Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
80 views153 pages

P2 U2and3

sjsykdtygjtyj

Uploaded by

10 Joshua Master
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
80 views153 pages

P2 U2and3

sjsykdtygjtyj

Uploaded by

10 Joshua Master
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 153
1, INTRODUCTION lotion of element icles in solids is responsible for the manifesta ‘in much the same way as THeAotion of ncharged paricles in fluids is responsible of viscosity. Here, as in viscosity, the statistical nature of the motion calls fora study should utilise methods of thermodynamics and statistical mechanics. ; Itis known from elementary physics that a metallic filament connected to the | or any other source of electric current becomes hot and incandescent and up current is high enough. This is one of the simplest and the more obvious manife: materials offer to the passage of an electric current. It is being exploited for purposes in a variety of domestic and industrial applications. In a broad sense, current appears whenever the motion of a charged particle in an electric field is motions or collisions with other charged or neutral particles. Its manifestations th glow of neon lamps to the controlling action of photo cells, vacuum tubes a recent manifestations have to do with problems such as direct energy conyt thermonuclear reactions. In solids the carriers of electric current are electrons alone, and. stable rigid ionic lattice can be approximated to some extent by vi between small electrons and individual large ions. We begin by conductors and then extend it to semiconductors. Following are s properties of metals: to the electric-field strength. —2 Metals have high electric and thermal conductivities, 3. At low temperatures, the resistivity is proportional to the fi pe Ts ms 4. The resistivity of metals at room temperature is of the Debye’s temperature varies linearly with temp OxT 5. For most metals, Tesistivity is inversely proportional 180 2 Etectnicat Propeenies of Merats 184 6, The resistivity of an impur specimen is iven by Mathiessen's rule P= Py + p(7) where Pp is a constant for the impure specimen and p(7) is the tempera wre nuisacomat iemperature dependent resistivity f_Near absolute zero, the resistivity of cert ‘phenomena of superconductivity : ‘ain metals tends towards zero; i.e,, exhibit the 8, The conductivity varies in the presence of may resistance 3 1.38% 107 293]! 911x107 154 x 10° m/s Equation (6.2) indicates that the root mean square velocity of the electron is to the square root of the absolute temperature of the metal. At room temperature, imparted to the electrons by an applied electric field is very much smaller than the av velocity. The time t taken by the electrons in traversing the distance A will thus be decie drift velocity due to the field but by much greater velocity ¢ due to the kinetic therm ce A vital test of the validity of any theory of electrical conductivity is whether or n for Ohm’s law. This law is usually expressed as V = RI, where / is the current terms of the current density, Thus, we have ke R pI ARTA LD tie i” NL ie Je 4 co Pp Where, the resistivity p and conductivity 6 of a conductor of length J and area of related by Let us now assume that there are » Neer sed fed, these are darting about inal DOr Cube metre, wt tha inthe os ius ; ficld , is appl With mo net velocity caries run of eth oaeen eso rand bapa seth thee acer TOM, di is piven by pal here, isthe x-component ofthe Velocity ofthe ith arising from the applied field. Since the right-han fay write equation (6:3) in another form: is] [8] electron, and the suf d side of this eq We have used to denote the average value of v, through unit area in unit time, so that s-component, J, must be the m ‘oneach in a volume of unit area and length v,; that i 1 Jean(- The minus sign, which is coupled to the charge on at conventional direction of current (ie., of J,) is in the opp On differentiating equation (6.5), we have Comparing this with equation (6.4), we see that (aCe ®° thal fora constant, this free electron modell Nothing could be farther from Ohm’ law! Us ee Fig, 641 Calculation of current density Itis clear from equation (6.6) that the law can only be satisfied (ia is, Since nothing that we have just written is actually wrong, the difficulty n - have omitted to include a decelerating term which, when added to [2 | es a et n> 0 a <'>| gives amet — = ™~ Pies ae a Once again we conclude that the electrons cannot be completely f 4 conductor always warms up when a current is flowing, we deduce tha collide and exchange energy with ions—a process normally referred { The equation governing the sum of the accelerations should, therefore, d 4 £ ]o.e| 4 epas gil atl E lace ae x ai We can, in principle, feed details of applied electric and magnetic fi gradients into this type of equation, and solve it in order to obtain the resultant Currents. Unfortunately, in practice, this is an exceedingly difficult operation. The m P obtaining the scattering term, which can only be obtained by quantum mechanical technique. this method normally ends up with an integro-differential equation which is extremely ‘We shall therefore, obtain an expression for [Z< >| bya a v jetta a difficulties. - Suppose at an instant, ¢ = 0, the average velocity of the electron is < v the field off, so that < v,> subsequently tends to zero. In other words, the velocity will be randomised by electron-latice collisions. Ne Tetus that this process follows the simplest law of decay: ye switch SV = ‘ation (6.1)]. Thus one kmol of a metal which has N, atoms will therefore pave N, free electrons, assuming that each atom contributes one valence electron to the electron ‘gas. nergy associated with one kmol of such a metal is 3 Now the ¢1 U= INakyT 2 (6.18) If heat is supplied to the metal, these free electrons also absorb Part of the heat, and the molar electronic specific heat is obtained as follows: du} 3 [Clas [2] 7 ky, =1.5R, (6.19) where R, is the universal gas constant. Be UC, Jey = 1-5 x 1.38 x 103 x 6.02 x 10° = 12.5 x 10 J/kmoV/K That is, molar electronic specific heat = 12.5 kI/kmoV/K. The above value of 1.5 R, due to free electrons is about hundred times greater than the experimentally predicted value. Since the heat capacity of a solid due to atomic vibrations in 3 R,, free electrons should make a significant contribution to the total specific heat of a metal. It is, however, that at least at high temperatures, Dulong and Petit law (lattice specific heat) holds good and the total specific heat of a solid is 3 R,; this means that the free electrons do not contribute significantly to the heat capacity of a metal. It is, therefore, concluded that the law of equipartition and hence classical Maxwell-Boltzmann statistics must not be applied to evaluate the electronic specific heat in metals, (ii) Computation of Mean Free Path ‘The microscopic expre: ion for the resistivity of a metal is given by the equation (6.12), z ne*t The resistivity of the most useful metal copper at 20°C is 1.69 x 10° ohm-m and the concentration of free electrons in copper, m = 8.5 x 10°8/m3. Thus m a ne>p - 911x107" © 85X10 x (16 x10")? x 1.69 10 .4T x 10-4 sec But (6.20) CHAPTER 6 Referring from equation (6.20) = 2.47 x 104 x 1. ‘The experimentally found value for A is about ten times above this value, not explain the lange variation in 2. values. Relation between Electrical Conductivity and Thermal Condus (Wiedemann-Franz Law) . Heat conduction in solids may take place either through the mechanism of the ator the free electrons. In metals, electrons are the principal carriers of heat energy. Free & being excited by energetic scattering centres and carrying the extra energy 10 in a cooler part of the metal. TH (@T/dx) Heat flow A Ir E+ (dE/dx) E Fig. 6.2 Flow of heat through a conductor” Consider a copper conductor in the form of a rod of uniform eros: assume that a thermal gradient be established. Consider three layers at A, Ba of the heat flow. Let each of them be separated by a distance 4, (the mean in the metal). [r+(S)ajirana 17-(2) sete sey emperatno te ae de “ dE Lette thermal energy ofan electrons in these ayers be| E+ “| andl ‘Thus the excess of energy carried by an electron from A to Bis [Z| ee de kinetic theory of gases, the number of electrons flowing in a given