CHAPTER 1 DEFECTS IN CRYSTALS AND RELATED PHENOMENA 1.1. INTRODUCTION Real crystals are never perfect: they always contain 2 considerable density of defects and imperfections that affect their physical, chemieal, hechanical and electronic properties. The existence «of defects also plays an important role in various technological processes and phenomena such 3s annealing, precipitation, diffusion, sintering, oxidation and others. It should be noted that defects do which not necessarily have adverse effects on the properties of materials. There are many situations 2 careful control of the types and amounts of imperfections can bring about specific characteristics desired in a system. This can be achieved by proper processing techniques. In fact, “defect engineering” is emerging as an important activity.» ‘© wo . Effects of Defects on Material Properties Briefly consider the effects of imperfections or crystal defects on a few important properties of solids. Ca) The electrical behaviour of semiconductors, for exemple, is lareely controlled by crystal imperfections. The conductivity of siizon can thus be altered in typ)(n o p) and by over eight orders ‘of magnitude through the addition of minute amount of electrically active dopant elements(In this, dyjpel = elsatility 4S case, each ator of dopant, substitutional incorporated, represents a point defect in the con lace. The fact that such small amounts of imaury atoms can significantly alter the electrical Sirvngth) roberts of semionaucer) responsible for the development of the transistor and has opened up ‘the entire feld of solid state device technology. Practically none of the semiconducting properties that led to these enginedring accomplishments are found in a “perfect” crystal. They are properties inherent tothe defective solid state. ire nant tin is linc ves meh pear Sunt epe cred ddcmntan onsen bstorer ote warns ty oso tel seng hotness cre kerwton aban eons bettie mg 13) The ability of a ferromagnetic material (such as iron, nickel or iron oxide) to be magnetized and emagnetised depends in large part on the presence of two-dimensional imperfections known as Bloch walls. These interfaces are boundaries between two regions of the crystal which havea different magnetic state. As mognetisation occurs, these defects migrate and by their motion provide the material with a net magnetic moment, Without the existence of Bloch walls all ferromagnetic ‘materials would be permanent magnets. In fact, electromagnets would not exist fit were not for this type of defect. 4) The presence of surface defects such as cracks causes brittle materials like glass to break at small ‘plied stresses. This fats familiar to anyone who has broken a glass tube by fist filing a small notch 4 Maman Al aed tM UCT Supplementary Wand Out on Pont elects eae Phone Page tota9{or crack) into the surface. Removal of cracks from the surface of glass either by etching in hydrofluoric acid or by flame polishing’almost always raises the fracture strength. For example, glass in the absence of any surface cracks has a fracture strength of 10% N/m? (as opposed to real glass which has.a fracture strength of 10 N/m?) i To sum, cracks prevent brittle materials from obtaining their theoretical cleavage stress, while dislocations prevent ductile materials from obtaining their theoretical shear stress. . All defects and imperfections can be considered under four main divisions: point defects, or dislocations, planar defects or interfacial defects, and volume/bulk defects. ine defects i -2 POINT DEFECTS AND RELATED PHENOMENA Real crystals contain a variety of imperfections, or defects, by nature. Some of them are of atomic mensions, and are called atomic or point defects, while others are extended in one, two, or three dimensions, producing linear, planar, or volume defects, Many of the phySical properties of crystalline solidere strongly influenced by those =e) thus being described as “micrastructure- Sensitive’ “The well-known examples are mechanical Properties of metals: strength, ductility, and toughness} Also, the kinetic processes, i.e, diffusion and phase transformations, which control the microstructure formation, are governed by the density and mobility of the defects present in the materialoint defects mediate movements of atoms, dislocations provide ductility, grain bound often make metals brittle but also strengthen them by an mot Quantitative knowledge on the properties and beha ipeding dislocation mation, Yor example. ur of those defects is therefore ifiportant. Perfect erystal structures with all atoms situated in predictable lattice sites are rarely realised, Only in Specially grown, oriented semiconductor single crystals is this ideal closely approximated in an engineering materiahover appreciable lattice dimensions. Polycrystalline solids consisting of numerous small single crystals or grains of random orientation invariably form during production and subsequent processing of materials. th them, errors abound in the perfect placement of atoms and - the stacking of planes. Such effects tend to be magnified at the interfaces where grains fit together. At the outset a distinction should be’ made between Such crystallographic defects and gross manufacturing defects and flaws like cracks and porosity. This chapter focuses on lattice defects having atomic size dimensions. What are the implications of such defects? How harmful are they? What properties are influenced by them? First it is necessary to introduce crystallographic defects from a structural point of view. ‘The existence of vacancies in crystals was originally postulated to explain solid-state diffusion in crystals. Because of the mobility of gas atoms and molecules, gaseous diffusion is easy to comprehend, but the movement of atoms inside crystals is more difficult to grasp. Sti two metals, for example, nickel and copper, will diffuse into each other if placed.in intimate contact and heated to a high temperature. Several mechanisms have been proposed to explain these diffusion phenomena, but the most generally acceptable has been the vacancy mechanism. is a well known fact that 12.1 HISTORY Solid-state physics was born when Max von Laue (1879~1960) detected diffraction of X-rays on crystals, His experiments demonstrated that solid matter usually occurs in three-dimensional periodic Dr. Mamun Al Rashed, OMME, BUET : ‘Supplementary Hand Out on Point Defects and Related Phenomena, Pare 5.0129arrangements of atoms. His discovery, published in 1912 together with Friedrich and Knipping, was awarded with the 1924 Nobel Prize in physics. However, the ideal crystal of Max von Laue is a ‘dead’ crystal. Solid-state diffusion and many other properties require deviations from ideality. The Russian physicist Jakov ich Frenkel (1894-1952) was the first to introduce the concept of disorder in the field of solid-state physics, He in 1926 suggested that thermal agitation causes transitions of atoms from their regular lattice sites into interstitial Positions leaving behind lattice vacancies. This kind of disorder is now called Frenkel disorder and” consists of pairs of vacant