - i.e. 5–10 cubic meters. This does not include boards, blocks, racks, etc. It would take about 20 million rations at a minimum. If we calculate the reliability (i.e. time between failures) of this system using formulas, it would be less than one second. In other words, such a computer would never have worked at all. For those who are interested, in my free time I could tell you how we worked on large EU class machines ( best car The EC is inferior to the outdated IBM-286 and works 10,000 times slower than the Pentium). So integrated circuits essentially decided the fate of computer technology.

Now about literature. We will not be studying the latest IC developments (understanding modern microelectronics requires other physics that you have not studied), and any book published after 1980 will be useful to you. But some are especially useful:

1. I. P. Stepanenko, Fundamentals of microelectronics, - M., Sov. radio, 1980, 423 p.

2. R. Maller and T. Kamins, Elements of Integrated Circuits. - M., Mir, 1989, 632 p.

3. Yu. M. Kalnibolotsky, Yu. V. Korolev, G. I. Bogdan, V. S. Rogoza, Calculation and design of microcircuits. - Kyiv, High School, 1983, 208 p.

4. Design and technology of microcircuits. Course design (edited by L. A. Koledov). - M., Higher School, 1984, 232 p.

5. W. Till and J. Lacson, Integrated circuits - materials, devices, manufacturing, M.-Mir, 1985, 504 p.

Distinguish integral And functional microcircuit. An example of the first is an IC produced on a single semiconductor substrate, say an operational amplifier or processor. An example of the second is a piezoelectric resonator.

A short excursion into history: electronics as a science originated in the 19th century thanks to research in the field of transport physics electric charges in various environments and interactions of charges in electric and magnetic fields. There are three directions in the development of electronics (from the point of view element base):

Vacuum: electric vacuum and gas-discharge lamps, cathode ray devices, photoelectric devices, X-ray tubes;

Solid State: semiconductor devices, ICs, optoelectronic devices;

Quantum: lasers and masers, radio astronomy instruments, holographic instruments.

Integrated circuit(or simply - an integrated circuit) is a collection

interconnected components (transistors, diodes, capacitors, resistors, etc.), manufactured in a single technological cycle (i.e. simultaneously), on the same supporting structure - substrate, and performing a certain function of converting information.

Components that are part of the IP and therefore cannot be separated from it as independent products are called elements IP or integral elements. In contrast to integral elements, we will call structurally isolated elements and parts characteristic of the “pre-microelectronic” era discrete components, and electronic components and blocks built on their basis - discrete circuits.

The words “microminiaturization” and “microelectronics” should not be confused. Microelectronics is a field of electronics covering research problems,

– 2 – Lecture 1

design, manufacture and use of electronic devices with high degree integrations implemented on a single substrate, so that the IC itself can be considered as a structurally unified radio element (with all the ensuing consequences: reliability - like one element, the number of pins - like one element, etc.).

Modern problems can no longer be solved on the basis of the old element base. The main factors underlying the change in the element base are: reliability, dimensions and weight, cost and power. A simple example illustrates the reasons for the transition from the stage of transistor technology to the stage of microelectronics.

Suppose you want to build a compact electronic device containing 10 8 components. If we try to solve this problem on discrete components characterized by an average power of 15 mW, an average size (including connections) of 1 cm 3, an average mass of 1 g, an average price of 50 kopecks, and a probability of failure of 10 -5 hours -1, then the result will be the following : power dissipated by the device - 1.5 thousand kW, dimensions 100 m 3, weight 100 tons, cost (excluding installation work) - 50 million hryvnia (about 10 million dollars).

In addition, the installation of the device, even with two-shift work, will require 10 man-years, and its serial production may be prohibitive for the national economy. (Let’s forget for now that even an experienced radio installer makes 3–5 errors per 1000 rations, which he then has to look for.)

However, the most important drawback is that the average failure rate (10 –5. 10 8) turns out to be equal to 10 3 / hour, i.e. about 1 failure per 3 seconds, which, of course, indicates that the device is not working.

The technological processes used in the manufacture of ICs are of a group nature (i.e., many microcircuits are created on one wafer at once). Let's say that there are 100 such ICs, and each has 1000 transistors and passive elements. This means, roughly speaking, the number of junctions would be about 100 thousand (actually much more), and the number of errors would be at least 300. In ICs, there are no such errors, since there is no need to solder individual elements to each other.

The elemental base of electronic devices has undergone several stages of development:

First generation. Discrete electronics with a predominance of electrovacuum and gas-discharge devices. Then transistors began to play the role of active radio components. REA was assembled on separate boards or modules, which were combined into blocks, blocks into racks, racks into devices.

Second generation. The appearance in 1948 of a point bipolar transistor (Bardin and Bronstein) and then a more reliable alloy transistor (Shockley) marked the second generation of REA and EVA. IN point semiconductor devices p-n-transitions were created at the point of contact of two sharpened wires with a semiconductor. However, point contacts were unstable.

IN alloy p-n-transitions contact p- And n-regions is achieved by the interaction of the liquid phase of the fused electrode containing a doping impurity with a solid semiconductor. Although alloy devices are more stable than point devices, they are characterized by a rather low current gain (due to a thick base of the order of 10 μm) and low f g (due to fairly large containers p-n-transitions due to large areas of these transitions).

Transistors produced using processes of diffusion of impurities into a semiconductor substrate have proven to be more promising. The thickness of the base here can be 0.2–0.3 microns. At the end of the 50s originated planar technology in which ideas for obtaining were embodied p-n-transitions based on diffusion processes. This technology prepared the birth of the element base third generation- IC.

And finally, fourth generation element base REA and EVA are functional scheme.

It must be said that all of the generations listed above exist today, continuously developing. Thus, the use of printed wiring, more advanced materials, transformers, connectors, reduction in the dimensions of electronic and semiconductor devices made it possible to significantly reduce the size

– 3 – Lecture 1- 2

radio engineering systems. For example, the specific capacitance of ceramic capacitors has increased by 250–300 times over 30 years, and the dimensions of resistors have been reduced by 2–10 times. There was even such a term as microminiaturization - the implementation of electronic circuits, blocks, components and equipment in general from microminiature radio components and components. A typical solution to the problem of microminiaturization is the creation of micromodules and their subsequent sealing. If in normal printed circuit Packing density is 3–5 parts/cm 3 , then in micromodules it is 10–20 parts/cm 3 . An increase in packaging density always entails an increase in equipment reliability.

High reliability of IC and micro electronic equipment This is also explained by the use of highly pure materials, special manufacturing conditions, the tightness of the IC, and the absence of solder joints. The very small dimensions of ICs and the current consumption make them very economical devices compared to conventional electronic devices.

Constant reduction of the area occupied by one element (from 50  50 μm 2 to 1  1 μm 2 or less) allows increasing the degree of integration k= log N, Where N- the number of elements and components included in the IC.

Yes, when N = 1 ... 10  k= 0 ... 1, we get an IC of the first degree of integration,

N = 10 ... 100  k= 1 ... 2 - IC of the 2nd degree of integration,

N = 100 ... 1000  k= 2 ... 3 - IC of the 3rd degree of integration,

N = 1000 ... 10 000  k= 3 ... 4 - IC of the 4th degree of integration,

N = 10 000 ... 100 000  k= 4 ... 5 - IC of the 5th degree of integration.

The following terms are also used:

and if k  1 (N 10) - simple IC,

b) if 1  k  2 (10  N 100) - average IC (SIS),

c) if 2  k  4(100  N 10,000) - large IC (LSI),

d) if k 4 ( N 10,000) - ultra-large IC (VLSI).

An example of VLSI is microprocessor(packing density of elements is the number of elements, most often transistors, per unit area - 500–1000 elements / mm 2 or more - today it is about 100,000).

According to the method of implementation, they are distinguished: 1) semiconductor ICs; 2) film ICs, thin film (films up to 1 micron thick) and thick film (more than 1 micron); 3) hybrid ICs (film passive elements and suspended active components); 4) combined ICs (film passive elements and semiconductor active elements in a semiconductor substrate).

(End of lecture 1)

Lecture 2

Section 2. BASICS OF IC TECHNOLOGY

2.1. Main stages of IC manufacturing

Silicon (Si) is used as the main semiconductor. It is a very common element on earth (25.7% of the earth's crust by weight).