direction through unit area cee] . aia time is -nc, where n is the number of free electrons per cubic metre and ¢ is the average the! velocity of the electron. Thus the excess of energy transported by the process ‘of conduction throt#! unit area in unit time at the middle layer B is = ay eas 4 greh () arly deficit of energy transported through B in the opposite 1 _, (dE ~o() , Hence net heat energy transported through unit area in unit [GG sis mst be equal tothe product ofthe thermal condvtvity a2). By 3 |la =o [fF From the general definition Q=o,Ar rr “eat]-aft 1 67> IC lem simi ‘Now Thus, ‘The electronic specific heat associated with ea Now Thus, (ay ‘This linear dependence of 2 on absolute temperature is known as the law of Wiede oO Franz (classical one). The multiplying constant is called Lorentz number L, that is Blkale v=3[*] = 1.12 10% WOK? e The validity of equation (6.25) can be checked using experimentally determined values op, 6, Equation (6.25) is 386 Wm'K* respectively. era Rae ee OxT 581x107 x293 or L=2.26 x 10% WOK? The value of Lorentz number does not agree with the value calculated from the classical in thermal conduction is not correct. Quantum theory with Fermi-Dirac statistics reveals that o electrons near the fermi level take part in thermal and electrical conductions. IV. RELAXATION TIME, ' (/) Relaxation Time Suppose ata given time = 0, the average velocity of the electron is < v, >y. Just at that instant we the field off, so that < v, > will subsequently tend to zero. In other words, the velocity will by electron-lattice collisions. Now let us assume that this process follows the simplest la =<, >)e% § The factor tis a constant called relaxation time, because it gives a measure of the system takes to relax when a constraint (the electric field) is removed, When t = quat becomes = S70 e Thus relaxation time may be stated as the time taken for the drift velocity to initial value. Etectricat Properries of Meats 195 pifferentiating equation (6.27), we get aan ae [Zen] Hes at elec lan (6.28) a (i) Collision Time tae last section we introduced the concept of relaxation time ina highly mathematical way. However, tre shall now show that it also has a very simple physical meaning. Suppose that the probability of an - Perde sjectron making a collision in time dt is =, so that on an average there are A. collisions per second. e %, * tis now our intention in what follows is to show that ¢, = t. Let us assume that all electron collisions are elastic; that is, energy is conserved, so that the «peed of the electron is the same before and after collision. This assumption is not strictly correct, because a conductor warms up when current is passing; it implies that energy is transferred to the lattice during electron-lattice collisions. However, the actual energy change at each collision is minute, and the fractional change in velocity in a metal is typically: | in 10S, so that the assumption is entirely reasonable. If we assume that after collision the velocities are completely random, we have < Vz > aftr = 0. SINCE before = <2 it follows that the change of < v, > on collision =~. 1 There are — collisions per second, so that te Large © Small ¢ Fig. 6.3 Relaxation of a system of electrons to equilibrium after removing the electric field as given by equation (6127) <> Te However, we could equally write the left-hand side of this equation as the rate of change of (6.29) ea CHAPTER 6 Average final direction 7 ¥, cos 0 Incident \ - F electron \ 7 “es A ri = IN ™ Fig, 6.4(a) Scattering equally probable Fig. 6.4(b) Scattering predorninantly in the fonwang n all directions, so.that the average direction at average angle 0. Change im any ingle is 90° herefore, on an average, v, (1 = €0s @) (ce | Pacem aeee so that on comparing equations (6.29) and (6.28) we see that t = 1, and the relaxation time, therefore equals the mean time between collisions. In deriving this result, we assumed that the velocities took up random directions after each collision (Fig. 6.4 (a); this is equivalent to stating that after each collision the electron had no memory of what went before. However, it is possible that the scattering is very weak, in which case such an assumption is unrealistic. If, on an average, the velocity direction changes by @ on collision [see Fig. 6.4. (b)], then the change of < v, > is on an average, ~ < v, > (I~ cos @). Consequently cos6 >] oC shere. 7 is the relaxation time and t, is the mean time between collisions. In other words, the relaxation time is then much longer than the mean time between collisions; sse electrons have a memory. iii) Mean Free Path The mean free path A of an electron is defined as below: Lz Gt, In case, the obstacles are hard spheres, the mean free path A is determined by the concentration of these obstacles. Then for a given concentration of obstacles, the collision time t. becomes inversely proportional to the velocity = of the electrons. Then, with < cos @ > independent of the velocity ofthe lectrons, in accordance with equation (6.30), the relaxation time t becomes proportional to 4, and ence, inversely proportional to the velocity @. But in the preceding article, we have taken relaxation ne to havea single value, having no relevance to the distribution of electron velocities. It may, howeveh shown that for metals the relaxation time t, which eventually determines the conduetivity of metals, ays pertains to a single velocity; i.e., the one corresponding to the Fermi level Ep. The proof of this ment involves Fermi statistics which is treated in the next section. Evectricat Properties Vv. QUANTUM THEORY OF FREE ELECTRONS sical model of a metal is eminently successful in explaining many properties of metals. The jurstanding properties of metals are their high electrical and thermal conductivities. In metals, the Ofience electrons are not merely dissociated from their parent atoms, but are assumed to be as free to vrove about inside the metal as the atoms or molecules of a perfect gas. In general, their motion is fandom, but in an electric field, the negatively charged electrons stream in the positive-field direction find produce an electric current within the metal. In order to prevent the electron from accelerating indefinitely, it is assumed in this model that they collide elastically with the metal ions. This leads to a steady-state current, which is proportional to the applied voltage, obeying Ohm's law. Although, the details of this classical model require modification, it should not be overlooked that it was proposed almost fifteen years before the periodic nature of atomic arrays in metals was firmly established by X-ray diffraction. Although it correctly predicted the magnitude of electrical resistivity, dependence on temperature was proportional to 7"? instead of the observed linear dependence. Another success of the free electron model is the semi-quantitative agreement of the Wiedemann- Franz law. One more characteristic property of metals is their high luster and complete opacity. Suppos an electromagnetic radiation, such as a light beam, falls on a metal crystal, the free electrons are set into forced oscillations having the same frequency as the incident light. Thus all the incident energy is absorbed by the free electrons, and the metal appears to be opaque. When an electron, thus excited, returns to its initial state, it does so by emitting a photon of the same energy as it absorbed initially. The light given off by the electron goes off equally in all directions, but only the light rays directed towards the metal’s surface can get through. Thus the metal appears to reflect virtually all the light that is incident on it giving it the characteristic metallic luster. The two most notable failures of classical free electron theory are: (i) Its inability to explain the discrepancy between the observed and predicted electronic specific heats of metals discussed earlier (equation 6.19) and (ii) each electron has a magnetic moment associated with its spin, and classically, should therefore give rise to a paramagnetic susceptibility inversely proportional to the temperature. Experimental results, on the other hand, show that the paramagnetism of metals is nearly independent of temperature. These quantitative errors were removed by Sommerfeld in 1928, who substituted the quantum statistics of Fermi-Dirac