lattice sites (vacancies) and lattice atoms on interstitial sites of the host crystal (self-interstitials). Only a few years later, Wagner and Schottky generalised the concept of disorder and treated disorder in binary AB compounds considering the occurrence of vacancies, self- interstitials and antisite defects both sublattices. Nowadays, itis common wisdom that atomic defects are necessary to mediate diffusion in crystals. The German physicist Walter Schottky (1886-1976) taughtat the universities of Rostock and Warzburg, Germany, and worked in the research laboratories of Siemens. He had a strong influence on the development of telecommunication. Among Schottky’ ‘many achievements a major one was the dévelopment of a theory for the rectifying behaviour of ‘metal-semiconductor contact, which revolutionised semiconductor technology. important for diffusion processes in crystalline solids. This statement mainly derives One is the ability of point defects to move through the crystal and to act as ‘vehicles for diffusion’ of atoms; another is their presence at thermal equilibrium. Of particular interest in this chapter are diffusion-relevant point defects, Le. defects which are present in appreciable thermal concentrations. Ina defect-free crystal, mass and charge density have the periodicity of the lattice. The creation of a point defect aistur this periodicity metals, the conduction electrons lead to an efficient electronic screening of defects. As a consequence, point defects in metals appear uncharged. In ionic crystals, the formation df a point defect, e.g., a vacancy in one sublattice disturbs the charge neutrality. Charge- preserving defect populations in ionic crystals include Frenkel disorder and Schottky disorder, both of which guarantee global charge neutrality. We already know that Frenkel disorder (more precisely, Frenkel defect) implies the formation of equal numbers of vacancies and self-interstitials in one sublattice, and Schottky disorder (more precisely, Schottky defect) consists of corresponding numbers of vacancies in the sublattices of cations and anions. For example, in AB compounds like NaCl ‘composed of cations and anions with equal charges opposite in sign the number of vacancies in both sublattices must be equal to preserve charge neutrality. Point defects in semiconductors introduce electronic energy levels within the band gap and thus can occur in neutral or ionised states, depending ‘on the position of the Fermi level. In what follows, we consider at first point defects in metals and then proceed to ionic crystals and semiconductors. 12.2 POINT DEFECTS Fig. 1 is a schematic illustration of common point defects in crystals. Atomic vacancies (a) and self- interstitial atoms (b, c) are elementary intrinsic defects, A pair of a vacancy and a self-interstitial atom is called a Frenkel defect. Foreign atoms (e, f) are extrinsic defects, and are referred to as impurity or solute atoms, depending on the context. A foreign atom may be present in a space between the host atoms, or ina place where a host atom is normally present. The former is an interstitial solute atom (e), and the latter is a substitutional solute atom (f). Complex defects may be formed by association Dr. Marnun AlRashed, DEAME, OUET . ‘Supplementary Hand Out on Point Defects and Related Phenomens cancels)i if solute atom (g, f these defects, such as pairs of vacancies (divacancies, d), pairs of a vacancy and a salute som i triplets of vacancies or larger clusters, and so on, because the association most free energy of the defect-containing crstal) 9909999909 OO-OOOMOOOO OOOOOO0O OSDOOO0HSS Figure 1 Point defects in an elemental crystal a: an atomic vacancy, ba sell-intersttial atom at an inter ‘a selfinterstitial atom in the split (or “dumb-bell") configuration, d: a pair of vacancies (divacancy), e: an interst foreign atom, f: a substitutional solute atom, ¢: pair ofa vacancy and an interstitial solute atom, ha pale ofa vacancy td subtiina coke atom, 1.23 VACANCIES AND SELF-INTERSTITIAL : : | 1.23.1 Vacancies 1 A vacancy is normally considered to be a'region of atomic dimensions characterised by the absence of an atom from a site which should be occupied, Clearly this will affect the mean positions of the other atoms in the crysta}, and the vacancy may be considered to be a negative centre of pressure with an elastic strain ri displacements produced will be very small, except for the atoms immediately adjacent to the vacant site, For these atoms, however, the changes in position may be so large that it {sno longer possible to specify unambiguously from which site an atom is missing, should properly be considered as an atomic region of bad crystal, within which ther Thus a vacancy Fe is one fewer ‘Dr. Mamun Al Rashed, OMME, BUET ‘Supplementary Hand Out on Polat Detects and Related Phenomena Page of29,atom than in a corresponding volume of good crystal. In the same way, a divacancy is an atomic region containing two fewer atoms than a corresponding volume of good crab may be formed by the amalgamation of two single vacancies, but its internal structure need not fesemble that of the single vacancy. These much-studied intrinsic defects exist in all classes of crystalline materials, As they are granted a probability of existence from an energy standpoint, vacancies cannot be eliminated from the lattice. fre presence of a vacancy can cause a local relaxation or redistribution of atoms and electrons irrounding it. This accounts for their importance in phenomena involving atomic motion or diffusion in solids. Lattice atoms or impurity atoms that are completely surrounded by nearest-neighbour lattice atoms are not mobile. But, if there is an adjacent vacancy, then the two can exchange places and atomic motion is posibledy If a pair of vacancies combines to make a single point defect, a divacancy is said to be created, It is difficult to estimate or measure the binding Energy of this type of deféct. One calculation, however, places it at about 0.3~0.4 eV in the case of copper. 1.2.3.2 Equilibrium Concentration of Vacancies Diffusion in crystals is explained, in terms of vacancies, by assuming that the vacancies move through the lattice, thereby producing random shifts of the atoms from one lattice position to another. Vacancy moves as a result of an atom jumping into a hole from a lattice position bordering the hole. In order to make the jump, the atom must overcome the net attractive force of its neighbours on the side opposite the hole. Work is therefore required to make the jump into the hole, or, as it may also be stated, an energy barrier must be overcome. Energy sufficient to overcome the barrier is furnished by the thermal or heat vibrations of the crystal lattice. The higher the temperature, the more intense the thermal vibrations, and the more frequently are the energy barriers overcome. Vacancy motion at high temperatures is very rapid and, as a consequence, the rate of diffusion increases rapidly with increasing temperature. An equation will now be derived that gives the equilibrium concentration of vacancies in a crystal as a function of