Atomic number - 14

Atomic weight - 28.06

Melting point - 1420 C

Bandgap E g = 1.12 eV (at 300 K)

Free electron mobility  n= 1350 cm 2 /(V. s) (at 300 K)

Hole mobility  p= 480 cm 2 /(V. s) (at 300 K)

Dielectric constant  = 12

Acceptors: B, Al

– 4 – Lecture 2

Donors: P, As, Sb

Resistivity  = 2.5 . 10 5 Ohm . cm (300 K)

Pair concentration n i = 1,5 . 10 10 cm -3 (300 K) (by the way, for germanium n i = 2,5 . 10 13 cm -3)

Lattice constant: 5.43 Å (300 K)

Number of atoms in 1 cubic meter cm of substance: 5 . 10 22 cm -3 .

Germanium is also used in semiconductor technology, but only for the production of discrete diodes and transistors. Germanium is not used for the production of ICs. Firstly, for Germany E g = 0.77 eV, so germanium is sensitive to temperature changes. Secondly, germanium dioxide GeO 2 can have two types of crystal lattice - cubic and hexagonal, of which the first is difficult to obtain, and the second is soluble in water. As we will see later, the oxide grown on the surface of the semiconductor plays an important role as a passivation layer. In general, GeO 2 cannot play the role of a passivating layer, but SiO 2 can. In addition, a layer of silicon nitride Si 3 N 4 can be grown on the silicon surface, which evenly lies on the surface of the semiconductor. Finally, SiO 2 plays the role of a mask due to its low temperature sensitivity and resistance to aqueous solutions.

But the most important reason is the low melting point of germanium, which does not allow high-temperature diffusion of impurities. As we will see below, in order to introduce impurities into a semiconductor in real time (even several hours), it must be heated to at least 1000 degrees. Germanium will already melt at this temperature (melting point 936 C). And it would take many months and years to drive the impurities at a temperature safe for germanium - 700  750 C.

For these reasons, silicon is more technologically advanced for the production of semiconductor ICs than germanium.

Another semiconductor for ICs is gallium arsenide GaAs (A III B V):

E g = 1.4 eV

 n= 11,000 cm 2 / (V. s)

 p= 450 cm 2 /(V . With)

 = 4. 10 8 Ohm . cm

n i= 1.5. 10 6 cm -3 .

Thanks to its wide E g it is even less sensitive to temperature than silicon. And due to the very high  n the gain factors of transistors are greater than in silicon devices, and, in addition, the frequency properties of transistors are significantly better than those based on silicon. Since gallium arsenide is a multi-valley semiconductor (explain what it is), it is used in microwave microcircuits to create generators (based on the Gunn effect). True, GaAs is much more expensive than Si. Additionally, although GaAs itself is non-toxic, its technology is highly toxic.

So, the main element of modern ICs is silicon.

Now let's look at the most simplified sequence of technological operations for manufacturing ICs (we will omit the stage of preparing the wafers - cutting, polishing, etc.). I will not list all the operations, but will give only a brief description of what has to be done. (Till and Lacson, p. 15.)

1. Circuit design

2. Design of photo masks

3. Making photo masks

4. Photolithography

5. Local introduction of impurities (4 and 5 are repeated several times)

6. Metallization

7. Metal etching

8. Control on the plate

9. Scribing (cutting the wafer into individual crystals)

10. Output control

– 5 – Lecture 2

11. Installation in the case

12. Creation of leads from the housing.

This is just a short list. Typically the number of operations is 50–200. Below we will look at only a small fragment of the technology - for the manufacture of the ICs themselves.

But first, I would like to make one point regarding point 1. Traditional methods of discrete circuit design, which make maximum use of passive elements, have to be rejected as useless, since transistors take up less space on the chip than passive components. Inductors simply don't exist, capacitors are large in area and small in capacitance, and high-value resistors require too much space.

And now - about the actual production:

To obtain a complete IC configuration, it is necessary to use 5–6 photomasks for a bipolar IC or up to 10 for MIS LSI. Making photo masks requires high precision. Then, using a computer, the original is copied on a two-coordinate plotter with a 20-fold reduction. The drawing is then reduced by another 20 times using a step-by-step animation process. The latter plays

(multiplies) the photo template 100–1000 times, clearly arranging copies in rows and columns - this is how hundreds and thousands of identical photo templates are obtained. Control areas are placed on the photomask, which make it possible to control the process of implementing the IC in stages.

Another way to obtain photomasks immediately in full size is to apply them with the beam of a scanning electron microscope, which is controlled by a computer.

– 6 – Lecture 2

After each procedure for receiving the next R- or n-layer, the semiconductor is coated with a layer of silicon dioxide SiO 2 . Then a layer of photoresist (positive or negative) and a photomask are applied to the SiO 2.

Positive FR hardens in places not irradiated by light, negative - vice versa. Next, the structure is exposed to light (exposure), the photon is removed, and the exposed structure is developed and washed. In the places where it happened tanning FR, the latter remains on the surface of the structure. For example, in the case of a positive FS, it is hardened in those places that were under the opaque sections of the FS, in the case of a negative one, vice versa. The tanned FR forms mask, which protects the SiO 2 surface during the subsequent etching operation. SiO 2 etching is carried out with hydrofluoric acid HF.

Hydrofluoric acid dissolves SiO 2 in those places that are not protected by a FR mask. This is how windows are formed in SiO 2.

Then do diffusion doping impurities into silicon, carefully maintaining the composition of the medium and temperature. So, in R-silicon for creation n-areas can use phosphorus and antimony. If the source plate is n-silicon, then to create R-areas commonly use boron. Then, after the diffusion process, the resulting structure passivate.

Everything described above is called photolithography. This process is repeated several times to form transistors, diodes, resistors and capacitors in the epitaxial layer.

– 7 – Lectures 2–3

(End of lecture 2)

Lecture 3

After all the necessary elements IC, etched in SiO 2

windows for receiving metal contacts (usually aluminum) in the right places. Aluminum is deposited in a vacuum over the entire wafer either by vapor deposition or electron beam. Excess metal is removed using a photomask, in the same way as described above, i.e. etching, and leave it in those places where required, for interconnections and contacts. This photolithographic process is sometimes called photoengraving. The resulting structure is then annealed to make the aluminum denser, and the structure is coated with a layer of passivating oxide SiO 2 .

(Using the course project and glass as an example, consider all the masks,

required for the manufacture of ICs.)

After manufacturing, the ICs are tested. Defective ICs are marked with ink. The plate is then scribed, either with a diamond cutter or a laser beam, and broken into individual crystals. Each such crystal (they are sometimes called chips) is placed in a separate case and external pins are connected to it using the thermocompression- press the hair to the contact pad and pass an electric discharge through the joint for welding.

And finally, at the final stage, the IC is sealed in the housing.

conclusions

1. n- And R-areas of a semiconductor IC are obtained as a result of operations such as epitaxy, diffusion, and some others (for example, ion implantation).

2. Selective etching and diffusion of impurities into the semiconductor substrate is carried out using a photolithographic process.

3. After each photolithography, the structure is passivated with a SiO 2 film.

2 .2. Epitaxy

Epitaxy is the process of growing single-crystal layers on a substrate in which the crystallographic orientation of the grown layer repeats the crystallographic orientation of the substrate.

Currently, epitaxy is usually used to obtain thin working layers of a homogeneous semiconductor on a relatively thick substrate, which plays the role of a supporting structure.

Typical - chloride The epitaxy process as applied to silicon is as follows. Monocrystalline silicon wafers are loaded into a crucible ("boat")

– 8 – Lecture 3

and placed in a quartz tube. A stream of hydrogen containing a small amount of silicon tetrachloride SiCl 4 is passed through the pipe. At high temperature (about 1200 C), which is provided by high-frequency heating of the crucible, a reaction occurs on the surface of the plates

SiCl 4 + 2H 2 = Si + 4HCl.

As a result of the reaction, a layer of pure silicon is gradually deposited on the substrate, and HCl vapors are carried away by the hydrogen flow. The epitaxial layer of deposited silicon is single-crystalline and has the same crystallographic orientation as the substrate. The chemical reaction, due to the selection of temperature, occurs only on the surface of the plate, and not in the surrounding space. To deposition silicon from the gas phase

was epitaxial, the wafer must be heated strongly so that the deposited silicon atoms can move to positions in which they form covalent bonds with the substrate. In this case, the atoms must have time to continue the single-crystal lattice before they are covered by the next layers of deposited atoms. This determines the epitaxy temperature - from 900 to 1250 C.

If pairs of boron (B 2 H 6) or phosphorus (PH 3) compounds are added to silicon tetrachloride vapors, then the epitaxial layer will no longer have its own conductivity, but hole or electronic conductivity, respectively, since during the reaction acceptor conductivity will be introduced into the deposited silicon. boron atoms or donor phosphorus atoms.