in place of the classical statistics used by Lorentz. Finally, it should be noted that even this quantum free-electron theory fails to explain two important aspects of materials. It is incapable of explaining why some crystals have metallic properties and others do not and why the atomic arrays in crystals, including metals, should prefer certain structures and not others. These properties of crystals can be explained by the more general zone theory discussed in the next section. The difficulty in the classical theory arises because the Maxwell-Boltzmann statistics permits all the free electrons to gain energy, leading to the prediction of much larger physical quantities than are on ly observed. When quantum statistics are used, however, it tums out that only about one per cent of wi tee electrons can thus absorb energy. The resulting specific heat and) paramagnetic susceptibility intent in much better agreement with experimental valves. The introduction of Fermi-Dirac statistics ee « calculations, therefore, is the prineipal accomplishment of the Sommerfeld theory. He retained eal as fee electrons moving in a uniform potential within the metal as in the classical theroy, but Faas ne5 era obeying the laws of quantum mechanics instead of those of classical mechanies. A amiforn ea tiation was Jalen introduced by postulating that the potential field inside the metal is not period instead, the field experienced by a moving free electron varies periodically with the y of the crystal. a The clas: a he bounk shown in Fig. 6.5: clectrons are trappet “The two main problems now to be treated afe: 1. What are the possible and permissible enerBY box? +, How are the very lage number of ree electrons diss states? “The first problem is salved by giving sui second one is answered by applying Fermi-Dirae distribution formula to Vi. CLASSICAL WAVE EQUATION Fig. 6.6 Cla Tsino=Tuno=7(2 a Over the element AB the rete at which the tangent of th ie angle # (ang) = 4 a dk fa?y] © \ pe tral change iS la Sx since 8/ and 8x are practically the qT srerefoe, the component al Balong BB, will be Now the net force tending to displace AB is, therefore Nov fp 2, \ (2) gale ee Ca (e-()s] 72 ‘This force must ‘be equal to the product of mass and acceleration, Hence, or Here v is the velocity of the transverse wave. Since y is a function of both ¢ and x, the total derivatives must be replaced by partial derivatives. (6.31) ‘This wave equation is also valid for several other types of waves. In practice, sound waves in air and light waves in vacuum are two of the important types o! waves that obey equation (6.31). Equation (6.31) admits many solutions depending on the initial displacement and the imposed toundary conditions. Of the numerous possible solutions, the solution corresponding to the case of undamped, monochromatic harmonic waves in the positive x-direction, is of particular interest. The eneral solution in this case has the form This can also be written as (6.32) (6.33) CHAPTER 6 ae ation (633) has significance in the case 3 string from its nO” not be Only the real part of equ where y represents the displacement of tht partis discarded as irrelevant. ‘This, however cant pecause the wave function of ‘ge Broglie, unlike therefore, be complex. ‘We now proceed (© get the Wa) ‘with material particles sue! sh as electrons. ‘The wave funetion $ ‘the solution for ‘undamped: ATION the positive ai direction. sER’S WAVE EQ oving freely 0 positive * rection ha jem ap for a particl wes in the harmonic Wa function of x and f a mown relations for WAVES: ‘Thus, Wisa my and We now use © = ‘yh, the well kt peAexp [--(-3] gic [--»(o-»9)] = Aexp [ ani(w -4)) +s quantum theory and Einstein's mass energy TENG Applying Planck E= hv and E = me? and hence V Momentum of the photon is ceo ith Thus for electron i : peer or e fee 5 where, v is the velocity of the electron. Substitut ; ing these values in equation (6.35). 35), we get (6.36) (637) The general expression for the wave function given by equation (6.37) is correct only for frely moving particles and hence may not work well in a situation where the motion of a particle is subject to certain restrictions. Therefore, we are advised t» obtain the fundamental differential equation. for, which may then be solved in a specific situation. We thus begin by differentiating, equation » twice with respect to x and once with respect to t eee i ip a Tae [- = aps) (@ wot pw (6.38) a (6.39) io The total energy of a particle is sum of its KE, and PE, E zm? Va) ne (==) 4 2\ m 2 ey 2m poe aa Multiplying by on both sides, we get Po avy 2m 64) Substituting the values of EY and pF from equations (6.38) and (6.39) in equation (6.40), ye. 2 [ae naw _ Wid i | teed oe koe re (641) Equation (6.41) is known as one-dimensional Schrédinger’s time-dependent Wave equation. Th use of this equation, rather than Newton's second law, to solve a dynamical problem is similar he ‘wave optics rather than geometrical optics in solving a problem in light. One point about equation (6.41) needs special mention. This equation is a complex equation because of the presence ofthe factor; ‘This means that Y is a complex function, but of real variable x and t. Y cannot, therefore, be observed like the displacement of a string, In three dimensions the time-dependent form of Schrédinger's equation is poe [eet law aw ow a here, the wave function Vis function of x, y, z and t, The quantity ay ow ay Oreo ften abbreviated as V¥, where V?, the laplacian operator is given by + (9) Vill. SCHRODINGER’S EQUATION FOR STATIONA! [he one-dimensional ‘Schriidinger’s time-dependent wave equation is how _(w \law as “sal es __This equation can be simplified for a great variety of problems for which the potential enerzy Micle does not depend upon time explicitly, the forces that act upon it, and hence V, varies witht ion of the particle only. Iris always possible to fin, ns of Cringe ; dent of time. I partite can he dary Re ein, Sioa an ey function of x alone and @ YS Avex i EYp) ow ote i ee % ~ A Vexp ‘iy cy ey xp (i Exipy ay i iby dx AEP iB iu 2, & a Fac (i En) Substituting these values in equation (6.41), we get 2) 42, (HEX) sxnieiny=[ 4 ) Beene i h ne alee alone and is ind Here y is a function of x ates form of Schrodinger equation. In three-dimension: f independent o 1 appreciated with the following simple analogy, faves of any wavel if boundary conditions are imposed), a'€ faves are set up with specific Here w is a function of (% ¥ ‘The difference between ‘¥ and W 'S wel a string of length a. one end of the string is free, the sing, However ifthe fee end is fixed (that is, range of values is n0 longer possible and stationary ¥ tanding wave may be represented By * where 1 isan integer. Such 2 spo A coe Nese Substituting = = sn the above equation, we have pe sin 2 cos mee T a=2 n=3 canding waves along a stretched string Fig. 6.7 where? i represents the transverse displacement of the string. Here ¥ is. the ¥(0)=A sin and 60) = cos )=cos22, Each point onthe string vibrates in simple the amplitude y ofthis motio along of this motion, alth« i th 5 although varying from point to point along strings rs respect to time. Thus equation (6.45) indicates that only certain ni ot ‘and therefore the energy of the motion) are possible. ; ; s IX. IMPORTANCE © The Schridin ger ‘avation, like Newton's laws of motion, is a fundamental relationship, rence to a v; he ast am ecrimestal choca Really : ai ental observati experiments have oe ae tae vations. However, Newton’s laws microscopic scale. In evaigaras not completely wails amt 7 , the Scheidinger atin ap eke ene es, Ths. the deductions Which have hog been confirmed experiment have been found to le. yation has been found to I ally, Ad t0 the text, Newton's laws seem to h; conte and further, same aE bee, hrOdinger ¢ easier (0 apply and apprec: much 9 atoms, equally parison to ator y par accurate om i ‘yom. But, When Newton’s laws d them. But system 8 the latter ree wit rodinger e «d events, the § ns and events, th Spite of its diffi X. PHYSICAL SIGNIFICANCE OF THE WAVE Fi iculties, rhe wave function Y enables all eoeral a complex function, does 1 pnly that an electron is present ori Possible inform, not have a direc physical mes present: we cay With a moving S no value of the wave function associated we can detect \ggested that the zin space at the tthe