**EE692** Advanced Semiconductor Devices Gong Gu

**Why Semiconductors?** Information processing (Amps, A/D, processors, tranceivers…) Information acquisition (sensors) Image, sound, temperature, pressure, … Image, sound, temperature, pressure, … Information transmission (wires, busses, cables, optical fibers, or just air!) • Brains and muscles of the system are made of semiconductors • Metals & dielectrics are used as transmission media • Why? Information processing (tranceivers, processors, …) Displays

**What’s common for all the core components?** Light, sound, temperature, pressure, … Voltage, current Vout Vin sensor A Vout Vin Vin Vout output input Vin Modulation of some physical quantity (output) by some others Some kind of gain, conversion ratio, sensitivity, etc

**Example: Field-Effect Transistors (FETs)** Semiconductor vs Metal For SiO2 dielectric, breakdown field Eb~ 107 V/cm. No matter how thick it is, the maximum induced carrier area density is r0Eb/q = 2 × 1013 /cm2. Vout Vin Vout Vin For a 1 m thick Si channel, ni = 1.45 × 1010 /cm3, the background carrier area density is ni × 104 cm = 1.45 × 106 /cm2. In principle, the area carrier density, and therefore the channel conductance, can be modulated by 7 orders of mag!!! FET’s are building blocks. G D S Schematic illustration of a FET For Al, n = 1.8 × 1023 /cm3. Even for 1 nm thin (monolayers!) Al, the background carrier area density is 1.8 × 1016 /cm2. The conductance can only be modulated by 0.1%!!! What are semiconductors, anyway???

**chemistry** Economics Semiconductor processing Materials science Information theory Solid-state physics Semiconductor physics Device physics circuits Transistor level Higher level Control theory Core knowledge body of the device engineer A Digression: The Vast Field of Electrical Engineering • Different disciplines are different levels of extraction • Device engineers are at the junction of many disciplines • Follow your passion

**chemistry** Economics Semiconductor processing Materials science Information theory Solid-state physics Semiconductor physics Device physics circuits Transistor level Higher level Control theory Core knowledge body of the device engineer A Digression: The Vast Field of Electrical Engineering • But, each small field can consume one’s entire life • So, how can one be a good device engineer??? Chuang Tzu: My life is limited while knowledge is unlimited. Pursuing the unlimited with the limited, it is just hopeless! 莊子： 吾生也有涯 而知也無涯 以有涯逐無涯 殆矣

**chemistry** Economics Semiconductor processing Materials science Information theory Solid-state physics Semiconductor physics Device physics circuits Transistor level Higher level Control theory Core knowledge body of the device engineer A Digression: The Vast Field of Electrical Engineering How can one be a good device engineer??? The big picture! This course is about the big picture. It willed be tailored to suit your research interest; we have a small class after all.

**Let’s get to know each other!** • Name, year • Previous exposure to quantum mechanics, solid-state physics, device physics, processing, ckt design (courses + hands-on) • Advisor • Research field, particular topic • Like it?

**Syllabus** Course Objective: To provide students with an understanding of device physics and advanced semiconductor device concepts. • Topics • Review of Semiconductor physics • Crystal structure, band structures, band structure modification by alloys, heterostructurs, and strain • Carrier statistics • Scattering, defects, phonons, mobility, transport in heterostructures • Device concepts • MOSFETs, MESFETs, MODFETs, TFTs • Heterojunction bipolar transistors (HBT) • Semiconductor processing • Photodiodes, LEDs, semiconductor lasers • (optional) resonant tunneling devices, quantum interference devices, single electron transistors, quantum dot computing, ... • Introduction to nanoelectronics

**Syllabus (Cont’d)** • Reference books • Jasprit Singh, Physics of Semiconductors and Their Heterostructurs • Reads like somebody’s notes. May not be the most elegant or strict from a physics point of view, but definitely serves semiconductor folks well. Intriguing and stimulating. • Jasprit Singh, Semiconductor Devices:Basic Principles • Book by the same author on Devices but including semiconductor physics & processing. • U. K. Mishra & J. Singh, Semiconductor Device Physics and Design • E-book available on line thru UT Lib. • Karl Hess, Advanced Theory of Semiconductor Devices • Thin, but covers lots of stuff at advanced levels • Ben Streetman, Solid State Electronic Devices • From basic physics to device concepts. Oldie goodie. • S. M. Sze, Physics of Semiconductor Devices • The “Bible” of device enginees. Not for beginners. Keep it in mind or on your shelf; an excellent reference book for your future career. • R. S. Muller & T. I. Kamins, Device Electronics for Integrated Circuits • An undergrad textbook on Si microelectronics, but good to have. I go back to it quite often. • J. D. Plummer, M. D. Deal, P. B. Griffin, Silicon VLSI technology: fundamentals, practice and modeling • Best textbook on processing, by the people who developed many of the models.