temperature. The important characteristics of a point defect are its free energy of formation, and its free energy of © motion. The former governs the number of point defects present at equilibrium, and the latter the rate at which a point defect can move inside the structure. An individual defect will usually have a small energy of formation, and since the presence of a few defects gives a large increase in the configurational entropy of a crystal, real crystals at finite temperatures will always contain a small number of defects in equilibrium, with the lattice. Let us first consider isolated vacancies, which were once believed to originate at the free surface of a crystal, or at grain boundaries, when atoms leave their original sites to build up new layers. The unoccupied sites thus created were supposed to diffuse into the interior. It is now recognised that equivalent processes may occur within a crystal, since dislocation lines may sometimes act as sources or sinks for vacancies. Let us assume that in a crystal containing ng atoms there are ny vacant lattice sites. The total number of lattice sites is, accordingly, ng + ny, or the sum of the occupied and unoccupied positions. Suppose Dr. Mamun Al Rashed, DMIME, BUET ‘Supplementary Hand Out on Point Defects and Related Phenomena Page 80129that vacancies are created by movements of atoms from positions inside the crystal to positions on the surface.of the crystal. ‘When a vacancy has been formed in this manner, a Schottky defect is said to have been formed. Let the work required to form a Schottky defect be represented by the symbol w. A crystal containing ny vacancies will therefore have an internal energy greater than that of a crystal without vacancies by an amount nyw. The free energy of a crystal containing vacancies will be different from that of a crystal free of vacancies. This free-energy increment may be written as follows: G,=Hy,-TS, where G, is the free energy due to vacancies, Hy is the enthalpy increase due to the vacancies, and S, is the entropy due to the vacancies. . |. Sesaccording to the above, oa . Hy =nyw and thus Gy = nw —TSy, Now the entropy of the crystal is increased in the presence of vacancies for two reasons. First, the ‘atoms adjacent to éach hole are less restrained than those completely surrounded by other atoms and can, therefore, vibrate in a more irregular or random fashion than the atoms removed from the hole. Each vacancy contributes a small amount to the total entropy of the crystal. Secondly, entropy increases in the présence of vacancies due to entropy of mixing, Let us designate the vibrational entropy associated with one vacancy by the symbol s. The total increase in entropy arising from this source is nys, where ny is the total number of vacancies, While a Consideration of this vibrational entropy is important in a thorough theoretical treatment of vacancies, it will be omitted in our present calculations because its effect on the number of vacancies present in the crystal is of secondary importance. The other entropy form arising in the presence of vacancies is an entropy of mixing. The entropy of mixing has already been derived for the mixing of two ideal gases, and is expressed by the well-known equation Sin = AS = —nk[Xq In Xq + (1X4) In(1 — X4)] where Syq is the mixing entropy, n is the total number of atoms (rt, + ng), k is the Boltzmann's constant, X, is the concentration of atom A = "4/,, and (1 — Xq) is the concentration of atom B= o/, 7 The above equation applies tly to the present problem if we consider mixing of lattice points, instead of atoms, in which there are two types of lattice points: one occupied by atoms, the other unoccupied. If there are ny occupied sites and n, unoccupied,-the unmixed state will correspond to one in which a lattice-of ny + ny sites has all positions on one side filled and all on the other empty. Or. Mamun Al Rashed, OMME, BUET ‘Supplementary Hand Outon Point Defects and Related Phenomena Pagesot29‘Therefore, we may make the following substitutions in the mixing-entropy equation: tg ty Tae he ay : 12k =e ety where Xs the concentration af vacances, and Xe isthe concentration of occupied lattice positions. Ifthe above quantities ate set into the mising-entropy equation, we have fs Sm = —(e tne In +) which becomes after simplification . a Sm = Kl(e + My) In(ng + ny) — ny Tany — nln Ne) The free-energy equation for vacancies may now be WTS oF, Gy = ny ~ TL (Mg + ny) Ing + ny) — ny ln ny — nel Be) This free energy must be a minimum ifthe crystal is in equilibrium; that is, the number of vacancies (nu) in the crystal will seck the value that makes or, Gy 2 minimum at any given temperature. AS 2 ‘result, the derivative of G, with respect to ny must equal zero with the temperature being held constant, Serena t mds tle +4) — nyt —Inng~ 9] or omwsirn | which, when expressed In exponential form, becomes yt . othe tn general, it has been found that the number of vacancies in 2 metal crystal is very small when! ‘compared with the number of atoms (the number of occupied sites nq). Therefore, 1 . Meet i Ry where the work to form one vacancy in kis the Boltzmann's constant in /K, Tis the absolute ‘temperature ink, ny is the number of vacancies, and ng is the number of atoms. {both the numerator and the denominator ofthe exponent ofthe above equation are now multiplied bby N, Avogadro's number (6.023 x 10), the equation will be unaltered with respect to the ‘aman Al bed ME BEE Splementay Hod Out on Pt elec nd Related Phen ce30028of vacances andthe temperature, However, the fandard thermodynamical notation. Therefore, let coc cco betwen he teste cnet ithenorespontos y= hw R=kn ae wire Hy i the heat of activation; thats, the work required to form one mole of vacancies, in joules er mole, i the Avogadro's number, kis the Boltzmann's constant, and Ri the gas constant 8.32 joules per mole . Therefore, the equibrium concentration of vacances is 1 we Prk eet a, sre crperimental value forthe actvaton enthalpy for the formation of vacancies in copper i ~ SBerermotely 83,700 joules per mole of vacances This value may be subtituted into the eben Savatoninah attempt to determine the eect of tmperture on the number ef wearers, dette : 3000/87 «1000077, ‘shisha sera, the eulrum numberof vacancies shoul be er, fr inthis age a and at 300 k, approximately room temperature By es Fee e = 445 x 10-18 145 x 10" and t 1350, si degrees below meting point,‘Temperature (C)— 8 400 300, 192 3 4 ~ 5 6 x 10-4 \Vacanty concentration —> Figure 2Equlibrlum concentration of vacances a function of temperature fo aluminium ‘Significance of This Equation ‘Two questions need to be considered: first, why should there be an equilibrium number of vacancies ata given temperature, and, second, whydoes the equi rium number change with temperature. 24 seLewrenstima (Gotesital atm one tha cela ae ina cg tht normaly would be urezne. Such ‘laces are the holes or interstitial positions that occur between the atoms lying on the normal lattice sites. Its necessary to differentiate between two basically different types of atoms that can occupy these interstitial sites.