time, But ¥ itself cannot never detect a part Particle atthe partic ‘ood of finding the particle there at actualy equal tis probability because of the following reasons: If Prep Soe particle somewhere at a given time, tant ca corresponding-to the certainty of its g But the amplitude of meaningless. Hence th oft. Bor su ime instant, is related to the likelih ree resents the probability of finding 1 between the two limits; Zero, ‘0 the certainty ofits presence » and w.negative probability is ‘on ', which is complex number ofthe for (4+ ib), cannot be an dbsenvable quantity and it has no direct physical meaning. While itself has no physical interpretation, we Notice that the square of its abs, [¥}, can be given a definite meaning by looking upon the ¢ wave. The intensity due to monochrom, solute magnitude, amplitude. By analogy, we may interpret that [#2 with the de Broglie wave described by the wave fu inction at the point Proportional to the value of |"? there at that instant. % Y, Zat the time instant ¢ is Since the wave function is a complex Quantity, it may be expressed in the form (x, y, 2.) = (@ + ib) where, a and b are real functions of the variables (x, y, z, ) and i= Jol. The complex conjugate of Y which is denoted by ‘Y’ is obtained by changing i to ~ i. Therefore, we have Ys y. 2.) = (a= ib) Multiplying the above two equations, we have ; WY" = (a + ib) (a — ib) =a? + The product on the left hand side of this equation is ae by P. Thus. a, P=I¥Gy40P=0+P . : i i db product of W and ¥*is real and positive if ¥ #0. Is Farag enan rats a, 4 *) land itis called the modulus of . The quantity 1 (x, y, z, 1) [is density is obtai ini 1 wave with an The mathematical expression for a complex wave is obtained by combining a real inary one, o oc If a Acos2n( w=) wr =A sn 2x [-$) vewsreacaae(w—) +isnae(ucd 2 ( 1-3] : : W = Aexp | 2mi| vi 2 ‘The fact expressed by equation (6.46) has been used by Maxborn to give the interp (ax, y, 21), which is as follows: For the motion of a particle, the quantity Pdv= (x, y, 21) YC, yz) dv =I 4G, y, 2 0 Pav isthe probability that the particle will be found in a volume element dy surrounding the point; (Gy, 2), at the instant 1. For the motion of a particle in one-dimension the quantity, Pdx=V(x, 1) ¥(x, 1) de=|¥(x, ) Pde the probability that the particle will be found over a small distance dx at position x, at the insta lis case | ‘V(x, 1) |” is called the probability per unit distance. Wy, z )=uG, y, oem WG, y, zt =u" (x,y, em alue of the constant energy E, we now get =ulx, y, 2) u(x, y, z) dv =| uC, y, z) Pedy or the wave function satisfying the time-independent Schrdinger equation, the p independent of time. Limitations of y ause of the above interpretation of lil as a probability density, the solutions that cd from the Schrédinger equation are subjected to certain limitations. The most ( W must be finite for all values of x, y, z. Gd) y must be single-valued, i-e., for each set of values of x, y, z, y must have (iid) y must be continuous in all regions, except in those where the potential energy | (iv) The first derivative of y must be continuous, _— a Euecrnicat Propesmies of Meras 207 jon ofthe tree-dimensional, time-independent Schréitinger w - slut The ® jon for the wave function for the particle at ‘ave equation for a particle gives The expression is, expr 5 a function of ( d contains nes ant: The) TaluhOHt eee a (x, , 2) and com as paetiite constant. The value of the constant with its sign is determined by the following onsiderations ; since |'¥@, 9, 2) dv is the probability that the particle will be found in a volume element dv ing the point at position (x, y, roundin 2) the total probability thatthe particle will be somewhere in must be equal to 1. Thus we have race ms i Iya,» 2) Pa spore, ia function ofthe space coordinates (x,y 2), From this “normalisation condition” we can find the value of the constant and its sign. A wave ‘anion which satisfies the above condition is said to be normalised (to unity). ‘ondition for the wave function for the motion of a particle in one-dimension is The normalisi From these equations we see that for one-dimensional case, the dimension of y (x) is L-!? and the three-dimensional case the dimension of y (x, y, z) is £32, The main task in applying the Schrédinger wave equation to a particular problem is to obtain a suitable expression for y showing how the wave amplitude varies with distances along x, y, z-axes, and then to derive solutions of the differential equations. There will be in every case, many expressions for satisfying the Schrodinger equation, all of which are, however, not acceptable. Only those functions which satisfy certain conditions are acceptable and are called eigenfunctions of the system, while the snergies B, corresponding to these eigenfunctions, are called eigen values. In looking for solutions to the Schridinger equation, the following conditions are operative: |. The solution must be everywhere finite (not everywhere zero) and single valued, implying ‘hat a solution for y exists which can only have unique solution for a given position in space. 2. The solutions must be continuous and its first derivative must also be continuous. ‘hen 4 panicle is nt subjected to any external force, so thatthe particle moves ina region in which its ‘ental energy is constant, it is said to be a free particle. Such a particle has definite value of total sca a dfnite value of momentum, But the positon ofthe particle i completely unknown. Inthis ne ‘We consider application of the Schrédinger equation to two simple cases: (1) One-dimensional ete, f a particle under no forces, and (2) one-dimensional motion of a stream of particles, and the OF «rectangular potential barrier on the motion. a CHAPTER 6 '208 — Souo Stare Prvsics i, (i) The Free Particle . Suppose a particle of mass m isin motion along the x-axis. Let no other foree act on the part the potential energy of the particle is constant. For convenience, the constant potential cies to be zero. The one-dimensional, time-independent wave equation under this condition is gieay® ay [7] Tiki + +| => |By=0 ; dx? le By 64 Arm 2 oe #, where p, is the momentum of the particle. im ie, or The solution of equation (6.47), we get (x) = Ae + Be (6.9) where, A and B are constants. % Equation (6.49) gives the time-independent part ofthe wave function. The complete wave facing for the particle is given by Cx, ) = yore Substituting the value of yl) from equation (6.49), we get sie > M(x, )= (Ae + Be 5 (x,t) = Aetot—f 4, Berke rk mln) side of the equation (6.50) represents the wave travelling in the positive x-direction, term represents the wave travelling in the negative x-direction. Therefore, for the motion of ti in the positive x-direction, we have ; 2m po lag WO) = Ae and Yr, ) = Actor ts) The a momentum operator operating on the wave function W(x, ) sive, ox ________ Bin i a equations (647) and (6.48), we have “fom e gut p,= hk nae ARY per T are eee ae 7 ae gh (6.50a) riseqnion shows thatthe wave function Ya, 2) for he pul is an eigenfuneton ofthe Pe umopetatorand he momentum, sth eigenvalue of the operator. Hence the momentum i arp with the VAlUE Py aa tity ffnding the panicle between x and (x+ ds) sien by P de = V(x, ) YC, ) de = Constant = A? ‘perefore, the probability density P forthe postion ofthe particle with the definite value of stant over the a-axis; ie., all positions of the particle are equally probable. This gunentu 18 CO tained from the principle of uncertainty. ison is a0 0 ccording to the interpretation of the wave function, the probability of finding the particle soenbere in space must be equal to 1 finding J ¥0,9¥@0de=1 re But W(x, 1) ¥"(x, 1) = Constant = A? ‘Therefore, the integral on the left side of equation (6.51) is infinite. Hence the wave function for ‘parle cannot be normalised and A must remain arbitrary. The difficulty arises because we are ‘eng with an ideal case. In practice, we cannot have an absolutely free particle, The particle will sings be confined within an enclosure, and hence its position can be determined. This means that its «eatum cannot be determined with absolute accuracy. (The Rectangular Potential Barrier: The Tunnel Effect : ‘patil i impinging on a potential barrier with energy certainly less than the height ofthe potential om ot necessarily be reflected by the barrier