**Syllabus (Cont’d)** • Journals • IEEE Electron Device Letters • IEEE Transactions on Electron Devices • Applied Physics Letters • Journal of Applied Physics • Websites • Wikipedia (Are you kidding? No!) • Ioffe Physico-Technical Institute • http://www.ioffe.ru/SVA/NSM/ • http://www.ioffe.ru/SVA/NSM/Semicond/index.html • Physical properties of many semiconductors.

**Syllabus (Cont’d: The Tough Part)** • Evaluation • Classroom participation, performance (15%) • Homework / Mini projects – simple (20%) • Term project: critical review of a selected paper, oral presentation on the topic of the paper, oral exam (65%) • The topic may or may not be closely related to your research, but cannot be your research topic per se. Need my okay on the topic before it’s too late. • The good news: It’s not that tough • … • The population is too small. Any distribution does not have any statistical meaning. Which means, you could all get A’s. On the other hand, you could …

**Back to Business** What are semiconductors, anyway??? Long way to go to answer this question.

**Review of Semiconductor Physics** Quantum mechanics • Shrödinger equation The equation that scared Einstein • Stationary states • Special case: free space • E-k dispersion: light wave vs de Broglie wave • The concept of eigenstates • Wave packets • The uncertainty principle

**Review of Semiconductor Physics** Quantum mechanics A few things that we should cover but haven’t: • Bound states • Atoms; Coulomb potential • Normalization of bound state wavefunctions • Unbound states • Difficulty of normalization of unbound states and the way around it • i vs. j; physics vs. EE • More quantum mechanics jargons you need to know • Eigenstates • Operators, eigenvalues Solving Shrödinger Eq is to find the eigenvalues of the Hamiltonian operator.

**Spin** The electron has an intrinsic angular momentum, with a value ħ/2. Along any direction, spin has two eigenvalues, ± ħ/2.

**Review of Semiconductor Physics** Quantum mechanics • Homework • Solve the Shrödinger eq for the following special cases: • 1D Infinitely deep well • 1D Finite well • 1D Harmonic oscillator • 3D general well • Barrier tunneling For 1., complete the math, visualize results, draw analogy with the electromagnetic resonant cavity. For 2., find and read thru the math in a book, visualize results, compare to dielectric cavity or waveguide For 3., find and read thru the math in a book, visualize results; compare to the above and get some sense of how energy level spacing is related to the shape of the potential. For 4., generalize and visualize as much as you can. For 5., visualize results and discuss physical meanings.

**Review of Semiconductor Physics** Solid-state physics The daunting task of solid state physics • Quantum mechanics gives us the fundamental equation • The equations are only analytically solvable for a handful of special cases • One cannot solve the equations for more than two bodies! • Solid-state physics is about many-body problems • There are 5 × 1022 atoms/cm3 in Si Si atom: 1s22s22p63s23p2 Core: Nueclear + 1s22s22p6, Valence electrons: 3s23p2 We’ll come back to this later Each particle is in the potential of all the other particles, which depends on their positions, which must solved from the equation… You have an equation with ~1023 unknows to solve. Mission impossible! • Solid state physic is all about approximations.

**Review of Semiconductor Physics** Crystal structures If we assume the atomic cores have known and fixed positions, we only need to solve the equations for the valence electrons. Life much easier! Static lattice approximation • Justification • Related/similar approximation: Born-Oppenheimer Crystal structures If you shine X-ray on a piece of solid, very likely you’ll have a diffraction pattern. Remember Bragg? That means periodicity in the structure.

**Review of Semiconductor Physics** Crystal structures Bravais Lattices • A mathematical concept: • No boundary or surface • No real (physical) thing – just points, hence no defects • No motion Unit cells (or primitive unit cells) The smallest unit that repeats itself. Fig. 4.10

**Fig. 4.2** Honeycomb