Thus, epitaxy makes it possible to grow single-crystal layers of any type of conductivity and any resistivity on a substrate that also has any type and value of conductivity.

The epitaxial film may differ from the substrate in chemical composition. The method for producing such films is called heteroepitaxy, Unlike homoepitaxy described above. Of course, with heteroepitaxy, the film and substrate materials must still have the same crystal lattice. For example, it is possible to grow a silicon film on a sapphire or spinel substrate. (Spinel - general chemical formula - A 2+ B 2 3+ O 4, where A - Fe, Mg, Zn, Mn, etc.;

B - Al, Cr, Ti, etc. Transparent red and blue spinel is a gemstone.)

The boundary between the epitaxial layer and the substrate is not perfectly sharp, since impurities partially diffuse from one layer to another during the epitaxy process. This circumstance makes it difficult to create ultrathin (less than 1 μm) and multilayer epitaxial structures. Single-layer epitaxy currently plays the main role. It has significantly expanded the arsenal of semiconductor technologies: obtaining such thin homogeneous layers (1–10 microns) as epitaxy provides is impossible by other means.

2.3. Doping - introducing impurities into a semiconductor

Selective introduction of impurities into a silicon single crystal makes it possible to establish mass production of electronic circuits with exceptionally small dimensions.

Imagine the lattice of a single crystal of silicon. Ideally, we would like to replace some horizontal and vertical rows of silicon atoms with acceptor or donor impurity atoms R- or n-type. Unfortunately, such a targeted ideal introduction of an impurity is impossible. Two available technologies for introducing impurities - diffusion And ion insertion (ion implantation).

The diffusion method involves bringing the impurity into contact with the surface of a single crystal of silicon. When the single crystal is heated, impurity atoms penetrate into the single crystal and replace the silicon atoms that have been “shaken out” from their places.

– 9 – Lecture 3

The ion doping method consists of directing a beam of high-velocity impurity ions onto the surface of a silicon single crystal. These ions penetrate into the thickness of the crystal and, as a result of interaction with the lattice atoms, slow down and stop.

We will consider the types of diffusion possible in silicon, the laws that govern the diffusion process, the behavior of the diffusion coefficient, as well as issues of the solubility of impurities in silicon. Solutions of diffusion equations (Fick's second law) will be carried out in two cases and some methods for calculating diffusion layers will be discussed. The process of ion insertion will also be described and the relationship between the total concentration

impurities and concentration distributions with the parameters of the implantation system: beam current and accelerating voltage.

Types of diffusion(C. Worth and R. Thomson, Solid State Physics, pp. 75–87)

This material is closely related to the nature of defects discussed in the MET course. Considering temperature as a thermal vibration, we can assume that the lattice is shaking so much that some atoms leave their places and are replaced by others. If an impurity atom with approximately the same size and valency happens to be nearby, it can replace the departed atom at a lattice site. This process is called diffusion of substitution or diffusion across vacancies. Impurity atoms jump around lattice sites stochastically (randomly) in all directions. Let us assume that this random process began in one area with a certain concentration of impurities. Then the number of atoms that can return to the region with a higher impurity concentration will be less than the number of atoms that can return to the region with a lower concentration. Thus, the movement of the impurity will generally be directed towards decreasing concentration. But for an impurity to move in a crystal, it is necessary that vacancies be encountered along the path of the impurity atom. This means that the rate of substitution diffusion depends on the rate of formation of vacancies in the lattice.

However, an impurity atom can find a place in the crystal without replacing the atom original element. These are the so-called internodes. At high temperatures, impurity atoms can jump from one interstitial site to another, thus moving along the crystal lattice. This movement is called diffusion of implementation. Interstitial diffusion has a much higher propagation speed than substitution diffusion, since interstices are, as a rule, free, and for substitution to occur one has to wait for the formation of vacancies.

A combination of these types of diffusion can take place in a crystal lattice. A certain proportion of the impurity performs interstitial diffusion, and the rest - substitution diffusion. To construct a model of the interstitial diffusion process, we will assume that this process can start from a lattice site or interstitial site and also end at both a lattice site and an interstitial site, but the movement of the atom occurs only along interstitial sites.

All donors and acceptors usually used for silicon diffuse through vacancies. Gold, iron, copper and lithium diffuse through the interstices. Iron, copper and lithium are generally undesirable, but gold is intentionally introduced by a diffusion process into some high-speed ICs. Gold atoms diffuse along interstices, but most of them stop at lattice sites, forming a substitutional impurity. Gold atoms do not fit well into the silicon lattice. They form something like heterogeneity in it. Since the recombination of an electron-hole pair occurs near an inhomogeneity, gold atoms cause an increase in the recombination rate and a corresponding decrease in the average lifetime of mobile

– 10 – Lectures 3 – 4

charge carriers. Gold atoms reduce the carrier lifetime and therefore increase the switching speed. True, for many reasons, they are now trying to replace the use of gold to increase switching speed with other methods - modification of the geometry of elements, manufacturing techniques and circuit design.

The rate of diffusion depends on a number of factors related to the structure of the material, in particular, the interatomic distance, the chemical number of the element, the type of bonding between atoms in the lattice, etc. However, since the frequency of jumps of individual atoms depends exponentially on temperature, the diffusion equation as a function of temperature is often written in the form

D = D 0 . e – Q /(kT) , (1)

Where D 0 is called the frequency factor (independent of temperature), and Q- activation energy. D 0 and Q for most practically used materials can be found in the literature (Worth and Thomson, p. 87, E. A. Matson and D. V. Kryzhanovsky [SPKM], etc.)

The concentration of an impurity in a material cannot be arbitrarily large: it is limited by a special parameter - the maximum solubility of the impurity, which itself depends on temperature: at a certain temperature it reaches maximum value, and then decreases again. I will now give you the maximum limiting solubilities along with the corresponding temperatures for some materials (Stepanenko [OM] - p. 162, Till and Lacson [ISMPI] - p. 74–75):


N pr, cm -3				Lecture course

Send your good work in the knowledge base is simple. Use the form below

Students, graduate students, young scientists who use the knowledge base in their studies and work will be very grateful to you.

Posted on http://www.allbest.ru/

Lecture 1 Introduction. Integrated circuits as a new type of device

Microelectronics. Basic concepts.

Microelectronics is a branch of electronics that covers the research and development of a qualitatively new type of electronic devices ( integrated circuits(IS, IMS)) and principles of their application.

Integrated circuits (ICs) are a set of a large number of interconnected components (transistors, diodes, capacitors, resistors, etc.), manufactured in a single technical cycle on the same supporting structure (substrate) and performing a specific function of converting information.

Integrated circuits as high quality new type devices:

b ICs independently perform a complete complex function.

b Increasing the functionality of the IS is not accompanied by a deterioration in the main indicators, but on the contrary contributes to their improvement.

b Preference for active elements over passive ones.

b It is unlikely that there will be a scatter of parameters among adjacent elements, since they are located at a distance of several microns from each other.

Main IP parameters

1) The functional complexity of an IC is characterized by the degree of integration, that is, the number of elements on the chip: k = ln*N, where N is the number of elements.

2) Packing density - the number of elements per unit area of the crystal. At the moment, a density of 100,000 elements per square millimeter has been achieved.

IC classifications

According to the design and technological design:

According to the degree of integration there are:

According to their functional purpose, IS are divided into:

1) Analog - generators, amplifiers, detectors, shapers, frequency filters, signal delay devices, converters, modulators.

2) Digital - triggers, logical ICs, storage devices, devices for processing digital information.

For use in equipment:

1) general use.

2) special application.

By design:

1) Case-based.

2) Unframed.

According to production technology:

1) Silicon technology.

2) Gallium Arsenide technology.

3) Silicon-germanium technology.

System symbols IC.

1) IC housing type:

K - wide application,

P, A - plastic body,

C, I - glass-ceramic body,

M - metal-ceramic or ceramic body,

E - metal-polymer body.

If there is no letter after K, and after one there is B (K_140UD1B-2), then the IC is unframed.

2) Structural and technological design:

1.6 - IC on bipolar transistors,

5, 7 - ICs based on field-effect transistors,

2, 4 - hybrid ICs,

3, 8 - reserve.

3) IC series number.

4) Affiliation of the circuit according to its functional purpose.

5) Serial number developments in this series (1 or 2 digits).

6) Type of IC pins:

1 - flexible leads,

2 - tape outputs,

3 - hard conclusions.

For the manufacture of ICs, a group method is used, i.e. A large number of ICs are manufactured on the same wafer; if the technological process allows, several such wafers are processed simultaneously.