but there is always a probability that it may mee a and continue its forward motion. This surprising phenomenon is known as tunnelling fay gan 258 of number of electronic devices. One ofthe most fain these isthe tunnel is used as high-frequency oscillator. Fig 68 Rectangular one-dimensional potential barrier of height Vp > & fy ‘A rectangular pote extends over the region (Ui) from constant and equal 0 Vy On both si ‘vhen the particle is in these regions, no forces act on it. tial barrier of height Vp and width a for a particle is sho, = 04t0.x=a in which the potential energy (1) of we des of the barrier, in regions (1) and (IT), Y= F Suppose a beam of particles travelling parallel tothe 2-axis from left to } potential barrier. In the regions (1) and (LID, the energy of a particle is totally kinetic, 4 } (itis party kinetic and parly potential. IF E< Vy thea according to classical mechs 4 of any patil reaching the region (ID) afer eosin the region (I) is zero, ‘ sian mechani, the ransmision probobiliy sa small but deft all 1 Se called tunnel effect. ; Let y,(2), Wal2), a(x) be the wave functions for the motions of the partic d (II respectively. In these regions the time-independent Schriidinger will i uatio @:-2> 1, 7 becomes very small and it decreases exponentially with increase in the thickness of the barrier. Equation (6.70) also shows that ifa is constant, and E/V,, decreases, then T decreases exponentially. This, variation is shown by the part AO of the curve for T (Fig. 6.9). (c)When the particle energy E approaches the potential energy Vp of the top of the barrier, then will be << 1 Thus, sin h Ba = Ba and sin h? Ba Hence, 2m(V) — E)a? ieee cee 4E(V, — E) ‘ 2mVia* SER E> Vp we get 06 a 02 ee 0 2 : thy fig 6.9 Varden cana RS Vo _| 2m(Vy = Ba? @ 2 1 Moa 2H? Fe sin h Ba = 0s, suring both sides the above equation, we get quaring S sin h? Ba = ~sin? oa substituting this value in equation (6,684), we get 218 ‘Sou Stave Prsics This expression is the same as that obtained in the previous case (for F< V4), (b) Equation (6.72) shows that for E> V,, T= 1, when oa =m where n= 1, 2, 3, Substituting y2m(E-Vo) energy (E ~ Vp). Hence T= 1, when a= (2) This means that if the width of the potential barrier is an integral multiple of h there is perfect transmission of the incident beam of particles. Equation (6.72) also increases, T oscillates between the maximum value 1 and a value less than one (Fig. 6.9 Reflection Probability The reflection probability is given by =a) tal BY? , ( Fl is called the complex conjugate of( i) The quantity R is called the 4A, Dividing equation (6.66) by equation (6.65), we get Bees ay me + 5 Wa (6.74) This is the expression for reflection coefficient. Adding equations (6.74) and (6.68), we get Ra (6.75) The tunnel effect provides explanations for the following phenomena; | |. The field emission of electrons from a cold metallic surface. 2. The reverse breakdown of semiconductor diodes. 3. The electrical breakdown of insulators. 4. The switching action of a tunnel diode. 5. The emission of ocparticles from a radioactive element. a Fa ‘few situations in which an exact solution of the Schrédinger equation may be obtained. lt te hypothetical, but they are approximate to real situations and also serve the purpose of i ‘pplication ofthe Schrédinger equation. In this section we propose to treat by the methods “cosy shames certain ofthe simple systems for which the Schrbinger equation canbe solved Sota Balms ean idealisation of naturally occurring systems, but the consideration of them hich are ee as they furnish an insight into the methods of quantum mechanics and give results ul in the discussion of many problems of physical sciences. ny Xq pour umsdoo asp Jo sone sa “HSH Kags ays #44 jg) vorenba jo wonnos pssus8 ayy, Tursuo vst ABiouD amp 2s Levon 03 wonenbo ys 29 Su wo 2pm 2) sus A aug pur 4A "X08 sua quoted Jo 1d Hon saouryswnasi asaxg 29H v “apy uosanas 29 U8 re ta 51 010 IH “HoHNjos ou oq uota [1m azaip asneaag “9 = V2? 10H” gba wos ¥ Jo onjea oy SunMDsy ABs0urg Jo sonenund Sous y = (x) “A fi Jo wonow xp 405 suoNdUNy 2AvM 2Xp 2200H t “A@rou9 o122 2004 wt Jou AB1ouD os97 yim ojoruEd *0= 10 =u 10} asnnaoq ‘0 = M9461 0UNPD AM cuarsAs amp JO 24 Jo uonnjos ay 305 u =4 =0y 2am, 0=>y us y= (Ar ‘any 2m ‘uonrpuos puosss amp 3 o=@n7= 7 UIs y= (A 2ou4 puw = sf 198 om SHH OM o=@ao=*” le ass) t= 2 us aan ‘asojosouy. gun aq sms x0q mp UH SION HATO: ajonued agg un Aigegaud (es SL, p> h?. This is an illustration of Bohr’s correspondence principle which states that quantum mechanics must yield identical results with those of classical physics in the limit of large quantum numbers. (ii) Particle in a Rectangular Three-Dimensional Box We now consider a particle of mass m constrained to nove in a fixed region of space, which for simplicity, we assume to be a rectangular box with edges of length a, b and ¢ and volume v = abe. The potential energy V may be put equal to zero within the box; at the boundaries of the box and in the remainder of space we put V= es. We take our coordinate system to be the cartesian coordinate system with the origin at one corner of the box and the x, y.and z-axes along the edges of length a, 6 and c respectively. The potential energy is then V, =0,0 on? a? Substituting these values in equation (6.84), we get we YZ ae +X ay +XY ae =- Dividing by XYZ on the above equation, we get ax ay az ( ed) YZ W the first term in tis equation isa function of x only the second a function of y n of z only, while the sum of the three is a constant. This can only be so, in fact © constant. We must therefore have ; ow the ctferential equations in, y and z-coordinates may be writen as The general solution of equation (6.89) will be of the form X(x) = A, sin kx + B, cos kx When x=0,X=0, hence B, = 0 When x=a,X=0=A, sin ka Therefore, ka=ne k= =, where . a Equation (6.92) becomes X=A, sin Applying the normalisation condition jue > Substituting this value in equation (6.93), we have elle ‘Similarly, the solution of the equation (6.90) will be of the form ¥y) =A, sin ky +B, cos ky When y=0,Y=0, hence B, =0 When y=b, Y=0=. sink . pee Per ‘This will be true only when kb = nym Le, ky = ia eee nt ‘Substituting B,=Oand ie op in equation (6.94a), we get nes Y=Asiny Applying the normalisation condition between y = 0 and y = 5, we have r fia 0 ng afin? BP ay = 6 b 2 5 5 1 2 anf ‘Thus equation (6.945) becomes 2 aes Similarly, we can solve equation (6.91) to obtain z= [ye jstes ope Equation (6.88) is © integers. Equation (6 ‘ave functions inside the rectangular box for the Stationary states, $i sigenvalues of the energy of the particle, These v Where m= nF +n? +2 ‘and the normalised wave functions are a From equations (6,99) and (6.100), we derive the follo 1. The three integers n,, n, and n, which are called # a Completely each energy state. No quantum number taken as zero, then w(x, y, 2) = 0, which would the box. he sum of the ; t Squares 2. It should be noted, however, that the energy F depends only a a of the quantum numbers n,, ‘ny and n, and not on their individual values. My ; tationary state 3. Several combinations of the three integers may give Cate et a i wave functions, but of the same energy value. Such states an a degenerate. Example: 1, n=1,n,=1,7,=2 2. n,=2,n,=1,0,=1 3. n,=1,n,=2,7,=1 The corresponding wave functions are then mw org) = Asin 7 sin 2Rx RY «RZ Ver =A sin sin = sin cet eeie ic a DT Te =Asin — sin —sin — Tee pe se in each case the energy value is exactly the same; that is 2 E oi 8a’ sh Es = level for which the energy is ~ there are three independent states having qua ima (21 I)and (12 1). The level is, therefore, said to be three fold degenerate of triply d noted that the groundstate (1 1 1), as also several other states like (2 2 2), are non-d Ttcan be shown with an example that the degeneracy breaks on introducing a's to the system. Consider, for example, the triply degenerate level, having E re value, there are three independent energy states having quantum numbers (21 1), (1 7 My le Let the length of the cubical box be increased by a small amount, da breadth and thickness unaltered. The corresponding change in energy of the. 4h an energy <2 associated with the x-direction, will be different from th : 2 and (1 1 2), which have an energy io associated with the x-direction, The rst state (2 1 1) is 4h? ama?” Bavtah day? = Br Bl whic yp is shown below: in Fig. 6.15 by comparing equa 6.101). lines which were si XIV. FERMEDIRAC STATISTICS AND ELECTRONIC DISTRIBUTION SOLIDS sal “The most important assumption of classical statistics is that any number of particles may haye idem energies. Though the classical approach gives an understanding of the conductivity of metal, i sive proper explanation forthe contribution by electrons othe specific heat and magnetic suscep As the electrons are free, the total molar specific heat of a metal at room temperature: 3R, ¥ 2 1e specific heat of a metal near room temperature is only 3, 3. 