The Crystal Structure

1.1 INTRODUCTION

Intermolecular attraction is minimum in the gaseous state and this disappears completely when the gas is ideal. The interaction is stronger in liquids and is strongest in solids. Thermal motion of the molecules increases or decreases by raising or lowering of temperature. The attractive interaction between the molecules tries to keep them together and the thermal motion is opposed to that. Hence, it is possible to change a substance from one state to another by changing its temperature. If a liquid is allowed to cool slowly, the molecules will arrange themselves in an orderly manner and this will finally result in a crystalline solid. If, on the other hand, cooling is rapid, the molecules will not be able to arrange themselves in order. Rapid densification will give a glass or an amorphous solid. It is not true that the molecules and atoms in a solid have rigidly fixed coordinates. But they move only a small distance about their equilibrium positions. In this book, we are concerned with crystalline solids and the word solid and crystalline solid will be used synonymously. What is the stable state of a given material will depend on its free energy. The stable state will be the one that has the lowest free energy under the given conditions. Free energy A is related to internal energy U and entropy S of the system as A = U ST ...(1.1)

Internal energy is lowered by an orderly arrangement of the atoms, molecules or ions as that will lead to maximum energy of interaction. But this will minimise the entropy. Since internal energy and entropy make opposite contribution to free energy, the state of matter will be determined by the relative contributions of U and ST to A. If interaction is strong, U is highly negative and ST can overcome the contribution of the former only at high temperatures. Such a substance will remain as a solid even at a relatively high temperature. The basic feature of a crystalline solid is the regular arrangement of the atoms and molecules. At the macro level, this translates into crystals having sharp boundaries with clear cut shapes. It is these beautiful shapes of natural crystals that attracted human attention for ages. This beautiful shape and colour added to their value as gems. Early study of crystals began with the observation of their shapes and this is known as Geometric Crystallography. The description of crystal symmetry in terms of point lattice began in the mid-nineteenth century. This was followed by X-ray crystal structure determination following the work of Laue and Bragg on X-ray diffraction by crystals. In the second quarter of the twentieth century, the presence of lattice defects and their role in determining the properties of crystals were recognised. We shall not try to follow the development of the subject in a chronological order as the development of knowledge in an area of science does not take place in the same

Solid State Chemistry

logical way as one would like to see it. But only after enough knowledge gets accumulated that a subject is put in a logical perspective. Here we shall follow the rational rather than the chronological course of development of the subject.

Crystal lattice
It is easy to imagine a crystal as a periodic arrangement of points as shown in figure 1.1. A point may represent an atom or a group atoms arranged around it in a real crystal. Let us begin with a single point. Repeated translation of this point through a fixed distance = (a periodic translation) will generate a linear array of points. This movement is denoted ^ by a translation vector t1. If we add a second translation t2, it will generate periodically repeating points on a plane and this is known as a plane lattice. If a third translation t3 is added, we get a three dimensional arrangement of the points that is called Figure 1.1: A two dimensional plane lattice a space lattice. The lattice points are imaginary. In a real crystal, they are occupied by atoms or groups of atoms that are arranged in a regular fashion about the lattice points. This atom or the group of atoms is the basis and the arrangement of the imaginary points is the lattice. The real crystal is then: basis + lattice = crystal. A two dimensional pattern as is usually found on a curtain cloth or a wall paper is analogous to a two dimensional crystal lattice. We can have an array of squarely arranged points (Figure 1.2a) or the points may be arranged along inclined lines (Figure 1.2b). We can select a single motif and place this motif in the same way about each lattice points. This will give two different patterns (Figure. 1.2c and 1.2d). By selecting a different motif, we may get a still different pattern and a large number of patterns can be generated from a limited number of motifs and lattice arrangements.
^ ^

Unit cell
We have seen that two noncollinear translations give rise to a plane lattice and introducing a third translation (not on the same plane) generates a space lattice. Since any line joining two lattice points is a translation and there can be wide choice of translation, the question arises as to which two translation should one select to describe a plane lattice. A few such combinations are shown in figure 1.3. It is seen that they generate two dimensional units called unit cell. Combination of ^ ^ ^ ^ t 1, t 2 or t 3, t 4 leads to cells having only one lattice point per cell. These are known as primitive ^ ^ unit cell. The combination t 5, t 6 generates a double cell. There can be many more multiple cells. The unit cell of a lattice can be primitive or multiple. A repetition of the two dimensional unit cell by translation in two directions generates the plane lattice. This may be extended to three dimensional lattice that may be generated by translation of a three dimensional unit cell.