-The first of these is composed of atoms whose sizes are small, such as carbon, nitrogen, hydrogen, and oxygen. Atoms of this type can sometimes occupy a small fraction of the Interstitial sites in a metal. When this occurs, itis said to be an interstitial solid solution (See Section 1.62). i ‘The other fype of nesta atom consists ofan atom that would normaly occupy a regular lattice ste, In the case of pure meal lke copper, such an interstitial atomn would be a copper atom) In FCC structures, for example, there are octahedral as well as tetrahedral interstit elther shor four age ltice atoms, respective, surround the smaller intrsil atom. The cence of the FCC unit cell (%, %, 24) is an example of the former, whereas the (%, %, %) site is an example of the later. The hardball model ofthe (10) plane ofan FC crystal in Fig. 3 clearly shows thatthe Interstates are too smal to hold sucha nrg atom without badly dstorting the lattice, As 8 consequence, while the activation energy to form vacancies in copper isalitle less thane, that for the formation ofintersitalcopper atoms is probably about 4 eV. This difeencein activation energy ‘means that inerstl atoms produced by thermal vbrations shouldbe very care in a metal such 2s copper. This an easly be shown by a computation using the same form of equation a that developed this chapter to compute the equllrium ati of vacancies o atoms. I sites where cages of While interstitial defects of the type under consideration do riot normally exist in sufficient concentrations in a metal to be sig cant, they can be cealy produced in metals a a result of ‘aation damage Colsions between fast neirons and 3 metal can result in knocking atoms out of ‘heir normal tice stes( Te removal ofan atom frm its lattice poston of course, produces both an interstal atom and 3 vacancy) has been estimated that each fst neutron can rel in the éreation of about 100-200 Iterstas and vacancies, ‘tama A Rashed, OnE, BUET ‘Supplementary Hand ton ont elec a Related PhenomenaInterstitial site Figure 3 The site ofthe interstitial site fe much smaller than the sie ofthe solvent atoms in metals the energy of formation of an interstitial atom is much higher than that for 2 vecanofond is of the order of 4 eVAAt temperatures just below the melting point, ‘the concentration of such point Getecsisony about 10™ pn theefreinestalsareofltle consequence inthe normal behaviour ‘of metals. They are, however, more important in ceramics because of tie more open crystal structure. “They ae aso of importandainthe deformation behaviour of ois when pont defects are produced by the non-conservative motion of jogs in screw dislocation and also of particylar importance in materials that have been subjected to ieraigtion by bigh-eneray particles Wie is ver cult to create an itestiia defect ofthe type presently being considered hey are very mobile once they are produced) The energy barre for ther motion Is smal and nthe ease of con Riasteenesino bef erro 6 Perso or ih town in Fig 4. Here, itis assumed that one is observing the (00} plane of an fe erytal. In ig, 4a the intertal atom is shown as a shaded circle and is designated by the symbol "suppor tht he atom were to move to the right. This motion would push atom ‘b’ into an interstitial position, as shown inst ths pour son scum nglotce ste Farber mtonel yy fame deco wou mat an itera tomo of ey ths ype ef maton, the congue associated with an interstitial atom can readily move thrSugh the lattice, Any given interstitial atom cry owsasnal eee adbeast drone seen rend ec eee stents maton ees itleeney The fe fete dtaton sound cee inoarthe ton ener ect equiarttothatbenedinte motonel decree é -S_ NON-EQUILIBRIUM PRODUCTION OF VACANCIES, cancies and selfntersttial atoms of concentrations far ‘produced by several method: leradiation with energe ‘AND SELF-INTERSTITIAL ATOMS above the equilibrium levels can be apid quenching from high temperatures, mechanical working, cles and Dy. Maman lashed one, UCT ‘Stplementar Hand Osten ont eles and Relate Phenomena JoseO © oO 0 O i | i Figure 4 The configuration when an interstitial atom can move readily through the crystal Cin the first case, the defects produced ata high temperature may be retained at the low temperature, where their migration is very slow. The defects are then “frozen-in” and remain supersaturated. This method is effective for vacancies, and was employed for basic studies of their properties. However, * application to self-interstitial’ atoms in metals is virtually impossible, since their equilibrium concentration is too low and, moreover, their mobility is too high to retain them’) a ir. Mamun Al Rashed, OMIME, BUET Supplementary Hand Out on Point Defects and Related Phenomena Page id of 29Row of . vacancies (a) . Row of . ve 7 interstitials (b) Figures Formation of pont detects by the aniilation of dislocations (3 Row of vacances (b] Row of interstitials. “tn the second ‘case, both vacancies and self-interstitial atoms are created by interaction between dislocations. For quantitative studies, however, this method is ndt convenient because controlling the amount ofthe defect is fg, and other defects, which ae Inetably produced topethen complicate the experiment Heviwr rng plastic deformation, the movement of dseesony fenerats point defects tno mechaianstenan-cansevative motionct fogs andthe amnion of pale locations of oposite sign producing ane ofvacandes or niet ees we eented by sles intersections sine they canoot de wth dsleatons, they have te dink owe sisocaon moves na screw destin they are smal segments having te chores afar cn Thestp plane ofthis segment int compatible wth that the dloation, Te nts poole ey br continous emisione cance: or ntertss. The second mechani edepted coreeen, in Fig. S. When the two dislocations cancel each other, they create a row of interstitials or vacancies itis planes dona coincide, The thee, gh eneryiradato, an eectie means toinrodue Frenkel defects Quenehing broduces moity vacancies and vacancy groups. The concentrations obtained are lewes thar meek Deformation onthe other hand can introduce higher concentrations of vacancies ond enenionge once finer th problems that tal introduces a number father substacul cheney Sie coments the sean Oiocations ae introduced, and they interact stony wk meg sefes One method of producing point defect doesnot preset these problems nadaven cq ‘etl by heer particles alow the invoducton of highconcentrsten of defects, When sald materials radiated with electrons, ions, or neutrons, collisions of incident particle with toms nthe solid occur, The target atom may be clplaced out of the regular site, leaving a vacancy behing, ifthe energy transferred from the incident partice i sufficiently high. The critical value for the “knock-on” event to take place Is a few ten ele *ctronwvolts in most metals, Fig. 6 shows the ‘*Perimental data of the displacement threshold ene, "BY, &, of common metals, St Maman lashed ne, ET “pPemertayHand Ovni eects ad Rated Phenomena rage sseta9When the kinetic energy ofthe displaced atom, called 2 primary knackon atom (PKA},is high, it may arrae an incident