0 PS be [34, Aas | 3K, being the contribution from the lattices and experimentally th 3k, asyocited withthe electron must be very small and not —*. In order to account for the unexpectedly low specific heat of metals, into the statistical behaviour of the electron gas. Such an electron gas behaves fi particles and hence obeys Fermi-Dirac statistics. Fermi-Dirac statistics is applie particles (called Fermions) which are governed by Pauli’s exclusion principle. S the behaviour of a free electron gas taking into account the quantum theory principle. He considered the distribution of a large number of electrons of thermal equilibrium among the various states in a three-dimensional box. Assuming that the metal is in the form of a cube of side a, we hay function of the stationary states as 8] nm — nny ge | sin sin za, sia within the metal and zero outside, and the energy levels are we mat Ing +n} +n2] where, n,, , and n, are integers. ara amar With According to the Pauli’s principle, not more than two electrons hat, atthe absolute zero of temperature, two electrons will go into » next higher energy, and so on, until all the electrons are allocated ince the number of electrons is very large, it is thus understandabl perature, some electrons have kinetic energies of several electron lacroscopic dimensions, say centimetre cube, the energy of the ground sta ‘der of 10°'S eV and hence may be taken to be zero for all p acing between consecutive energy level is less than 10-eV, a garded as almost continuous or sometimes “quasi-continuous”, 2 92 el ny +n; +n? Now we shall consider how the many electrons of a real materi ergy states. Since we are talking about an almost continuous rang »bability of occupying a given state by a continuous distribution i etron occupying a given energy level is given by is called the Fermi function, Bis the ene! jer FE ; ~1s the energy of the level whose occupancy is being sed, Eis the Fermi level and is a constant forthe particular system. At absolute zene FUE Sf, and FUE) = | for E < Ep. Thus at absolute zero, the Fermi level divides the occupied sales fF sanoccupied states: eC isthe highest energy state forthe electrons to occupy at absolute 20, fo Caras Degree of wth) degeneracy @2) 749) 214 @2y Way ArH Six-fold degenerate Gy TD B19 Three-fold degenerate Three-fold degenerate ‘Three-fold degenerate 321) |. Six-fold degenerate Non-degencrate Three-fold degenerate ‘Three-fold degenerate ‘Three-fold degenerate Non-degenerate Fig. 6.16 Energy levels of particle ina cubical box The Fermi function does not, by itself, give us the number of electrons which have a certain energy; it {ives us only the probability of occupation of an energy state by a single electron. Since even at the highest energy, the difference between neighbouring energy levels is as small as 10° eV, we can say ‘hatina macroscopically small energy interval dE there are still many discrete energy levels. So we can ‘aoduce the concept of density of energy states which will simplify our calculation considerably. To ‘oow the actual number of electrons with a given energy, one must know the number of states in the Sem Which have the energy under consideration. Then by multiplying the number of states by the batilty occupation we get the actual number of electrons. If N(E) is the number of electrons in a “em that have energy E, and Z(E) is the number of states at that energy, then N(E) dE = Z(E) dB F(E) (6.104) ‘nay ,._.V€ know that, the number of energy states, with a particular value of , depends on how "combinations of the quantum numbers result in the same value of n. Since we are dealing with ae almost a continuum of energy levels, we may constru ce of points represented by the nn, and n, and let each point with integer values ofthe coordinates represent an energy suds A radius vector n from the origin may be drawn to point n,n, in this space, and al the surface ofa sphere of radius n will have the same energy. ‘Thus n represents a vector to n, (=n? +n? +n2) in three-dimensional space. In this space every integer specifies a unit cube contains exactly one state. Hence the number of states in any volume is ‘numerical value of the volume expressed in units of cubes of lattice parameters, a Fig. 6.17 Density of energy states calculation ‘Therefore, the number of available states within a sphere of raditis n is given by 1 a 4g accounts forthe fact that only positive integers are allowable and thus only ere alone be considered. Again the number of states within a sphere of radius (n 1 4n g ( Be + dn)’. Thus the number of energy statés having energy values lying bi (ée., in the energy interval dE) is given by 1(4n 14m x =-/— Bele lee (E) dE g intan (2) x T AB) dE = —U(n + day — 13) Gl + dn’ + 3nd x © = = Gn? dn) = =n? g Orda) = 0? dn fi -(3) Sma? n= \ >a || ae (ae substuing these values in equation (6.105), we get 2(6) d= (E)ntnam Gs ME) dE should be now remembered that the P. se, so thatthe number of energy levels Fig. 6:18 Density of energy states fora free electron gas re zeae = *[%e)" BIE a H) dB in a given energy interval dE will now be tual number of electrons NCE) dE in a given energy inter \de ti as sie of energy states in the interval by the Fermi distribution function, 5 n btn ee (EB) dE = Z(E) dE F(E) n)f8m]" wovae-(5) pt) ‘This distribution is shown in Fig. 6.19. Fig, 6.19 The population density of a free electron ga At absolute zero, the distribution those above E, being, of course, problem is simple, all states up to F are empty. At higher temperature the typical tailing app very small compared with the scale of the distribution until reached. The relation for the Fermi level in a metal is possible at any tempet degree absolute. The general expression for N(E) dE is n[ 8ma?]"* ale] Bide Re T=0K, F@)=1 since < By Thus, [@ae= (See) [en ; 8ma? ]*” w= (5) omg , wie] ve fore, the number of clectrons per unit volume (called density of electton) b ; the literature, @ temperature, Ty ‘rm velocity are used, They are calculated as follows: Fo F(B) FE), T=0K 1 1 Er 20 The Fermi dist Fig. 6 yee tribution function at different 2x32 16%10" 911x p= LL x 106 mis kyl p= Ep Ep and Figure 6.20 (a) shows the Fermi function ata finite temperature, Say, 7=300K. The Fermi curve at this temperature differs from the Fermi plot at K in the smearing ofthe drop from 1 t0 0.In ‘the actual case, the extent of the smearing is very small except at very high temperatures. ‘This can be seen by considering an example, say F(E) = 0.2 at 300 K. FE) =02= —s 0.2402 =1 0.2e' = 0.8 e=4 hy Tp=37100K 32x 160x109 138x107 Fe) 10" rit Eleciron concen- |" Ferm ener) atl) | Femi ¥atoely a) ae OK (Es) ineV | OK (v4) in 106 mis 4n 42 129° 2.65 323 107 140 212 0.86 ; 091 158 075 8.50 7.05 137 5.85 5.49. 139) 5.89 552 139 . 8.60 7.13 158 a 460 468 128 | ms aa 3065 13 | a Al 18.06 1163 ne a 49 8.60 7a Pa : “XVII. HEAT CAPACITY OF Table 6-1 Parameters cotineted with Fermi energy for metals sius the extent of the smearing (about 0.035 eV) is very small compared with the energy V. on of 5 1, Atabsolute zero, -F(E)= | for E< Ey eal F(B) = 0 for E> Ey % tO K, Ey is the maximum energy of the occupied state. 