The Crystal Structure

(a)

(c)

(b)

(d)

Figure 1.2: Two different plain lattices with identical motif leading to two different patterns

t3 t4 t6

t2 t1

t5

Figure 1.3: Different combinations of translation and the unit cell

The three translations are selected along the three edges of the chosen unit cell. The choice of a unit cell for real crystal is done based on the convenience and conventions. It can be a primitive cell or a multiple cell. The three selected axes along the edges of the unit cell are called the crystallographic axes a, b and c and the angles between them , and as shown in figure 1.4. As has already been said, the early years of study of solids were concerned mostly with crystal geometry. This was followed by the study of crystal symmetry. Bravais in 1848 had shown that all structures can be generated by using only 14 types of space lattice (14 types of unit cell). These 14 types of unit cells will give rise to 230 types of lattice structures (space group) by performing such symmetry operations as

g
a

Figure 1.4: A unit cell showing the edges and the angles

Solid State Chemistry

(i) translation, (ii) rotation, (iii) translation + rotation and (iv) reflection. It was shown by Federov and independently by Barlow that it is not possible to have arrangement of lattice points other than than these 230 types that can repeat itself infinitely in three dimensions. The 14 types of Bravais lattices are shown in figure 1.5.

sc

bcc

fcc

hexagonal

Simple

bcc orthorhombic

tetragonal

rhombohedral

monoclinic

triclinic

Figure 1.5: Fourteen types of Bravais lattice

All real crystals belong to anyone of the 230 space groups. The imaginary lattice points are occupied by atoms or groups of atoms. Table 1.1 shows the classification and the geometric properties of the various Bravais lattices. Table 1.1: Classification of the fourteen Bravais lattices
System Cubic Space Lattice simple cubic body centered cubic face centered cubic hexagonal prism Condition a = b = c = = = 90 a = b c = = 90; = 120 a = b c = = = 90 (Contd.)

Hexagonal

Tetragonal

tetragonal prism tetragonal bcc prism

The Crystal Structure

Orthorhombic

rectangular prism bc rectangular prism fc rectangular prism base centered prism rhombohedron parallelopiped bc parallelopiped triclinic parallelopiped

a b c = = = 90

Rhombohedral Monoclinic

a = b = c = = 90 a b c = = 90; 90 a b c

Triclinic

1.2

CRYSTAL PLANES AND MILLER INDICES

In a real crystal, the lattice points are occupied by atoms. One can think of a very large number of lattice planes and one set of parallel planes can be distinguished from another set by their orientation. Miller indices are the labels used to distinguish one set of parallel planes from another. It is a set of three numbers h k l that defines a set of parallel planes in a crystal. The following procedure is generally followed to determine the Miller indices. 1. Choose an origin; 2. Find out the intercept that the first such plane makes with the three crystallographic axes; 3. Obtain their reciprocals; 4. Eliminate fractions. The set of numbers thus generated in relation with the axes a, b and c are h, k and l respectively. A set of parallel planes are labelled by a set of hkl numbers. Different set of parallel planes have different hkl or Miller indices. This is illustrated in figure 1.6. Here, the plane nearest 1 1 1 , the a axis at and the c axis at is shown. This to the origin and cutting the b axis at 2b 3a 2c 1 1 1 makes intercept along a axis, intercept along b axis and along c axis. 3 2 2
c

c b

b
a

Figure 1.6: (322) set of planes

Solid State Chemistry

We can write intercept reciprocal

a 1/3 3

b 1/2 2

c 1/2 2

Hence this plane and a set of parallel planes separated by a distance d have the Miller indices (322). It should be noted that a point on a paper actually represents line of points when one considers the three dimensional lattice. Hence on the plane of paper a line of points actually represents a plane. The major advantage of the Miller indices is that it permits to express interplanar distance dhkl of a set of hkl planes in terms of lattice parameters a, b, c, , and . For a cubic crystal a dhkl = 2 h + k2 + l2