An element is a part of a microcircuit that implements the function of some electrical radio element (transistor, diode, etc.), which cannot be isolated as an independent product.

A component is a part of a microcircuit that performs the function of some kind of electrical radio element, which can be separated as an independent product.

Film IC (FIC) is a microcircuit whose elements are made in the form of various kinds of films deposited on the surface of a dielectric substrate.

A combined IC (SIS) is a microcircuit in which the active elements are made in the near-surface layer of a semiconductor crystal, and the passive elements are deposited in the form of films on a previously insulated surface of the same crystal.

Semiconductor IC (SPIS) is a microcircuit all of whose elements are made in the near-surface layer or in the volume of a semiconductor crystal.

After completing all the main technological operations, the wafer with the IC is sent to the operations of test monitoring of electrical parameters. At this stage, all ICs whose parameters do not correspond to the required values are rejected and marked. Then the wafer is divided into individual chips (ICs) using the scribing method (either using a diamond cutter or a laser), then the wafer is broken into individual chips and suitable chips are mounted in packages.

Scribing is the fastest, high-performance method of dividing a group of animated blanks into individual printed circuit boards, which has a low cost.

Scribing with a diamond cutter.

1 - cutting edge of the cutter; 2 - tracks for scribing in the layer of protective dielectric; 3 - semiconductor microcircuits; 4 - silicon wafer. a) - applying marks; b) - plate with risks; c) - design of a diamond pyramid.

Laser scribing.

a) - Scheme of laser scribing of semiconductor wafers; b) - Breaking semiconductor wafers into crystals with a roller: 1 - roller; 2 - protective film; 3 - crystal. c) - Breaking semiconductor wafers by rolling between rollers: 1 - plate; 2 - elastic roller; 3 - protective film; 4 - steel roller; 5 - carrier film.

Lecture 2 Hybrid integrated circuits

Hybrid IC (HIS) is a microcircuit that is a combination of film passive elements and active components located on a common dielectric substrate (active components are mounted).

The main element of the GIS is the substrate - the structural basis of the GIS. The following requirements apply to substrate materials:

1) High resistivity of elements

2) Mechanical strength

3) Physical and chemical resistance

4) Resistance to high temperatures

5) Good polishability of surfaces

6) Low cost

The substrate material and manufacturing technology must ensure high class surface cleanliness to ensure uniformity and reproducibility of electrical parameters of circuit elements.

Passive elements are resistors, capacitors and inductors. Active elements are called transistors, diodes and other elements of microcircuits.

Passive elements of GIS.

Resistors.

R-resistance; a is the length of the resistor; b-width of the resistor; h is the thickness of the resistor; Kf-shape coefficient; - specific surface resistance, characterizes the resistance of a square surface.

Capacitor.

C is the capacitance of the capacitor; a is the length of the resistor; b-width of the resistor; S - lining area; d is the distance between the capacitor plates.

Inductor.

Methods for forming thin films.

There are 3 main methods for forming thin films:

1. Thermal vacuum spraying;

2. Ion-plasma sputtering:

a) Cathode sputtering;

b) Ion-plasma sputtering;

3. Electrochemical deposition.

1. Thermal vacuum spraying.

2. Base plate.

4. Substrate.

5. Holder.

6. Heater.

7. Evaporator.

8. Rotary valve.

The criterion for the required vacuum is that the mean free path of an atom must be several times greater than the distance between the evaporator and the substrate (10 -11 - 10 -12 mm Hg).

Advantages of the method:

1) spraying speed;

2) simplicity and sophistication of the technology;

3) the possibility of obtaining clean films due to the high degree of vacuum.

Disadvantages of the method:

1) high material consumption;

2) low adhesion (adhesion strength of the film to the substrate or other film);

3) low quality of films due to uneven deposition;

4) the difficulty of reproducing the chemical composition of the evaporated substance;

5) the difficulty of spraying refractory materials.

2. Cathode sputtering.

1. Metal or glass cap.

2. Base plate.

3. Vacuum maintaining gasket.

4. Substrate.

5. Holder.

6. A cathode made of a sprayed substance.

7. Heater.

8. Fitting.

The cathode 6 consists either of the sprayed substance or is in electrical contact with it. The role of the anode is performed by a substrate with a holder.

The subcap space is first pumped out to 10 -5 - 10 -6 mm Hg. Art., and then a certain amount of purified neutral gas (often argon) is introduced, creating a pressure of 10 -1 - 10 -2 mm Hg. Art. When a high voltage (2 - 3 kV) is applied to the cathode, an anomalous glow discharge occurs in the anode-cathode space, accompanied by the formation of a quasi-neutral electron-ion plasma. A strong electric field is formed in the cathode space. Positive gas ions, accelerated by this field, bombard the cathode, knocking out not only electrons but also neutral atoms. The cathode is destroyed, and neutral atoms are deposited on the substrate, forming a film.

Advantages of the method:

1) low temperature;

2) low consumption of the sprayed substance because Spraying occurs only on the substrate.

Disadvantages of the method:

1) the material must have high electrical conductivity (impossibility of sputtering oxides and dielectrics);

2) contamination of films due to low vacuum;

3) low spraying speed.

3. Ion-plasma sputtering.

1. Metal or glass cap.

2. Base plate.

3. Gasket that maintains the vacuum.

4. Substrate.

5. Holder.

6. Heated cathode.

8. Fitting.

9. Target.

A non-self-sustaining arc discharge is ignited between electrodes 6 and 7, characterized by a special source of electrons in the form of an incandescent cathode 6, low operating voltages (tens of volts) and a high density of electron-ion plasma.

The under-cap space is filled with neutral gas with a pressure of 10 -3 - 10 -4 mm Hg. Art.

A negative potential (2-3 kV) is applied to the target, sufficient to cause an anomalous glow discharge and intense bombardment of the target with positive plasma ions. The knocked-out target atoms land on the substrate and are deposited on it.

The use of a mechanical shutter allows for ion cleaning of the target, which improves the quality of the sputtered film.

Compared to cathode sputtering, this method has the following advantages:

1) high speed;

2) greater process flexibility;

3) higher vacuum (correspondingly higher film quality).

4. Electrochemical deposition.

The method is based on the electrolysis of a solution containing ions of the necessary impurities. Material ions in solution have a positive charge, so the substrate is used as a cathode.

Advantages of the method:

1) high spraying speed;

2) the ability to adjust the thickness of the resulting films by changing the current.

Disadvantages of the method: very low quality of the resulting films.

Applying thin continuous films to the surface of the substrate is not the main goal of the technology; the main task is to create the required topology (pattern geometry). To do this, it is necessary to convert the continuous film into the corresponding drawing (image) by removing unnecessary parts.

There are several methods for generating GIS elements:

1. Removable mask method;

2. Contact mask method.

Removable mask method.

The removable mask method is based on the deposition of films through masks. The mask is a thin bimetallic foil with holes (windows).

Disadvantages of the method:

1) in the process, spraying occurs on the stencils themselves and renders them unusable;

2) stencil sagging;

3) metal masks are of little use for cathode and ion-plasma sputtering because they distort the electric field, which affects the sputtering speed.

Contact mask method.

The process of creating or transferring a geometric pattern (topology) onto the surface of a substrate is called lithography.

To form an image on the surface, special materials called resists are used (materials sensitive to active radiation, applied by centrifugation or spraying to the substrate). After treatment with active radiation, the resist film undergoes physical and chemical changes, as a result of which it becomes resistant to aggressive environments. Unnecessary parts of the film are removed with the so-called developer. The remaining drawing is called a mask.

If, after interaction with active radiation, the resist material is destroyed and can then be removed, such a resist is called positive. If it polymerizes and acquires a three-dimensional structure, and the non-irradiated material can be removed, then such a resist is called negative.

Lithography using electromagnetic waves of visible and UV radiation as active radiation is called photolithography.

Photolithography.

The design of the future mask is made in the form of a so-called photomask. A photomask is a thick glass plate on one side of which a thin opaque film with the necessary pattern in the form of transparent holes is applied.

The production of a thin-film resistor by photolithography is shown in the figure on the left, where 1 is a photomask, 2 is a photoresist, 3 is a conductive layer, 4 is a resistive layer.

Photolithography is carried out as follows. Continuous films of the necessary materials are applied to the substrate, for example, to create a resistor, a resistive layer and on top of it a conductive layer (a). Then the surface is coated with a photoresist and, using an appropriate photomask, a pattern of the conductive layer is created in it (b), etching is carried out to remove unnecessary parts of the conductive layer (c), and the remaining photoresist is removed (d). As a result, ready-made contact pads are obtained on the still continuous surface of the resistive layer. Photoresist is applied again and a resistor strip pattern is created using another photomask (e). Then etching is carried out, the photoresist is removed and a finished resistor configuration with contact pad(s) is obtained.