particle and may cause further knack-ons.Achain of multiple atomic displacements iagered by a bigh energy PKA s called a displacement cascade, How many defects are produced by apa depends on its Kinetic energy. cording to the simple model for displacement of atoms by inchin and Pease (2955), above Ethe average numberof dsplacementsis proportional tothe enerey PKA Taking nickel as an example, whose eisplacement thresholds 23 eV, with 1 MeV electrons the Fravimum enerey transfer is about 70 eV but the average Kinetic energy of a PKA is of similar ‘magnitude to a(n this partiylar cace itis about 4 eV), Therefore, the defects introduced by electrons ofthis energy are mostly closely spaced sel-nterstital atoms and vacancies, Thisis the technique that fas widely been used in experimental studies of Frenkel defects in metals. On the other hand, fast neutrons of the same energy may transfer up to 66 keV to a target atom in the same material. The verage PKA energy is about hal of the maximum energy transfer, and the numberof displacements ‘expected for ths case from the Kinchin-Pease model is over 700. 50 aol. EgeV 1 KTqfeV gure 6 Deplacement threshold nergy of pure metal pleted agnnst Kn where Ta the meting temperature, ‘This can be understood as follows. We know that there are many different kinds of high-energy radiation (e.g. neutrons, electrons, a particles, protons, deuterons, uranium fission fragments, Vays, X-rays) and all of them are capable ff producing some form of ‘radiation damage’ in the materials they rradiate, While all are of importance tosome aspects ofthe solid state, of particular interests the behaviour of materials under radiation ina nuclear reactor. This fs because the neutrons produced ina reactor by fission reaction. have extremely high energies of about 2 milion electron volts (i.e. 2 MeV}, and being electrically uncharged, and consequently unaffected by the electrical fields surrounding an atomic nucleus, can Mum Raha OnE, ET “orlemanta Hand ut on Ft Delos nd Raed Piraens page 360129ructure. The resultant damage is therefore not localised but is ances through 2 travel large distances through the form of ‘damage spikes’ (or, displacement cascade). distributed throughout the solid in ‘The fast neutrons! (they are given this name because 2 MeV corresponds to 2 ce oan 4) are slowed down, in order to produce further fission, by the moderator in the pile until they ; thermal equilibrium with their surroundings. The neutrons ina pile wil therefore, have a spectrum o energies which ranges from about 1/40 eV at room temperature (thermal neutrons) to 2 MeV (Fast neutrons). However, when non-fssle material is placed ina reactor and irradiated most of the damage is caused by the fast neutrons coliding with the atomic nuclei of the material. ‘The nucleus of an atom has.a small diameter (e.g, 10° m), and consequently the largest area, or cross section, which it presents to the neutron for collision is also small. The unit of cross section is a barn, ite. 10° m so that in a material witha cross section of 2 barn, an average of 10? neutrons would have ‘to pass through an atom (cross-sectional area 10* m?) for one to hit the nucleus. Conversely, the ‘mean ree'path between collisions s about 10? atom spacings or about 0.3 m. Ifa petal such as copper {cross section, 4 barns) were irradiated for 1 day (20° ) in a neutron flux of 10%" ms* the number of neutrons passing though, unit area, ie. the integrated flux, would be 10” n m® and the chance of a tiven atom being hit = integrated flux x cross section) would be 4 x 10%, i, about 1 atom in 250 000 would have its nucleus struck. For most metals the colision between an atomic nucleus and a neutron (or other fast particle of mass ’m) is usually purely elastic, and the struck atom mass M will have equal probability of receiving any Kinetic eneray between zero and the masimum Emo = *8=M™/ | », where Ey ithe enerey ‘of the fast neutron, Thus, the most energetic neutrons can impart an energy of as much as 200 000 €V, to.a copper atom initially at ragt, Such an atom, called @ primary ‘knock-on atom (PKA, will do ‘much further damage on its subsequent passage through the structure often producing secondary and tertiary knock-on atoms, so that severe local,damage results. The neutron, of course, also ‘continues its passage through the structure producing further primary displacements until the energy ‘transferred in collisions i less than the energy £, (= 25 eV for copper) necessary to displace an atom from its lattice site. ‘The damage produced in irradiation® consists largely of interstitials ie. atoms knocked into interstitial positions in the lattice, and vacancies, ie. the holes they leave behind, The damaged region, estimated Stow neutrons give rise to transmutation products. Of particular importance isthe production ofthe noble gs hebrea 3607 aw otla- 19 PM Az Gopm N ~—pPOpD Feet Ocal fe nd eahdel esr te nd er ees Strnad ou Serene ttt bed ae ane | | f | | |12:7 powr perecrs im attoys A pure metal consisting of . Fuca “ “ 4 f only one type of atom just isn’t possible; impurity or foreign atoms will : reer and some wl ext 9 enaline nth acees seo er les fc, ven wth reve ‘costed techies, is liao eine mets uty n exes 09900 Hen on therderot 010 10mpurtyatoms wb present mom eaemetesfoseneR een oe : re Is are not highly pure; rather, they are alloys, in which Impurity atoms have been added : into to impr sec haces tothe malay say ond ss imorove mechani sere ad corecn reine) rel seg ser SE sive scopper alo fanomalanbiertenronmert reser bh conn esas 4 0 ver st Alon with coger Seneanty enone te metals ane enreciting the corrosion resitance appreciably The addon otinpury some toamealleeh inthe frmaton of soe soliton andor anewseond pase, ependng one tins of anean shi cncersatorsanhetempeue a he) 4 ‘Several terms relating to impurities and solid solutions deserve mention. With regard. alloys, solute ‘and solvent are terms that are commonly employed. Solvent represents the element or compound that Is present inthe greatest amount; on occasion, solvent atoms are also called host atoms. Solute is used to denote an element or compound present in a minor éancentration, ‘A solid solution forms when, as the solute atoms are added tothe host material, the enstal structure 's maintained and no new structures are formed, Perhaps is useful to draw an analogy with a Vivid solution. if two liquids, soluble in each other (such as water and alchel) are combined, a liquid solution is produced as the molecules intermix, and its composition is homogencous thrSushout. A sold solution is also compositionally homogeneous; the impurity atoms are randomly and uniformly lspersed within the sli 7.4. Interstitial Defect Can Interstitial defect I formed when an extra atom orion f inserted into the erstal structure at 2 ‘normally unoccupied postion, as shown in Fig. 9. Interstitial atoms or ions, although much smaller ‘than the atoms or ions located a the atic points, re sil arger than the interstitial sits that they ‘occupy consequently, the surrounding crystal region i compressed and distorted. Intersil atoms such as