2, As temperature increases, an electron may get energy of the vacant state, and so Fermi function falls, But for energies below AT, the value of F(E) will be unity; ie., only in the region of Ep, i EE, F(B) will be > é Clasical statistical mechanics predicts that all particles in an assembly: ad so contribute to the heat capacity. Hence, the electronic contril A 3k, ; ‘ovalestmetl willbe 5 (Nyky)= =". But the observed electronic is usual LT as ey level a lly less than 0.01 of this value. This discrepancy is explained as ‘When we heat the specimen from absolute zero not every el ected classical, but only those electrons in orbitals within ‘one are excited thermally; these electrons gain an energy whit i eee os thet ano 7 the i » qualitative solution to the the heat Pied energy states. This gives an immediate qualitative 60 free elocg Pty %f the conduction electron gas. Thus the mum cnr ie fig Will be E,. Hence the fraction of ‘electrons that will be exc! =3 Kinga kyT _ 0.025 _ 9,005 = 0.5% . Ey at Ordinary temperatures, less than 1% of the val TeSUIt contrasts with the classical view in which all $0 contribute to the heat capacity. [ ic energy of the order 5 ky: Hence the ener ence electrons contribute to the heat ie electrons in a system would absorb Each of the electron may absorb a kinet with a kmol of the metal will be fo U=M% Tae (347) E, J\2 v=[ (Gr) ee ‘Thus, while the lattice contribution is 3R,. This value of 0.015 R, agrees with the experit at very low temperature, the lattice contribution is small and falls as T? and the elk is significant. “0, MEAN ENERGY OF ELE ‘The Fermi energy at absolute zero is given by equation (6.110), ie., Yan? (rales It is now possible to estimate the average electron energy at 0 K. To dot energy U, at 0 K, which is given by the following equation: fy Coes J Z(B) dE F(E)E with F(B) = 1 0 From equation (6.106) _(3)[ 8ma? |” on=(5)["] fear d (ix. EFFECT OF TEM FUNCTION Since the Fermi temperature of the electron over a wide range of| ‘much, equations (6.122) and (6.125) may be assumed for further ‘matter. The following details are not required up to equation (6,125) In the above treatment of free clectron theory of metals, we assumed that the zero degree absolute and, in accordance with the Pauli’s principle, we fill the upall electrons at the Fermi level of energy E,- But the kinetic energy of the the temperature is increased and some energy levels are occupied which were and some levels are vacant which were previously occupied at absolute zero ‘The number of electrons dN having energy values lying bet dN = Z(E) dE fe) N= Jaw dE fE) 0 For solving this integral, we shall first evaluate the integral given by e 7 d I= | fE) 7p (BE ° IRE. is the Fermi distibution funtion and g(E) any other function, such that g(B) —> 0 ‘ating equation (6.115) by parts T= (0) Ely - [bf de ‘The first term on the right side of the above equation vanishes at the upper limit probability of finding an electron with infinite energy is zero, The same term Vanishes atthe because we suppose g(0) = 0. Now expanding g() by Taylor's theorem about Eas “ a(E) = wll) +(E-E) + ‘Substituting this in equation (6.116) f 1 j 1=- fr i Hip +(E-E) £(E) + EEF gE } 1 yp) + Ly Ep + LE + ¥ where, ly =~ f [(E) dé 5 b= -fe-Ep four f 1.=-()fe-etroae ; However, at low temperatures such that ky7”<< E,, the derivative f(E) is large only at r £, as shown in Fig. 6.21; its negligible for other values of, particularly for negative vah lower limit on the integrals may thus be replaced by ~2o; we shall now solve the int Ds: § ) aE) OE —E E Fig, 6.21 The functions QE) and - ne he [rede ==. = RDI 4 = es | bal” Ite GF Lew 1 Jee) Hey over, f) | Lixerg @-Eyigr Thus F'@=-| lee ry ) about wel kyr Yultse E-E if E-E, ie Tin alee ee: [ise | 7hs imtegral vanishes because f“(2) isan even function of (E-B), andthe itegral is thus an “Hlucim, which when integrated from ro t0 = wil yskdney, : 1)\7 oe Xow ty=-(5) fee rroae 2, eat “A arp Ee ie a | Pat E-E, ae i ie, fs ee (kD) de = dB “siting these values in 1, relation, Lf kat) x%e! ig) ae 5 bey el) y (heer)? 2 . , 7 The solution of this standard integral is. P tyr? Thus equation (6.117) becomes (referring equation 6.115) ; 2 = [eB fl) dE = t+ Ug'e' Gp ., 1 Equation (6.118) is a convenient form for working out the Fermi integrals given (6.112) and (6.114). But itis restricted by the condition g(E) = 0 when E'=0 and by the kT << Ey - Thus, to evaluate the integral given by equation (6.112), we choose & aB)= [28 ae 6 : ‘th g’ = Z(E) and g” = Z'(E) Referring equations (6.112) and (6.118), we write i a J Z(E) dE + 0 © . 2 N= [E26 de= 7 0 ‘i AtOK, FE) =1 for 0 < E<&, and thus Ey N= J 2) dE FE) with F(E) =1 3 Fp w= | 20 ae a Equating equations (6,119) and (6.120), we get P 2 J AB) d+ * (DP (ZC), = 0 ‘We assume here that Z(Z2) does not vary much in the interval from E,toE, will be only a small fraction of £,. Thus the above equation can be written as. 2 2E) (,— E,)+ TUhyl 26) =0 we ABD is of the form CE"? [refer equation (6,106)], we write the above equation as ie CEP (E- + oe | =0 Nr (6.121) js the Fermi level at absolute zero and E, that at higher temperature such that kT << Ep en ¥ ; gover, the second term of equation (6.121) is a very small correction term that should be ow i: tively large term E,. For example, the Fermi energy of copper is about 7 eV. Thus cohracted from 2 relat (kal) 2 7 jt hand side. Thus equation (6.121) can be writen as oy Ep onthe ti tenenber tht itis applicable only at temperatures such that kT’ << E,, which happily ‘ilrange of temperatures for which the metals are solids. , For determining the mean energy of an electron at other temperatures, ‘we have ‘neg given by equation (6.114) choosing : a(6)= [EE aE 0 i a f= ” = [EZ(E)) §' = EX(B) and g” = (EZ(E)] Referring equation (6.119), the total energy of the electronic system may be i ; Pe a a : Us Jao dE f(E) (B) dE + 6 SpE o o a 5, By we a q ele Z(E) dE + —(kpTy? 2 us Jrzcerae . june + yt) {yet} , ,, Replacing the first term of the above equation by Uy from equation (6,110) y 2, ~ © ane U=U,+ ieee dE + (kD) iz vez) 7 The procedure used in equation (6.1204) is carried out here also to solve the secong above equation. In the last term Z(E) is evaluated at E,, rather than at E,. Thus, pre a U=U,+ZE DE, a + 5 |e ce", Ta 3 U= Uy + HE) Ep(E,— Ep) + (gl? [cfe"] CE? is nothing but Z(E) mW ry U=Uy+ HE,) Ep(E,— E,) + “a kel? ZE,) ‘Now substituting the value of (E,— E,) from equation (6.122) 1 (ky)? me 4 Ueto a 2 2Ep) Ep + “(kg TY ZE,) 2 © U=Uy+ Kal? ZE;) From equation (6.106) at E = E,, and from equation (6.108) otsitatng this in equation (6.124) substi (=) )? 3.0 DUNERL fe aeep 1 this value of Z(E,) in equation (6.123) HE ,) substittin N Er w 3 U=Uy+ EUs 5 err | E, Patical purposes, the Fermi level may be ¢ ‘The significance gf the above equations m The specific heat at constant volume is equation (6.125)] metal due ‘Thus for a monovalent metal, the heat capacity per kmol of the alelehoe| shell Kale w)\ a) KaNa For E,=5eV and 300K than that predicted by equation (6.1102) and this result oo ‘This value is higher a (6.126) differs from the classical result of ake bya fa erimental values. Expression ( °C, Thus the quantum mechanical theory P ich amounts to 0.025 for E; 5 eV at 3 3 alue of 5Ry- ‘much smaller than the classical v Ratio of electronic specific heat calculated quantum mechanically and class a [Cla (Q-Mech) _ 2 “\ Ep [C,]x (Classical) 3p ahs = 057 x 104T specific heat varies as the cube of the absolute temperature at Iow temp specific heat at low temperature is given by a F (Clr ae constants. But the lattice specific heat contribution at high temperature regiom dent of temperature. , perature. Equation (6.127) is thus written as ion ie @ q. 2 (p) shows that a plot of >> as a function of 7? gives a straight line, 6.22 ie ee sph the values of the constants A and Bf can be determined, which prom this BEP hy predicted specific heat values of metals, The plot for copper ing the ee for A. Using E,= 7.1 eV, one gets A = 0.508 using the equation (6.126) gece ale 0.095 1" ‘smaller than the observed value, This is resolved if we remember d a ian are not actually completely free, and that they rather move in a ‘hc effective mass m* instead of m (for copper nt = cei 6 we cnt with the experimental results immediately appear. mote 10 T(K) 20 @ Fig. 6.22 (a) Different contribution to the total s} (b) Kok and Keesom plot for as insulators with zero Ferm! SSS have high electric conductivity; where: conductivity. ati Since, for a free electron ; [I m Jn electron in the Férmi surface under the influence of v the equation of motion of eac of intensity, E is a a —|=Ee el ‘This means that in the absence of collisions, the Fermi sphere will be at a constant rate in k-space. The following approach yields the quantum mechanical pression for electrical conductivity. Integrating the above of motion, one gets end Ko 40) = [4 | lectric field is applied at time 