1.3

DIFFRACTION OF X-RAYS

In 1912, von Laue first suggested that since the lattice points in a real crystal are occupied by atoms, the crystal lattice should act as a three-dimensional diffraction grating for X-rays. This should happen because X-rays have wavelength of the dimension of interplanar distances in a real crystal. Shortly after this, W.L. Bragg showed that wavelength of the X-ray undergoing diffraction by a crystal is related to the interplanar distances by the famous Braggs equation.

1.3.1 Braggs Law of Diffraction

Let there be a set of lattice planes consisting of an array of atoms as shown in figure 1.7. The X-ray beam incident on a plane at angle will be reflected from the plane such that the angle of
c c c
2a a a 2a

b
a 2a

a
(100) c 3a

a
(010)

(001)

c
3a 2a a

c
3a

a 2a 3a 3a

b
3a

a 2a 3a

a
(111)

a
(101)

a
(011)

Figure 1.7: Different sets of crystal planes in a cubic crystal

The Crystal Structure

reflection is also and a part of the intensity will pass though the crystal undeviated from its path. Reflection is caused by the interaction of the electromagnetic radiation with the electrons of the atoms in the lattice. In order that the intensity of the reflection is sufficiently strong, reflected waves from the successive planes separated by dhkl should be in phase. From figure 1.8, it is seen that the path difference of the waves from successive planes is 2d sin . In order that the waves travelling from successive planes are in phase the condition n = 2dhkl sin should be satisfied. This is Braggs condition of reflection and is known as Braggs law.
Path diff. = BC + BD nl = 2AB sin q = 2d sin q q A q q C B D

(1.2)

Figure 1.8: Diffraction of X-ray from a set of planes

1.4

RECIPROCAL LATTICE

The concept of reciprocal lattice is very useful in X-ray crystallography. It was Ewald who developed the relation between the diffracted X-ray beams. The crystal, instead of being seen as different sets of parallel planes, may be represented by a normal drawn perpendicular to each set of parallel planes from a common point as origin. The length of the normal is proportional to 1/dhkl. This length and direction of the normal is used to represent a set of parallel planes. If a point is placed at the end of each such normal, an array of points is generated. Each point then represents a set of parallel equidistant lattice planes and hence, each point is represented by a set of Miller indices (hkl) of the crystal. This array of points is known as the reciprocal lattice. The reciprocal lattice vector dhkl has a direction same as the normal to the dhk l planes and its magnitude is 1/dhkl. We see that the arrangements of the points in the reciprocal lattice has the same symmetry as the lattice points of the real crystal. The concept of reciprocal lattice is particularly helpful in understanding diffraction of Xrays by crystal planes. Let us rewrite the Braggs equation in reciprocal lattice as sin hkl =

1 d hkl 2 = 2 dhkl

(1.3)

Here, we have tried to relate the magnitude of the reciprocal lattice vector to diffraction angle and the wavelength of the X-ray. In order to see the geometric consequence of this equation, let us imagine a sphere of radius 1/ = AO as shown in figure 1.9.