hydrogen are often present as impurities, whereas carbon atoms are intentionally added to Iron to produce stecl) For small concentrations, earbon atoms occupy interstitial sites inthe ron + ‘erystal structure, introducing a ses in the localized region ofthe crystal in their icity. AS we will {roduction of interstitial atoms is one Important way of increasing the strength of metalic see, th materials Unlike vacancies, once introduced, the number of interstitial atoms or lns inthe structure ‘remains nearly constant, even when the temperatures changed. we relatively high atomic packing factors, these ntrstial postions are the atomic diameter ofan interstitial impurity must be substantially (Croce et relatively small. Consequently, smaller than that of the host atoms. Normal, th ns is tow (less than 10%. Even very smal impurity atoms are ordinary larger than the ‘duce somelatticstraison the adjacent ost atom) ofearbon Impurity ator ingersttialsites, and as consequence they int smiessiasaatin when sea ath minum cocton te stom is much less than that for iron: 0.071 nm versus carbon forms: ts about 2%. The atomic radius ofthe carbon a , o.224nm. rae : Pam eaten Panto and Rts Panis : .&S{Substitutional] impurity.a Interstitial impurity atom Figure 9 Two-dimensional schematic representations of substitutional and inte impurity atoms. 1.2.7.2 Substitutional Defect A substitutional defect is introduced when one atom or ion is replaced by a different type of atom or ion, as shown in Fig. The substitutional atoms or ions occupy the normal latice site. Substitutional atoms or ions may either be larger than the normal atoms or ions in the crystal structure, in which case the surrounding interatomic spatings are reduced, or smaller causing the surrounding atoms to have larger interatomic spacings. In either case, the substitutional defects disturb the surrounding crystal. Again, the substitutional defect can be introduced either as an impurity or as a deliberate alloying addition, and, once introduced, the number of defects is relatively independent of temperature. (several features of the solute and solvent atoms determine the degree to which the former dissolves in the latter, as follows. This is known-as Hume-Rothery Law. ——— 1. Atomic size factor. Appreciable quantities of a solute may be accommodated in this type of solid solution only when the difference in atomic radii between the two atom types is less than about +15%, Otherwise the solute atoms will create substantial lattice distortions and a new phase will form. Crystal structure. For appreciable solid solubility the crystal structures for metals of both atom types must be the same. 3. Electronegativity. The more electropositive one element and the more electronegative the other, the greater the likelihood that they will form an intermetallic compound instead of a substitutional solid solution. 4. Valences. Other factors being equal, a metal will haye more of a tendency to dissolve another metal of higher valency than one of a lower valency. ‘De. Mamun Al Rashed, OMME, BUET 7 Supplementary Hand Out on Point Defects and Related Phenomena Page 22.0129Examples of substitutional defects include incorporation of dopants such as phosphorus (Por boron (6) lato Si, siilary, if we add copper to gicke, copper atoms wil occupy crystallographic sites where nickel atoms would normally be Pest ote ‘example ofa substitutional slid solution is found for copper and nickel. These two elements are completely soluble in one another at all proportions. With regard tothe aforéientioned rules that govern degree of solubility, the atomic radi for copper and nickel are 0.128 and 0.425 nm, respectively; both have the FCC crystal structure; and thelr electronegativties are 1.9 and 1.8; finally, the most common valences are +1 for copper although it sometimes can be #2) and-42 for nickel : \ 12.8 powr perecrsin oxic comPounns Ina ads and metal exes where poiiv ard native lone populate late ste, pon defect structures re ore campo Mntnance of huge netsh reso, Te vale the ses Involved tu oneider Fi 108 which depicts an electra aula eamposed of monovalent (positive) cations and (negative) anions. Creation of a cation vacancy means the absence of a single positive charge or an effective negative lattice charge. This defect cannot exist by itself because the Ite no longer elec esta Therefore ana on vary i ereatdatthe sae tine an the ssn piste ion vacancy negate fon veo s naw a2 Shot dete, This defect unique to ole mater and comment Youn a many geamie mates. The ‘mechanism involved: when vacancles occur in an ionically bonded material, a stoichvornetTe number Of aon nd enlons mus be ising om repr ston posto f eee neat is tbe preereFor example one vacancy none O* vacancy in MgO const Shot pain 10, torene te" wscany thereat 0O* vacances 2283 c0Q @ BSE PS Figure 10 Point defects in onic rsa (2 Schothy defect and (0) Frenkel defect (Goeiercec hia 8 std xen hp a er Tre area tarentrtelatmttnawm a3 Frentl fe. Thelen Aa) Alloying alkali halides can complicate the defect picture. For example, if CaCh is added to NaCI the two ‘Ch-ions occupy two normal anion sites, but the single Ca" ion can occupy only one cation site (Fig. 103). To compensate forts extra charge an additional cation vacancy must be created. In such a case we must inguish between two types of cation vacancies: those pfoduced intrinsically simply by heating and those introduced extrinsically by alloying. Either way these defects play animportant role in both material and charge transport processes in all types of ionic solids. And as we shal later see, semiconductor doping has features in common with alloying onic compounds x Maun Ashe, OND, ET “Sippementar Hand Goton ot Defects and Rete Phenomens age 0128 tinea eation of An important substitutional point defect occurs when an ion of one charge replaces an ofa different charge. This might be the case when anion with a valence of +2 replaces an ion with a valence f I i tucture. To m: i i xtra positive charge is introduced into the structure f +1 (Fig. 11). In this case, an extra posi rain 3 eee fer vacancy might be created where a +1 cation normally would be located. Again, this imperfection is observed in materials that have pronounced ionic bonding. eo. \ \) Q : BR 82888 ¥~ O8O8OQ C6OSO Figure 11 When a divalent cation replace: “ae tain a ‘monovalent cation, a second monovalent cation must also be removed, ‘creating a vacancy. . inte solids, when point defects are intraduced, the following rules have to be observed: (2) a charge balance must be maintaine electrically aD (bh a mass balance mustbe maintained; and + be conserve d so that the crystalline material as a whole is ) the number of crystallographic sites must For example, in nickel oxide (NiO) if one oxygen ion is mi sing, (designated as V). Each dot ()in the superscript, position indicates: To maintain stoichiometry, mass balance, (designated as Vy), Each accent (') in the (em VACANCY MOTION Vi it creates an oxygen ion vacancy ‘an effective positive charge of one, and charge balance, we must also create nickelion vacancy Superscript indicates an effective