Materials used in the manufacture of GIS.

For resistive films, chromium, nichrome (Ni - 80%, Cr - 20%) and cermet from a mixture of chromium and silicon monoxide (1:1) are most often used.

Aluminum is used for capacitor plates, and before sputtering the bottom plate it is necessary to spray a thin layer of CrTi, since the adhesion of aluminum directly to the substrate is insufficient. For dielectric layers, the most common are silicon monoxide SiO and germanium monoxide GeO. A special place among dielectrics is occupied by the oxides Ta 2 O 5 and Al 2 O 3.

For conductive films and ohmic contacts, as a rule, either gold with a CrTi sublayer or copper with a vanadium sublayer is used. The thickness of conductive films is 0.5 - 1 mm.

Substrates must have the properties of good insulation, low dielectric constant, high thermal conductivity, sufficient mechanical strength, polishability (roughness should not exceed 25 - 50 nm), LTLE must be close to the LTLE of the films used. The most common materials for substrates are glass ceramics and ceramics.

Lecture 3 Physics of semiconductor structures

There are two main mechanisms of charge carrier transport in semiconductors:

1) Diffusion - directed movement of charge carriers in a crystal in the direction of decreasing their concentration.

2) Drift - ordered movement of charge carriers under the influence of an external electric field.

Semiconductor-semiconductor contact

p+-n unbiased junction

Since the concentration of holes in the p + -region is higher than the concentration of electrons in the n-region, a process of diffusion from the p- to n-region (diffusion current) occurs. At the same time, diffusion of electrons from the n- to p-region begins. As a result, a double electric layer of space charge is formed at the boundary of the p- and n-regions. The field of this double layer creates a potential barrier that prevents further diffusion of holes into the n-region and electrons into the p-region. For electrons to move to the p-region and holes to move to the n-region, they need to overcome the potential difference (band gap). Thus, a p-n junction is formed at the interface of semiconductors with different types of conductivity.

If a direct voltage is applied to the p-n junction, an electric field opposite to the field is created in it, the resulting field is weakened, the potential barrier is reduced, additional electrons enter the p-region, and holes enter the n-region, this process is called charge injection. When a reverse voltage is applied to the p-n junction, a field is created that coincides in direction with its own, therefore the resulting field is enhanced, the potential barrier increases, and the movement of charges is minimized.

Metal-semiconductor contact.

Can be ohmic or rectifying.

The work function of electrons from a solid is the energy required to escape from the crystal (μm is the work function of electrons from the metal, μs is the work function of electrons from the semiconductor).

Ohmic contact.

An ohmic contact is a physical contact between a metal and a semiconductor whose electrical resistance is low and does not depend on the direction of the current in a given range of current values.

The presence of an enriched layer in both cases means that the resistance of the system as a whole is determined by the neutral layer of the semiconductor and does not depend on either the magnitude or the polarity of the applied voltage. Such a contact is called ohmic.

Rectifying contact.

In the depleted layer, the concentration of OHC is less than the equilibrium one (far from the contact); therefore, the near-contact layer has an increased resistivity and determines the resistance of the system as a whole. The potential barrier in the contact layer is called the Schottke barrier. Such contacts have rectifying properties and can be the basis for creating dioids. Aluminum is mainly used for metallization of semiconductors in ICs. A Schottke barrier will appear at the interface between aluminum and n-silicon; to prevent this, the contacts are burned in.

Dielectric-semiconductor contact.

Silicon dioxide SiO 2 is mainly used as a dielectric. Silicon dioxide layers always contain donor-type impurities, which upon contact with silicon are concentrated near the boundary. Therefore, silicon dioxide films at the interface with silicon form a thin layer of positively charged donor atoms, and the electrons donated to them pass into the silicon layer.

Bipolar transistors (BT). Principle of operation.

BTs are two back-to-back pn junctions. Interaction is ensured when the connection distance is less than the diffusion length of the NS.

Offset type

Reverse

Types of displacement at transitions for various operating modes of BT:

When the transistor is turned on normally (forward-biased EP, reverse-biased EP), the potential barrier of the EP decreases and the EP increases. Electrons injected from the emitter into the base pass through it almost without recombination and enter the collector, which is under a positive potential, thus the collector collects the NSCs that enter the base.

The resistance of the reverse-biased gearbox is very high - several megohms, so you can turn on very large load resistances without changing the collector current. Consequently, significant power can be released in the load circuit: P = I 2 R.

The resistance of a forward-biased EP, on the contrary, is very small (at I e = 1 mA, R e = 25 Ohm) therefore, at almost the same currents, the power consumed in the emitter circuit is much less than the power released in the load circuit, therefore, the transistor amplifies the power (it is an amplification device).

Connection schemes:

With a common base

Common emitter

With common collector

Lecture 4 Semiconductor integrated circuits. Technological foundations of semiconductor microelectronics

microcircuit dielectric bipolar semiconductor

A semiconductor IC is a microcircuit that is a semiconductor crystal, in the near-surface layer and volume of which areas are formed that are equivalent to the elements of an electrical circuit and the connections between them.

Basic technological operations for the production of PPIS

1. Preparatory.

Preparatory single-crystalline silicon ingots are usually obtained by crystallization from a melt (Czochralski method), in which a rod with a seed in the form of a single crystal of silicon, after contact with the melt, is slowly raised with simultaneous rotation. Following the seed, the growing and solidifying ingot is pulled out. The diameter of the rod is 1.5-3 cm, length - up to 3 meters (usually 70-100 cm). The silicon ingot is then cut into wafers 0.5 to 4 cm thick.

Before starting the main technological operations, the plates are polished for:

1) removal of mechanical defects;

2) ensuring parallelism of planes;

3) ensuring the required thickness of the plates (0.15 - 0.25 mm).

After grinding, the plates are polished to reduce surface roughness to hundredths of a micron.

Etching is carried out.

2. Epitaxy.

Epitaxy is a process of growing single-crystal layers onto a substrate, in which the crystallographic orientation of the grown layer follows the crystallographic orientation of the substrate.

Epitaxy can be:

1) From the gas phase.

A standard chloride process is used. A single crystal of a silicon wafer is placed in a crucible in a quartz tube through which a stream of hydrogen with a small amount of silicon tetrochloride SiCl 4 is passed. At high temperatures, a reaction occurs on the surface of the silicon wafer:

As a result, a layer of pure silicon is created on the surface of the wafer, and hydrochloric acid vapor is removed with a stream of hydrogen. If B 2 H 6 (boron hexohydride) is added to SiCl 4 vapors, then the resulting film will have a hole type of conductivity, but if PH 3 (phosphorus trihydride) it will have an electronic type.

2) From the liquid phase - by the Czochralski method.

To improve process control, electroepitaxy is used - electricity, at a certain polarity, cooling (film growth) or heating of the interface between two phases occurs (dissolution of the upper layer of the substrate in the melt).

3) Molecular beam epitaxy - condensation of molecular beams in ultra-high vacuum (analogue of vacuum deposition).

3. Thermal oxidation.

Thermal oxidation of silicon is one of the most characteristic processes in semiconductor IC manufacturing technology. The dioxide film performs the following functions:

1) protective (passivation, surface insulation);

2) masking;

3) the function of a thin dielectric under the gate of a MOS transistor (MOS).

Silicon oxidation is carried out at temperatures from 1000? up to 1200? C in an atmosphere of oxygen (dry oxidation) or a mixture of oxygen with water vapor (wet).

4. Doping (introduction of impurities).

This is carried out in either single-zone ovens (if a liquid or gaseous diffusion source is used) or in two-zone ovens if a solid diffusion source is used. In this case, in the first zone, the diffusant is mixed with the flow of neutral carrier gas and transferred to the second zone, where diffusion occurs at a higher temperature.

The theoretical foundations of diffusion are based on Fick's laws:

D - diffusion coefficient, N - concentration, I - particle flux density.

Characterizes the rate of accumulation of particles in a crystal or the accumulation of impurity atoms at a certain depth. In order to calculate the distribution of impurities in depth at any time, it is necessary to set boundary conditions.

N(0,t) - impurity concentration on the surface.

N(?,t) is the impurity concentration on the opposite side of the substrate.

L N is the depth to which the impurity must be introduced.

N S - surface concentration.

N 0 - intrinsic concentration.