Solid State Chemistry

P shkl 2q q q C O

Figure 1.9: Relation between reciprocal lattice point and X-ray diffraction

Let AO also be the direction of the X-ray beam incident on the crystal plane at C, the centre of the sphere. If is the Braggs angle, the reflected beam will strike the sphere (shown as a circle here) at the point P making an angle 2 with the passing beam. It should be noted that the angle between the incident beam and AP is . Since 2 sin hk l/ = 1/dhk l, we can see that 1/dhk l = OP. Thus, the point P is the reciprocal lattice point for the set of planes from which the X-ray beam is reflected. Also, since the Braggs diffraction conditions are satisfied, the diffracted beam will touch the sphere (circle in the picture) at point P. The same will be the case with other set of parallel planes except that they will strike the sphere at some other point which are the reciprocal lattice points for the respective set of planes. The three-dimensional sphere is called the sphere of reflection or Ewald sphere.

1.5

POWDER METHOD

The most common method of determining the structure of a crystalline solid is X-ray diffraction. There are many experimental variations. Complete determination of the crystal structure by locating the coordinates of all the atoms requires a good single crystal of, at least, about 1 mm size. But most substances crystallize as polycrystalline solids which means that each particle is made of a number of randomly oriented tiny crystals. Growing a single crystal needs some
Diffracted beam

X-ray O

2q

2q

Diffracted beam

Figure 1.10: Diffraction of X-ray by randomly oriented crystals

The Crystal Structure

special techniques and is not always easy. However, it is possible to get important structural information by recording the X-ray diffraction pattern of the powdered polycrystalline samples. This is commonly known as the powder method. Synthetic chemists prepare many solid materials in the laboratory. Before proceeding further, one would like to know whether the desired structure has been formed. This is done by recording the X-ray powder diffraction pattern of the material and comparing this with that of the known pattern. Extensive powder diffraction data have been compiled in the ASTM X-Ray Data Files that makes such comparison possible. Moreover, the powder diffraction patterns can provide important structural information such as the type of Bravais lattice, size of the unit cell and the space group. In the case of a simple crystal, it is even possible to determine the coordinates of all the atoms by analyzing these patterns. No surprise that this method has become a useful tool of the solid state chemists. If a set of Miller planes satisfy the Braggs condition, the reflected beam will emerge making an angle 2 with the undeflected beam. Since the crystals are randomly oriented, the same Miller planes in another crystal may satisfy Braggs condition, but the deflected beam at angle 2 may have a different direction as shown in the figure 1.10. Since the number of crystals is very large, all kind of orientations are possible and the diffracted X-ray will form a cone with angle 4.

1.5.1 Debye-Scherrer Method

Debye and Scherrer devised a specially designed camera that used photographic film for detecting the diffracted X-ray beams. The finely powdered sample in the shape of a thin rod is mounted vertically inside the camera which is surrounded by a strip of photographic film. The rod-like shape of the powdered sample is achieved by filling the powder in a short capillary made of glass or some polymer. The crystals inside the tube are randomly oriented. Hence planes with different Miller indices will satisfy the Braggs condition of reflection of X-ray. So as not to miss any set of parallel planes fulfilling the reflection condition, the rod is slowly rotated by a motor. In randomly oriented crystals, there are innumerable orientations of the same set of planes making the angle hkl with the incident X-ray beam. Hence the reflected X-ray will appear as a cone that makes an angle 2hkl with the direction of the beam. This cone will strike the wall of the camera in the form of a circle that will be registered in the thin strip of the film as two lines equidistant from the centre of the circle (see figure 1.11). If S is the measured distance between the centre of the two lines (part of the circle), then S 4hkl = radians R This gives S 180 degree hkl = 4R Thus, by measuring S, the corresponding value of can be found out. The interplanar distance dhkl can be found out using Braggs equation.

180 = 57.3, calculations can be simplified if the radius of the camera is made 57.3 mm or its integral multiple. Then S can be measured in mm and can be easily converted to in degree. This is the reason why cameras are made either of 57.3 mm or 114.6 mm radius.
Since The main disadvantages of the Debye-Scherrer method are: (i) the precision is limited by the measurement of S, (ii) the time of exposure is large, often several hours, (iii) measurement of the

10

Solid State Chemistry

intensity of the lines is not very satisfactory and it is time consuming. For these reasons, this method is hardly used in recent times and has been replaced by the automatic X-ray powder diffractometer. We shall not discuss here the different sources of error while recording the X-ray diffraction pattern using the Debye-Scherrer camera as it is rarely used nowadays.