charge of -13, /acancy movement is shown in Fig. 12. We have determined the equilibrium ratio of vacancies to atoms ata given temperature T, Nothing, however, has been sald about the time required to attain an equlibrurn number of veg This imay be very long at low temperatures or very rapid at high temperatures, Since the movement of Necancies is caused by successive jumps of atoms into vacancies, a study of the fundamental ag overning the jumps is important. We know that an energy barrier, shown schematically in Fig. 13, ‘must be overcome when a jump is made, and that the required energy supplied by the thermal, or heat, vibrations of the crystal. —— ee {We use the Krdger-Vink notation to write the defect chemistry equations. The main letter in this notation eteribes 2 vacancy or the nome of the element. The superscript indicates the effecie, charge on the defect, and the subscript desribes the location of the defect, A dot ( indicates an effective charge of +1 and an accent C)represents an effective charge of -1, De. Marmun A Rashed, OMME,BUET ‘Supplementary Hand Out on Point Defects and Related Phenomena Page a2 of 21 2 2 4 2000000 2000000 = 0000000 = ga00000S* 990000~ 900009 a000f00_ 6900000 99609000 90000000 90000000 6000-000 9000000 — 9900000 = 9000000" 0000000. 000000 — 0000000 = 9000000 += 0000000 . 8 7 6 5 9200000 9000000 = 9900000 += 0000000 9900000 9000000 0000000 = d000000 9900000 9=000000 9000000 000+0000 9000000 0000000" 6000000 + G000000 0000000 = 0000000 = 6000600 = 6000000 - Flere 12 Movementofvscnler '¥ a is the helght ofthe energy barrier hat an atom, such as atom ain Fig. 10, must overcome in order ta jump into vacancy, then the ump can any occur if atom a possesses vibrational energy greater ‘than do. Conversely ifthe vibrational energy slow than gg, the jump cannot occur, The chance that me ven atom possesses an energy greater than qo has been found to be proportional to the function eH or p= constant et where pis the probability that an atom possesses an energy equal toot greater than a given enerey Ge, kis Boltzmann's constant, and Tis the absolute temperature, This equation was originally derived for te enerey distribution of atoms ina perfect gas (Maxwellotzmann distribution). However, this same function has alo been found to predict quite accurately the vibrational energy distribution of "he atoms in a erytalline soli 5 Snce the abovefunctons the probability that a given atom has an eneray greater than that required fora jump, the probebity of lumping must be proportional to this funtch, Therefore, we con wre the folowing equation ee where ris the number of atom jumps per second into a vacancy, Ais 3 constant, qals the activation ‘neray per atom {height of the ener barrie), and k snd T have the usual slanifcenee, the ‘numerator and denominator of the exponent in the above equation are both multiplied by N, ‘Avogadro's number 6.02 x10, we have ee ere Hm isthe activation enthalpy for the movement of vacances in joules per male, and R the gaz ‘constant (8.31 per moe K). 5 2 Maman ae OE BET ‘Semen Ot oh ant Ele deed Penna gerecsasFigure 13 The energy barrier that an atom must overcome in order to jump into a vacancy The constant A depends on a number of factors. Among these is the number of atoms bordering the hole. The larger the number of atoms able to jump, the greater is the frequency of jumping. A second factor Is the vibration frequency of the atoms. The higher the rate of vibration, the more times per second an atom approaches the hole, and the greateris ts chance of making ajump. « Let us consider the above situation with respect to an actual metal, The value of A for copper is approximately 10", and the activation enthalpy Hm = 121 Ki per mole. If these values are substituted into the jump-rate equation, we have at 1350 K (just below melting point of copper) : ty = 2.x 10?° jumps/s and at 300 K (room temperature) 1, = 106 Jumps/s. ‘The tremendous difference in the rate with which vacancies move near the melting point, compared with their rate at room temperature, is quite apparent. In one second at 1350 K the vacancy moves approximately 30 billion times, while at room temperature a time interval of about 10° sec, or 11 days, occurs between jumps) {3 STACKING FAULTS (Paras aetect) Perfect crystals can be described*as a stack of atom layers arranged in a regular sequence. For the simple metallic structures, the atomic layers are identical, A stacking fault is a planar defect and, as its name implies, it is a local region in the crystal where the regular stacking sequence is interrupted, Stacking faults are not expected in planes with ABABAB...sequences in body-centred or face-centred cubic metals because there is no alternative site for an A layer resting on a 8 layer. However, for ABCABC.., Or ABABAB... stacking of the close-packed planes in close-packed structures there are two Dr. Mamun Al Rashed, DMME, BUET ‘Supplementary Hand Out on Point Detects and Related Phenomena Page25 0f29|Papas Pastlons of one layer resting on another. According tothe har Love of tomsresting onan layer canrestequalywelinetnar sn ‘here's no reaton forthe selection ofa pa sheremotel dese stad 3 potion neem lor ostion. ma facecentedeubatert any stacking fault are péssible, referred to as intrinsic and extrinsic, " These are best described by considering the change in sequence resulting from the removal or , the atcha Of 28 era layer. In Fig. 14a part of aC layer has been removed reach breakin the stacking sequence, This an intense faut andit con be seen tee stacking sequences above and below the faut plane are continuous right up o the faut set, ‘The normal stacking sequence of 212 planes ie denoted by parted by those witha stacking ero bya Ini sath faut (8) enrnetatig ful In Fig. 4b an extra layer has been introduced bétween a Band aC layer. There are two breaks inthe ‘tacking Sequence andi referred to at an extrinsic au, The entra layer does not belong to the \ Continuing patterns of the atic either above or below the faut, 1.4 PLANAR DEFECTS Zhe several different types of interfacial or planar imperfections, in solids canbe grouped into the following categories: Interfaces between solids and gases, which ae called free surtaces Interfaces between regions where there is a change in the electron structure, but no change inthe perlodety of atom arrangement, known as domain boundaries; interfaces between two crystals or grains ofthe same phase where there i an orientation diference in the atom arrangement across the interface; these interfaces are called grain boundaries; interfaces between different phases, called phase boundaries, where theres generally a change of chemical composition and atom arrangement agross the interface. . . . 