Ш The case of an unlimited source of diffusant.

where erfc - additional function errors, close to the exponential function e - z.

To create a p-n junction, it is necessary at a depth L N to obtain equality between the concentrations of the introduced impurity and the intrinsic concentration of the crystal:

Ш Case of a limited source of diffusant.

First, a certain number of diffusant atoms are introduced into the near-surface layer of the plate, then the diffusion source is turned off, the temperature is increased, and the impurity atoms are redistributed in depth with their total number unchanged.

where Q is the number of impurity atoms per unit area.

Ш Ion implantation.

This is a method of doping plates by bombarding them with impurity atoms, accelerated to an energy sufficient to embed the atoms deep into the solid. The energy of atoms is about 100-150 KeV. The impurity concentration in the implanted layer depends on the current density in the beam and the process time (it lasts from several minutes to several hours).

Advantages:

1) low process temperature;

2) controllability.

Disadvantages: beam scanning.

5. Etching.

1) etching itself - a chemical reaction of a liquid etchant with a solid to form a soluble compound. Unlike mechanical etching, etching occurs smoothly, one molecular layer after another. By selecting the etchant, its concentration, temperature and process time regulate the thickness of the layer being removed.

2) Electrolytic etching - the process occurs under conditions of current flow through a liquid. Solid body - anode.

3) Ion etching - the plate is placed in a vacuum, a glow discharge is created in it, a negative potential is applied to the silicon and plasma ions bombard the plate (etch). Ion energy is 2-3 keV.

6. Lithography.

Lithography is the process of creating or transferring a geometric pattern (topology) onto the surface of a substrate.

1) Photolithography.

2) Electronic lithography. In this type of lithography, a focused beam of electrons scans the surface of a wafer coated with resist or silicon dioxide. In those places that should be illuminated, the beam current is maximum. In areas exposed to light, silicon dioxide is etched faster.

3) Laser lithography. It is based on the impact of a scanning beam of an excimer laser on a silicon substrate placed in a chlorine environment. As a result of the chemical reaction of chlorine ions with silicon atoms under the thermal influence of a laser beam, a gaseous compound SiCO 4 is formed, which volatilizes, and an etched area is formed on the substrate.

Aluminum is mainly used for metallization of silicon. If a film of aluminum is simply deposited onto the surface of silicon, Schottky barriers are formed. Rectifying junctions appear at the border with n-regions. To avoid this, aluminum is burned into silicon at a temperature of 600? C. In this case, a layer is formed at the aluminum-silicon interface in which almost all aluminum is dissolved. After solidification, a surface layer of silicon doped with aluminum with a concentration of 5 10 13 cm -3 is obtained. Since aluminum is an acceptor with respect to silicon, the formation of p-n junctions in n-layers is possible. To avoid this, the n-region near the contact is heavily doped, creating an n + layer with a concentration of 10 20 cm -3.

Lecture 5 Semiconductor integrated circuits. PPIS on bipolar elements

Manufacturing technology of PPIS on BT.

PPIS elements must be isolated from each other so that the necessary connections are made only by metal wiring. All insulation methods can be divided into 2 main types:

1) insulation by a reverse biased p-n junction;

2) dielectric insulation.

Isolation by reverse biased p-n junction.

Provides for the implementation of 2 back-to-back diodes between the insulated elements.

1. Triple diffusion method.

Local diffusion is carried out through windows in a silica mask. A donor impurity is introduced into the original p-silicon wafer and an n-layer (collector region) is obtained. During the second diffusion, an acceptor impurity with a higher concentration is introduced into the resulting n-layer to a shallower depth. A third diffusion is then carried out to obtain the emitter layer. Each diffusion includes the following processes: wafer oxidation, photomask application, exposure, etching and diffusion.

Flaws:

2) The difficulty of selecting diffusants, since the concentration of the impurity is limited by the maximum solubility.

3) The collector layer is heterogeneous, the impurity concentration increases from the bottom to the surface, as a result of which the breakdown voltage is low.

2. Planar epitaxial technology.

There is an original p-silicon wafer with a polished surface.

1) General oxidation of the plate is carried out.

2) Photolithography: formation of windows in the oxide under the n + layer.

3) Diffusion: formation of n+ layer.

4) Bleeding of oxide.

5) Growth of the epitaxial n - layer, diffusion from n + to n layer (under the influence of temperature).

6) General oxidation.

7) Second photolithography, formation of windows in the oxide for separation diffusion.

8) Second diffusion, formation of p - separation layers (n - pockets), boron diffusant.

9) Third photolithography. Formation of windows for the base area.

10) Third diffusion, formation of base areas. Boron diffuser. Diffusion is carried out in two stages: driving and acceleration.

11) The fourth photolithography, the formation of windows in the oxide under the emitter and ohmic contacts of the collector.

12) Fourth diffusion, creation of n + layers. Phosphorus diffuser.

13) Fifth photolithography, formation of windows in the oxide for ohmic contacts.

14) Aluminum spraying.

15) The sixth photolithography, the formation of windows in the oxide for metal wiring.

16) Aluminum etching.

17) Heat treatment for burning aluminum.

The hidden n+ layer is necessary to reduce the horizontal resistance of the collector.

Profile of impurity distribution along the depth of an epiplanar bipolar integrated transistor.

Effective concentration distribution profile.

3. Isolation by collector diffusion (ICD technology).

In a p-silicon semiconductor, hidden n+ layers are created, then a p-type epitaxial layer is grown. Next, separation diffusion is carried out using a donor impurity, forming n + separation layers. The result is p-type pockets (base layers) and n+ separation layers along with hidden layers(collectors). Diffusion is then carried out to form emitter regions.

Advantages:

1) In KID technology, the number of photolithographs is reduced compared to the previous process.

2) The collector area is heavily doped, so there is no need to carry out additional diffusion of gold or other impurities that reduce the lifetime of minority current carriers to increase the performance of the IC.

3) A thin epitaxial layer limits the collector-base breakdown voltage due to the spread of space charge into the base region.

Dielectric insulation.

1. EPIC technology.

The original n-silicon wafer is covered with an epitaxial n + -layer, 2-3 microns thick. Then grooves 10-15 µm deep are etched into the plate. The entire surface is oxidized and a layer of polycrystalline silicon (200-300 microns) is sprayed. The structure is turned over and ground down to the grooves. The result is n-pockets with a buried n + -layer, in which elements of semiconductor ICs are formed.

2. SOS technology (silicon on sapphire).

Manufacturing large-scale integrated circuits (LSIs) based on silicon-on-sapphire (SOS) structures has a number of advantages compared to manufacturing them on a silicon substrate, such as speed, reduced power consumption, and high packing density of elements. Before growing the silicon layer in the growth chamber, the substrate was annealed at a temperature of T = 1400°C for 30 minutes.

After reducing the substrate temperature to a given one (500-850°C), a silicon layer is deposited at a rate of ~ 1 μm/hour. This layer is then etched to the substrate and pocket islands are formed in which the elements are formed.

Combined isolation methods.

1. Isoplanar technology.

The method is based on local oxidation of the n-Si epitaxial layer through a mask of silicon nitride SiN 4 . As a result, the epitaxial layer is divided into separate pockets. The side insulating layers are dielectric, and the bottom parts of the pockets are insulated by back-to-back p-n junctions.

Notes:

The side insulated layers are not semiconducting, but insulating because oxidation occurs throughout the entire depth of the epitaxial layer (no more than 1 μm).

The bottom parts of the pockets are separated by p-n- junctions, so the method is a combined one.

Each pocket is in turn divided by oxide into two parts. In the main part there is a base and an emitter in the second - a collector. Both parts are connected through a hidden n+ layer, which reduces the collector capacitance.

Local oxidation of the epitaxial layer is created throughout its entire thickness, while the thickness of the epitaxial layer does not exceed 1 micron. Obtaining a SiO a film with a thickness of more than 1 micron through thermal oxidation is a very long process.

This method provides a high density of arrangement of elements and is used for the manufacture of LSI insulation.

2. V-groove insulation method.

The method is based on through etching of the epitaxial n-layer using the anisotropic etching method, in which the surface must have an orientation of 1.0.0, etching occurs along the 1.1.1 planes. The edges of these planes converge just below the boundary of the epitaxial layer. The relief is oxidized and then polycrystalline silicon is sprayed onto the entire surface for leveling.

Elements of PPIS.

1. A type of bipolar integrated transistors.

1) Multi-emitter transistors.

MET - form the basis of the class digital circuits TTM.