Diffraction cones

X-ray in

X-ray out

Film

(a)

(b)

Figure 1.11: Diffraction from a powdered sample using Debye-Scherrer camera: (a) diffracted cones and (b) the part of the cones as pairs of lines in the uncoiled film

1.5.2 The Powder Diffractometer

In this instrument, the powder sample in the form of a thin circular disc of about 15 mm diameter 1-2 mm thickness is placed on a holder standing vertically. The holder can be rotated about an axis perpendicular to the table. The X-ray beam is allowed to fall on the sample making an angle with the surface. An X-ray photon counter rotates in a circle around the sample detecting all the reflected beams in turn. For better focussing of the reflected beam on the detector, the sample is rotated at a speed half of that of the detector. The detector converts the intensity into current and the diffraction angles is plotted as 2 against the current on a strip-chart recorder. As an example, the diffraction pattern of potassium bromide crystals is shown in figure 1.12. This method is fast and is much more precise both in terms of and intensity measurement.

80

70

60 2q

50

40

30

20

Figure 1.12: Powder diffraction pattern of potassium bromide

The Crystal Structure

11

1.5.3 Indexing the Powder Patterns

Identifying every d value in the powder pattern of a pure solid substance with as set of Miller indices hkl is known as indexing of a powder pattern. This is done using the equations that relate the lattice parameters for the cyrstal a, b, c and , , with dhkl. Indexing is simple for cubic crystals and not difficult for tetragonal and hexagonal crystals, but may be quite difficult for crystals of low symmetry like those with monoclinic or triclinic unit cells. For cubic crystals, a2 = (h2 + k2 + l2)d2hkl (1.4) Since = is constant for a given substance, it is possible to choose the hkl values for all d values in such a way so that the right hand side of the above equation gives a constant value. Thus the lattice parameter = of the cubic crystal can be found out. Rearranging the above equation, we get (1.5) d The value of = is chosen such that =2/d2 is an integer for all the observed values of d. This makes indexing easier.
2

a2

= h2 + k2 + l2

An alternative to this numerical method of indexing is a graphical technique. According to this method, some d values are calculated with assumed values of a using equation (1.5) for different hkl. Values of a between 0 to 20 are sufficient. a is then plotted against d using a 20 20 cm chart paper giving a number of straight lines each corresponding to a set of hkl. A strip of chart paper is cut and the experimental d values are marked using the same scale as used in the chart paper. The strip is then slid over the chart paper (see figure 1.13) so that each and every mark on it coincides with one of the straight lines of some hkl. The hkl values for each d is thus identified. For indexing a tetragonal crystal, one can use the equation

1
2 d hkl

h2 + k 2 a2

l2 c2

(1.6)

When l = 0,

a2 (1.7) h2 + k 2 This equation makes it possible to find out a for those planes with l = 0. After getting a, value of c can be found out using equation (1.6) for the rest of the lines. A graphical method for indexing tetragonal crystals was devised by Hull and Davey.
2 d hk 0 =

For a hexagonal system, the equation used is

1
2 d hkl

4 l2 2 2 ( h + hk + K ) + 3a 2 c2

(1.8)

When l = 0, it becomes

3 a2 (1.9) 2 4 h + hk + k 2 which is used for finding a. Knowing a, c can be found out using equation (1.8). Powder patterns of crystals of low symmetry are difficult to index and will not be considered here.
2 d hk 0 =