14.1 FREE SURFACES + Because of their finite size, all solid materials have free surfaces. The arrangement of atoms ata free Surface differs slightly from the interior structure because the surface atoms donot have neighbouring tors on one side. Usualy the atoms near the surface have the same crystal structure buta slightly larger latice parameter than the interior atoms. Pechaps the most important aspect of free surfaces isthe surface energy (y) associated with surfaces * of any soli. The source ofthis surface energy may be seen by considering the surroundings of atoms fon the surface and in the interior ofa solid. To bring an atom from the interior to the surface, we must Sean Ae OO cat regattaeither break or distort some bonds - thereby increasing the energy, The surface energy is defined as the increase in energy per unit area of new surface formed. In crystalline solids, the surface energy depends on the crystallographic orientation of the surface - those surfaces that are planes of densest atomic packing are also the planes of lowest surface cnergy. This is because atoms on these surfaces have fewer of their bonds broken or, equivalently, have a larger number of nearest neighbours within {thé plane of the surface. Typical values of surface energies of solids range from about 10" to 1 J/m?. Génerally, the stronger the bonding in the crystal, the higher is the surface cnerey. Surface energies can be reduced by the adsorption of foreign atoms cr molecules from the Surrounding atmosphere. For example, in mica the surface energy of freshly cleaved material in a ‘vacuum is much higher than the surface energy of the same surface cleaved in air, In this instance, Snygen Is adsorbed from the air to partially satisfy the broken bonds at the surface. Impurity atom adsorption makes it almost impossible to maintain atomically clean surfaces. As a result, surface Properties such as electron emission, rates of evaporation and ratés of chemical reactions are extremely dependent on the presence of any adsorbed impurities. These profertie: will be different ifthe measurergepts are made under concitions giving different surface adsorption, 1.4.2 GRAIN BOUNDARIES Grain boundaries separate regions of different crystallographic orientation, simplest form of a grain boundary is an interface composed of a parallel ary ‘Ts particular type of boundary is called att boundary because the misorientation isin the orcs or 2 simple tilt about an axis, parallel to the dislocations, boundaries because the angle of misorientation is gener a misorientation greater than 10° or 15" made up of dislocations becau lose their individual as scen in Fig. 15. The ay of edge dislocations. Tilt boundaries are referred to as low-angle ally less than 10", When a grain boundary has it is no longer practical to think of the boundary as being sg the spacing of the dislocations would be so small that they would identity. The'grain boundary represents a region a few atomic diameters wide where there is a transition in atomic periodicity between adjacent crystals or grains. Grain boundaries have an interfacial energy because of the disruption in atomic periodicity in the Vicinity of the boundary and the broken bonds that exist across the interface. The interfacial energy of grain boundaries is generally ess than that ofa free surface because the atoms ina grain boundary are surrounded on all sides by other atoms and have ‘only a few broken or distorted bonds. Solids with grain boundaries are referred to as polycrystalline, since the structure is. ‘composed of many crystals - each with a different crystallographic orientation. In the case of iron the grain boundary. structure can be revealed by preferential chemical attack (etching) at the grain boundaries, while the grain structure in polyethylene is revealed by the use of polarised light. * In polycrystalline samples the individual grains usually have a random crystallographic orientation with respect to one another, and the grain structure is referred to as randomly oriented. In some instances, however, the grains all have the same orientation to within afew degrees. n thi is said to have a preferred orientation or texture. 14.3 PHASE BOUNDARIES A phase is defined as a homogencous, physically distinct and mechanically separable portion of the material with a given chemical composition and structure, Phases may be substitutional or interstitial 1. Mamnun Al Rashed, OMME, BUET page 270129 ‘Supplementary Hand Out on Point Defects and Related Phenomena 8solid solutions, ordered alloys or compounds, amorphous substances or even pure elements; a crystalline phase in the solid state may be either polycrystalline or exist as a single crystal. Grain Boundary Figure 15 Grain boundary Solids composed of more than one element may - and often do - consist of a number of phases. For example, a dentist's drill, something painfully familiar to all of us, consists of a mixture of small single crystals of tungsten carbidé surrounded by a matrix of cobalt. Here the cobalt forms a continuous Phase. Polyphase materials such as the dentist's drill are generally referred to as composite materials. Composite materials have great importance in the engineering world because they have many attractive properties that set them apart from single-phase materials. Fer example, the dentist's drill has good abrasive characteristics (due to the hard carbide particles) and good toughness and impact resistance (due to the continuous cobalt matrix). Neither the tungsten carbide nor the cobalt has both abrasion resistance and impact resistance, yet the proper combination of the two phases yields a composite structure with the desired properties. ‘The nature of the interface separating various phases is very much like a grain boundary foundries between two phases of different chemical composition and different crystal structure at grain boundaries, while boundaries between different phases similar crystal structures and * ‘erystallographic orientations may be analogous to low-angle grain boundaries)in both energy and structure. The concept of a solid consisting of a continuous phase and a dis¢ontinuous phase (or phases) leads to a simple classification of the various types of composite materials. “1,5 VOLUME DEFECTS Crystal defects such as precipitates, voids and bubbles can accur under certain circumstances and have important effects on the properties of crystalline solids) As an example, it will be seen later how the interaction of dislocations with precipitates has played a vital role in the development of high- strength alloys. similar to ir. Mamun Al Rathed, OMME, BUET Supplementary Hand Out on Point Detects and Related Phenomena Ps "age 28 0129REFERENCES 1. DRAskeland, PP Fulay, Wl Wright, THE SCIENCE AND ENGINEERING OF MATERIALS, 6" Edition, Cengage Learning, 2011 D Hull, DJ Bacon, INTRODUCTION TO DISLOCATIONS, S“ Edition, Elsevier, 2011 WO Callister Jr, OG Rethwisch, MATERIALS SCIENCE AND ENGINEERING: AN INTRODUCTION, 8 Edition, Wiley, 2010 R Abbaschian, L Abbaschian and RE Reed-Hill, PHYSICAL METALLURGY PRINCIPLES, Cengage Learning, 4" Edition, 2009 O Laughlin and K Hono (Editors RW Cahn and P Haasen), PHYSICAL METALLURGY, 5" Edition, Elsevier, 2014 RE Smallman and AHW Ngan, MODERN PHYSICAL METALLURGY, 8* Edition, Butterworth- Heinemann, 2014 ah “SH 2... Christian, THE THEORY OF TRANSFORMATIONS IN METALS AND ALLOYS, 3 Edition, Elsevier, 2002 F. Mamun Al Rashed, OMME, BUET Supplementary Hand Out on Point Detects and Related Phenomena Page29.of29