1 = 0 to a filled tred at the origin of k-space, then in a will have moved to a new centre at eEAp vp Jeck=|—~\y Gfoea Sos he teal oe ico cach neve we ET os that the comeat sm fact came ont by sppreaches lead w the seme sat Ot mental comciesions. Tas the Feros sortase of coment is by clecones close m he Fo these phenom now be coxnpened fom the shows sa SSS 4 copper is L7 x 107 Om Tass vith mt = LOL = a | = er __taag f 1.01x10"' x 911 rep 85x10 x16" x10 x LT yg 1O1x911X107" «994 19-8 R5xL6 XLT =2.4x104s Axl04x v, p= 2X 70116210 _ 94x 108 ot, Fo 101x 10"! x 9.11 yp = 1.55 x 10° mis "i ig 4x 1.55 x 10 = 3.78 x10*= 378 x10- i” A, = 37.8 nm “e This value agrees with the experimental value fairly well £ |ETAI total number of atoms in the metal, and the mobility is limited primarily by crystal lattice which scatter the electrons and determine their mean free path in an electric field. At ¥ temperatures, however, the lattice vibrations are not limiting; in fact, one would expect them’ at 0K and the conductivity to become infinite, in a pure, perfect crystal. As shown in Fig. 623) resistivity in such a crystal should go as TS near 0 K as this is the temperature dependence scattering due to latice vibrations, and the electron concentration is independent of 7. No real completely pure or perfect, however, and imperfections of any kind-chemical impurities, lattice de even the crystal surface act as scattering centres for the electrons even in the absence of lattice’ ‘Thus the resistivity does not fall to zero at 0K, and the resistivity curve is displaced up amount which reflects the extent to which these imperfections are present and effective as S high temperatures, this displacement is insignificant compared to the much larger effect of the l= vibrations. For this reason the resistance ratio, the ratio of the resistivities at room temperature (0 helium temperature is used as a simple measure of the purity and per‘ection of a metal crystal the sensitivity of the resistivity to different imperfections varies widely, as seen in Fig. 623 (Os aj cetin impurities in coppers shown, Tt canbe seen thats litle as one part per million of 1 resistivity at 42 K by more than an order of ‘magnitude. Th resistivity of metals at ai cae, temperatures varies slowly with temperature, incre veld ia a 4 oe re gene ca characteristics of felectncl conduction 1 1, he i (onm’s 18), > For pure specimens, the eletrical condutivi sth temperature as follows: o« T and o, 5. Formetals that exhibit the phenomenon of temperature above 0 K and below the c transition (T, = 4.15 K) for mercury. as where py is a constant that increases temperature dependent part of the resist Matthiessen's rule. 5. For most metals the electrical resistivity Fg. 6.23 Resistivity of metal (a) its temperatures depené lis (ts dependence upon small concentrations of selected imp 6. The resistivity of alloys that exhibit orderdisorder transitions shoWS Pronounced jg, corresponding to ordered phases. ivi eneral e (6.12) helps on ‘The relation for electrical conductivity given by the general Ted Ona ca ne relaxation time + from the experimentally determined values of the ft “ nigh 4 a Meta For monovalent metals, the number of free electrons in a given volume is ed\ of; : : i vvected metals are piven in Table 6 Wa am present in that volume, The values of 7 at 0°C for some Se aeS es ee g 2 Wein : i mes a e order of 10°" sec from the table that the relaxation times are of the ord s A time does not give any significant clue except that it is small, it i ate e oF gee a rethe terms of the mean free path of the electron. Equation (6. id a & with the Fermi velocity v,.. Thus, if the scattering in i s , : oa hence A= p= vp tp. A values for eight familiar metals are coiven in Table 6.2. The point of special interest is the fact that ‘he mean free path of the electron is of the order of several hundred Angstroms. This Value is much larger than one would _~ expect if the electron really did collide with the ions as itlustrated in Fig. 6.24 and such a model inevitably leads to i) ean free path ofa few Angstroms, Especially in close-packed Fig. 6.24 Classical picture of scatter structures, in which the atoms are densciy packed, itis difficult to see how the electrons could travel so far between collisions. ‘of an electron by positive jon cores ‘The other difficulties of this model are the explanation of the temperature dependence, presse dependence, influence of impurities on the conductivity, etc. This paradox can be explained only by te use of quantum concepts, ‘Now a model must be developed, and it is necessary to take account of the fact that the electrons are moving through a region containing a regularly arranged lattice of positive ions. These ions produce a pattern of periodically varying electric fields (or potentials), and as the electrons travel within the ‘metal, they experience the influence of this periodic potential. The problem is analogous to that X-ray diffraction Drude-Lorentz assumes that the mean free path is of the same order of magnitude as the ine. atomic spacing, thus a constant; and the temperature variation of conductivity is due to the variation of the average velocity of the electrons. According to quantum statistics, however, the average ener hence the average velocity, of the conduction electrons changes very little with temperature, cand itis necessary to assume, therefore, thatthe observed variation of conductivity is caused by a change inthe mean free path. To calculate this mean free path (2.,), a knowledge of the source of resistance in metals. As essential. ‘The wavelength ofthe electron in a crystal lattice is given by the de Broglie relation. beh mv, {tis well known from the theory of wave propagation in discrete structures that when a Wate Passes through aperiodic lattice, it continues propagating indefinitely without scatering, The effet the atoms orions in the given lattice is to absorb ener gy from the wave and radiate itback, so that te result is that the wave continues without modification in either direction or intensity, The veloci Propagation, however, is modified. This is what happens in the case of an electron wave in a regu lattice, except that in this case we are dealing with @ matter wave. In other words, the electron wavell 4 perfect crystal resembles the electromagnetic wave corresponding to visible light which pas 4 Euecrricat Propernes oF SN {aC l. The only effect the crystal has is to introduce the refra we medium is. Therefore we see that, if the ions form a perfect lati ti } stall, te, b= cand hence 1 =o, which in turn leads to infinite conductivity. But i" ool the mean free path is about LO nm. The finiteness cf the conductivity must thus be eed af the ltice from perfect periodicity. Thus, in order to discuss the electrical resistance 12 pee ays im wich the periodicity ofthe lattice may be upset iy the Way 9 an types of lati imperfections which ean contribute tothe electrical resistivity are: ve wo main) ; f rm eal viraon ofthe ate, Which il prevent the toms from ever al being on tei tes at the same time. correct site 5 rhe presence of impurity toms and other point defects that will upset the lattice periodicity gif ELECTRON SCATTERING MECHANISMS! ANDI VARIATION OF! " RESISTIVITY, WITT TEMPEST acs sesible to explain the temperature dependence of resistivity using the following model: We assume hi i cnystalis free from the deviations of the type (2) mentioned above, and estimate Tp OF hp as emied only by latice vibrations, We, thus, proceed by assuming that the scattering of the electron sineis due othe finite size of the vibration amplitude of the atoms, that is, the vibration amplitude “ives a ealering cross-section; and that the mean free path is obtained directly by using kinetic: sey formal tis also assumed that at high temperatures the lattice v:brations are essentially classical. IEMs the mass of the atom, and (=f) is the restoring fore onthe displace atom, the equation afte atomic oscillator is it dx qi thao Geaahl ‘ . ea ct ees | i veo=2my =| (| 1s the amplitude of vibration. The potential energy of the oscillator will be (6.130) iE S8esot freedom, AC high temperatures, as the Einstein and the Debye frequencies are essentially Ste and most of the modes are the high frequeney modes, we write i = 0} = 03 (6.131) “ath hop = ky, :

You might also like