Flaws:

Ш It is necessary to reduce the inverse transfer coefficient of the emitter current; for this, the base resistance is artificially increased, removing the ohmic contact from the active region of the transistor.

2) Multicollector transistors.

This is a multi-emitter transistor operating in inverse mode. They form the basis of the I 2 L class of digital circuits. The main problem is an increase in the current transfer coefficient, which means that the collectors are located as close to each other as possible.

3) Schottky transistors

They form the basis of the TTLSh class of digital circuits.

4) Super V-transistors (transistors with an ultra-thin base).

These are transistors with a gain of 3000 to 5000, but the breakdown voltage of such a transistor is very low, so these transistors are placed in the input stages of operational amplifiers.

Problems:

Ш Technical problems, two diffusions are carried out through one window.

Ш Pushing back the collector junction.

Ш Fundamental physical problems.

2. Integral diodes.

These are diode connections of integrated transistors (either not all contacts are used, or some are short-circuited):

Anode - cathode (A - K)

3. Integral resistors.

Diffusion up to 20 kOhm

Pinch resistors up to 50-60 kOhm

Ion-doped resistors up to 200-300 kOhm

4. Integral capacitors.

Integral diffusion capacitor design:

1 - aluminum lead from the top plate of the capacitor; 2 - aluminum lead from the bottom plate of the capacitor; 3 - contact to the substrate; 4 - p-type substrate; 5 - collector n-region (lower plate of the capacitor); 6 - base p-region (upper plate of the capacitor); 7 - silicon oxide film.

Posted on Allbest.ru

Fundamentals of Microelectronics

The first most detailed level is

electrical diagram . She defines electrical connections elements (transistors, diodes, resistors, etc.).

At this level, a connection is established between electrical parameters circuit and the parameters of its elements.

Fundamentals of Microelectronics

Electrical diagram – conventional graphic designation of an electrical circuit. The electrical diagram depicts its elements - idealized models of real-life electrical devices (transistors, diodes, resistors, etc.).

An electrical circuit is a collection of interconnected electrical devices and elements through which electric current can flow.

The second level is a block diagram.

It defines the functional connection of individual stages described by electrical circuits.

Source	Amplifier
	Amplifier	amplifier
		amplifier

Source

Fundamentals of Microelectronics

According to their functional purpose, microcircuits are divided into analog and digital.

In analog microcircuits, signals vary according to the law of a continuous function. A typical example of an analog chip is an operational amplifier.

A digital microcircuit is designed to convert and process signals that vary according to the law of a discrete function.

Fundamentals of Microelectronics

1.2 Design and technological types of ICs

The design and technological classification of microcircuits takes into account the manufacturing methods and the resulting structure.

Based on design and technological characteristics, they are distinguished semiconductor and hybrid

microcircuits.

Fundamentals of Microelectronics

All elements in a semiconductor chip

and inter-element connections are made in

volume and on the surface of the semiconductor . The structure containing elements, interconnections and pads is called an integrated circuit chip.

Publisher: Publishing center "Academy"

The year of publishing: 2006

Language: Russian

Pages: 240

Preface

Introduction

Chapter 1. Basic provisions and directions of development of microelectronics

Depending on the type of signals being processed, all integrated circuits are divided into analog and digital. Analog integrated circuits are designed to convert and process signals that vary according to the law of a continuous function. Their areas of application are primarily television and communications equipment, as well as measuring instruments and control systems. Digital integrated circuits are designed to process signals that vary according to the law of a discrete, usually binary, function. They are used to build digital computers, digital units of measuring instruments, automatic control systems, etc. Currently, there is a tendency for increasingly widespread and successful penetration of digital methods (and therefore microcircuits) into traditionally analogue areas. An example is digital methods of processing and recording sound, which have made it possible to obtain previously unattainable quality.

According to the structure and basic manufacturing technology, microcircuits are divided into two fundamentally different types: semiconductor and film. A peculiar mixture of these two technologies makes it possible to produce hybrid as well as combined integrated circuits.

The basis of modern microelectronics is made up of semiconductor ICs, the elements of which are made in a thin (1... 10 µm) near-surface layer of a semiconductor substrate, the role of which is played by a silicon single crystal 200... 300 µm thick. Depending on the degree of integration, the area of the substrate can vary within very wide limits: from several units to 600...700 mm 2.

The elements of a film microcircuit are made in the form of various types of conductive and non-conducting films deposited on a dielectric (usually glass or ceramic) substrate. Pure film ICs contain only passive elements (resistors, capacitors, sometimes inductive elements), since film technology does not allow the production of active elements (transistors) on the substrate, so the use of film ICs is limited.

A hybrid IC is a film microcircuit on which, after its manufacture, specially manufactured unpackaged diodes and transistors are placed in the form of hanging elements.

However, an increase in the degree of integration sharply narrows the scope of VLSI circuits, since they become too specialized and are therefore manufactured in limited quantities, which is not economically viable. The way out is the development and production of basic matrix crystals. Such a crystal contains a large number of identical topological cells that form a matrix. Each cell contains a certain number of unconnected elements, selected in such a way that several functional elements can be formed from them (flip-flop, group of logic gates, etc.). By performing metal wiring inside topological cells and connecting them to each other, it is possible to obtain electronic units that are very complex in design and differ in functionality. Based on one basic matrix crystal, a large number of LSI modifications can be realized by simply replacing metallization photomasks.

The possibilities of microelectronics are far from being exhausted, and the predicted limit of its development as a scientific and technological discipline is constantly being pushed back in time. However, long-term forecasts in such a dynamically developing field as microelectronics are a thankless task. And even if such a limit is reached, this does not mean that progress in the field of electronics will stop. Semiconductor technology will be replaced by new technologies based on different physical principles. Perhaps it will be functional electronics, optical, quantum or, finally, bioelectronics.

Chapter 2. Physical foundations of semiconductor electronics

The first group includes conductors for which even a weak external energy influence easily transfers electrons to higher (and closely spaced) energy levels, which explains the high electrical conductivity of conductors.

When an external electric field is applied, the electrons (charge carriers) in the conduction band drift in the direction opposite to the direction of the field, thereby creating an electric current.

In the second group, the transfer of electrons to higher allowed energy levels is characterized by the existence of a certain threshold of external energy influence exceeding the band gap. Materials with a wide bandgap (more than 3 eV) are classified as dielectrics. Materials with a relatively narrow band gap form a class of semiconductors. The difference between semiconductors and dielectrics from the point of view of band theory is purely quantitative, consisting in different band gaps.

The electrical conductivity of all crystalline substances (and especially semiconductors) significantly depends on the presence of impurities in them. A pure semiconductor is called intrinsic. At absolute zero there is no free media charge and it does not conduct current at all. However, with increasing temperature, some of the covalent bonds between the lattice atoms are destroyed. In this case, a free electron simultaneously appears in the conduction band and a hole in the valence band, which under the influence of an external electric field can move, creating a current. The process of formation of electron-hole pairs under the influence of heat is called thermal generation. The concentrations of free electrons and holes in an intrinsic semiconductor are strictly the same. The conductivity of an intrinsic semiconductor, caused by paired carriers of thermal origin, is also called intrinsic. An index is used to designate quantities characterizing an intrinsic semiconductor. In an intrinsic semiconductor, the Fermi level is located strictly in the middle of the band gap.

Conductivity due to the presence of impurity atoms is called impurity. Almost all impurities used in electronic technology are substitutional impurities. In this case, a “foreign” atom embedded in a semiconductor single crystal takes the place of a “native” one in a node of the crystal lattice.

If the valence of the impurity atoms exceeds the valence of the atoms of the starting material, then the excess unpaired electron as a result of thermal action is easily torn off and becomes free, adding to its own conduction electrons. When such an impurity is introduced near the “bottom” of the conduction band, an additional (impurity) band appears, electrons from which, as a result of thermal (or other) effects of even low intensity, can easily be “thrown” into the conduction band. As a result, the conductivity of the material becomes mainly electronic in nature. Such semiconductors are called l-type semiconductors, and impurities that provide a predominantly electronic nature of electrical conductivity are called donors. It is obvious that the Fermi level in l-type semiconductors is located slightly below the impurity band.

If the valence of the impurity atoms is less than the valence of the atoms of the starting material, then the electron missing to complete the energetically stable shell is “borrowed” from one of the nearest neighbors by breaking its valence bond. As a result, the impurity atom turns into a stationary ion with an uncompensated negative charge, and a hole appears in place of the unfilled bond, which, when an electric field is applied, can drift across the crystal, creating a current.

Chapter 3. Elements of semiconductor digital electronics

Information is understood as a set of information about events, processes and facts taking place in the surrounding world. From a mathematical point of view, the very concept of information and the method of measuring its quantity are usually given on the basis of a probabilistic approach to assessing the state of a system. Let, for example, there be a certain system about which it is known a priori that it can be in one of N possible states. To find out which of these N states the system is currently in, we need information - a set of information that reduces the uncertainty that existed before receiving this information. Information may or may not be complete. Having incomplete information, we, without completely solving the problem, can still exclude some of the impossible states and thereby reduce the previously existing uncertainty. A piece of new information will help further narrow the range of possible conditions. This is exactly how (in an extremely simplified form) the process of cognition develops.

Obviously, a medium is required to store and transmit information. The information carrier can be paper (book), magnetic tape, electric current, sound waves(conversation), etc. However, information cannot be transmitted directly; it must be encoded. Encoding information is a prerequisite for the possibility of its storage and distribution. The purpose of encoding information is served by language - a certain set of signs and rules for their use. A sign is a set of characteristics that allow one to identify an object.

The transfer of information from the sender to the recipient is called communication. A certain piece of information (a set of characters) is called a message.

After receiving a message, the question inevitably arises about the degree of reliability of the information received.

In digital devices, electrical signals take on a number of discrete values, which are represented by binary code.

Functional dependencies in the system are realized by directly solving the equations of the system using one or another numerical methods according to a preset program. The device that implements this solution is called a processor.

The main advantages of digital technology (compared to analogue) are as follows:

Better noise immunity;

Increased accuracy;

Versatility;

Possibility of creating more complex devices.

Let us examine, for example, the issue of noise immunity. basis analog circuits are the simplest cells in the form of amplification stages. In the vast majority of cases, analog devices operate in linear mode, the main feature of which is that the response (output signal) is proportional to the effect (input signal).

Digital circuits are based on their simplest cells - electronic keys - analogues of contact pairs, which are characterized by two stable states: closed and open. However, in circuit terms, both cells are very similar, for example, a transistor connected according to a circuit with an OE. The only difference is in the choice of its operating mode.

Chapter 4. Basic components and blocks of digital technology

In any sufficiently complex objects, which include digital devices, there is a certain hierarchy. Based on the logical elements, the design of which was discussed in the previous chapter, individual nodes are built, from which, in turn, blocks are assembled.

Flip-flops are the simplest sequential circuits (circuits with memory) that can store 1 bit of information. Unlike combinational devices in these circuits, the output state depends not only on the input signals at the moment, but also on the previous history, i.e. from the state of the trigger in the previous cycle. Thus, a flip-flop is a bistable cell with two stable states that can persist indefinitely (as long as there is power). Triggers form the basis for the construction of most functional devices of the sequential type: registers, counters, frequency dividers and other units, so it makes sense to dwell in more detail on their structure and operating principles. Despite the extreme variety of types and, even more so, specific trigger schemes, their construction is based on the basic structure of the trigger, given along with its conventional graphic designation.

If you need to establish the state of a trigger based on a given combination of input signals, then during analysis it is reasonable to adhere to a certain sequence. They start by considering the signals at the preparatory (usually asynchronous and priority) inputs, since it is possible that they determine the state of the trigger. Secondly, the signal at the synchronization input is analyzed and, lastly, the state of the information inputs.

A digital counter is a device that counts the pulses received at the input, generates the counting result in a given code and, if necessary, stores this result. As input signals arrive, the counter sequentially cycles through its states in an order specific to a given circuit. The basis of any counter, as a rule, is a line of triggers of a two-stage structure or triggers of a dynamic type. Single-stage static triggers are not suitable for constructing counters. In this case, not only the G-trigger can be used as a base one, but also the B- or F-trigger in the appropriate switching mode.

Currently, computers use several levels of cache memory. The first, and often the second, caching levels are geographically located in the microprocessor, which allows exchanges to be performed at the highest possible speed (up to the core frequency). The second and third levels of cache memory are located on motherboard in close proximity to the central processor. Multi-level caching means that a lower-level cache caches a higher-level cache.

The principle of hierarchical organization (caching) is very widely used in a variety of memory subsystems. Virtual memory is the caching of disk memory using RAM. Obviously, the efficiency of the caching algorithm determines the probability of finding the requested data in the cache memory and, consequently, the gain in memory and computer performance as a whole. The main problem that needs to be solved when using the caching principle is the problem of coherence, i.e. correspondence of the contents of the cache to the contents of the area of which it should be a copy.

RAM and SRAM are volatile devices because the information stored in them is not retained when the power is turned off.

Together, RAM and ROM form the main memory of the machine. A kind of hybrid of RAM and ROM is semi-permanent memory, which in some cases replaces ROM.

Chapter 5. Microprocessors

A microprocessor is a software-controlled device designed to process digital information and control the process of this processing, made in the form of one or more integrated circuits with a high degree of integration. Depending on their functional features, processors are divided into universal and specialized. Universal processors are used to solve a wide range of problems in measuring and control systems and are the central link of any microcomputer. Specialized processors are focused on solving one specific task, but they perform this task many times faster than general-purpose processors. Specialized microprocessors and microprocessor systems based on them are usually called microcontrollers.

A processor is a device that processes digital information in accordance with a given program. By changing the program, the possibility of solving a variety of specific problems is realized. It is important to note that the processor is designed to work together with storage devices and input/output devices (I/O). The memory device stores commands of the processor operating program, as well as the values of constants and variables involved in calculations. The I/O block performs the interface function microprocessor system with the control object. In single-chip computers (microcontrollers), the processor itself, the memory and the computer are located inside a common housing on one chip.

The elementary cycle of operation of a microprocessor system usually consists of three stages.

1. Retrieving code from memory, which is treated by the processor as a command.

2. Execute the command.

3. Preparing to select the next team.

Such cycles are executed by a microprocessor system regardless of the content of the problem being solved.

The normal functioning of such a complex system as a processor is impossible without clear coordination, i.e. synchronization of the work of all its parts. The microprocessor system operates synchronously with the appearance of special clock pulses coming from the clock generator. An elementary action that a system can perform takes one period (cycle) of the clock generator and is called a microoperation.

Several microoperations can be performed in parallel by different elements of the processor at the same clock cycle. A set of simultaneously executed micro-operations is called a micro-instruction. The time interval during which a micro-instruction is executed is called a machine clock. A certain number of clock intervals form a machine cycle. The specific number of clock cycles in a machine cycle is determined by the code of the instruction being executed. For one processor access to memory or airwaves, one machine cycle is required. At the beginning of each machine cycle, the processor generates a special synchronization signal, thereby achieving coordinated operation of the processor with the memory and airborne devices.

The characteristics of the processor and the algorithm for its operation largely determine the capabilities for hardware and software implementation of the functions necessary to build a microprocessor system. Such characteristics are defined by the concept of microprocessor architecture, which reflects the following basic elements:

Structure, i.e. set of components ( components) microprocessor and connections between them;

Methods of presentation and formats of data;

A set of operations performed by a microprocessor;

Methods of addressing internal and external devices;

Characteristics of control words and signals generated by the microprocessor and supplied to it from the outside;

Reaction to external signals (interrupt processing system, etc.).

Based on the way the memory space of a microprocessor system is organized, two main types of architectures are distinguished.

An organization that uses shared memory space to store programs and data is called von Neumann architecture, named after the famous mathematician who proposed it. In this case, programs and data are stored in a single space and there are no signs indicating the type of information in the memory cell. The advantages of this architecture are a simpler internal structure of the microprocessor and a smaller number of control signals.

An organization in which program memory and data memory are separated and have their own address spaces and methods of accessing them is called Harvard architecture, after the name of the Harvard University laboratory that proposed it. This architecture is more complex and requires additional control signals. However, it allows for more flexible information processing algorithms, which in some cases can speed up the operation of the microprocessor.

Chapter 6. Integrated circuits. Construction of microelectronic devices, devices and systems

Bibliography

465 views

Fundamentals of microelectronics (course of lectures). Technological foundations of microelectronics

Lecture 1 (bring literature and exhibits to the lecture)

Section 1. BASIC PROVISIONS

Lecture 2

Section 2. BASICS OF IC TECHNOLOGY

Lecture 3

conclusions

Send your good work in the knowledge base is simple. Use the form below

Similar documents

Fundamentals of Microelectronics

Fundamentals of Microelectronics

The second level is a block diagram.

Fundamentals of Microelectronics

Fundamentals of Microelectronics

Fundamentals of Microelectronics

Popular articles

Latest articles

Sections

Pages

Special projects

Contacts