12
321 400 410 330 411 331 420 222 311 310 221 300 220 211 210 200 111 110

Solid State Chemistry

20.00

100

15.00

d=0

10.00

a,

5.00

5.00

10.00 d,

15.00

20.00

Figure 1.13: Indexing a cubic crystal using graphical method

What is measured in the diffraction pattern are the angles 2 and the intensity of the peaks. Larger the angle, more accurately it can be measured. Large angles correspond to higher hkl planes. Hence, the lattice parameter calculated for the high hkl planes is more accurate. For example, let us say that we have calculated the lattice parameter a for a cubic crystal using equation (1.4) for each hkl. Instead of averaging these values of a, it will make more sense if we average only those values that have been calculated for the large hkl planes. If it is suspected that the instrument itself is a source of inaccuracy, the sample may be mixed with some crystalline solid for which the d values are accurately known. These lines can then be used for calibration. Accurate measurement of density of the crystal is very useful as it helps in finding out the number of formula units per unit cell. For this, the number of formula units per unit cell is assumed and the mass of an unit cell is calculated from the atomic weight of the elements. The volume of the unit cell is calculated from the lattice constants obtained from X-ray diffraction.

The Crystal Structure

13

Density (x-ray) is calculated as

nm N g/cm3 volume where n is the number of formula units, m is the mass (in amu) of the formula unit, N is Avogadro number. This density known as X-ray density is then compared with the density measured by any standard experimental method. A comparison of the two densities will give the correct number of formula units per unit cell.

xray =

Intensity of the Lines

Intensity of the diffracted lines depend on the scattering power of the atoms that occupy a Miller plane. Higher the number of electrons in the atom, higher is its scattering power. This is so because the scattering of the electromagnetic radiation takes place by interaction with electrons. Planes that have heavy atoms will give more intense diffraction lines, whereas it is difficult to detect very light atoms by X-ray diffraction because of their poor scattering power. It is easy to predict the powder pattern of a solid with cubic unit cell. Let us take the example of a solid having one kind of atom such as a metal. A primitive cubic structure is expected to show the diffraction lines corresponding to all possible hkl values. But two cubic metals, tungsten and copper show diffraction patterns that are quite different. Copper that has a fcc unit cell gives reflection corresponding to (111), (200), (220), (311), (222), (400), (331) etc. whereas the powder pattern of bcc tungsten gives X-ray diffraction lines for planes (110), (200), (211), (220), (301), (222), (321) .... This marked difference between the two diffraction patterns can be understood by considering the fcc and bcc lattice as consisting of two interpenetrating primitive cubic sublattices A and B in two diffeent ways. The distance travelled by parallel X-ray beams from the same hkl planes of the two sublattices is not equal and there will be a path difference between the two waves. If the path difference is an integral multiple of the X-ray wavelength , the two waves will superimpose giving high intensity to the diffracted beam. We say the two waves are in phase. If the two waves are out of phase, they will interfere destructively and no diffraction line will be observed. The phase difference is phase difference = 2(hx + ky + lz) where the atom of B sublattice is located at the point (ax + by + cz) assuming the atom of A sublattice as the origin. Note that the plane under consideration in both the sublattice is the same hkl plane. When the phase difference is radians, the amplitudes of the two waves interfere destructively and if the atoms at the two sublattices are the same (with identical scattering 1 power), the intensity will completely vanish. If the unit cell is bcc, x = y = z = and the phase 2 difference = (h + k + l). Hence all reflections for planes for which h + k + l is odd will vanish. The systematic absence of the lines for planes for which (h + k + l) is odd indicates the cubic crystal as bcc. It will only show those lines for which (h + k + l) is an even number. No such restriction exists for a primitive cubic unit cell. It can be similarly shown that for fcc unit cell, (h, k, l) should either be all even or all odd in order that the waves from both sublattices are in phase. So, a fcc metal will show only those lines for which (h, k, l) are all even or all odd. In the light of this discussion, the difference between the X-ray diffraction patterns of copper and tungsten can be appreciated. Let us next consider the X-ray powder diffraction patterns of two similar fcc binary crystals KCl and NaCl. Their diffraction patterns are shown in figure 1.14.

14
111 200 220 311 222 400 331 420 422 511 333 440 531 600 442 620 533 622 444

Solid State Chemistry

711 511 640

(a)

(b)

200

220

222

400

420 422

440 600 442 620

622

444

640

Figure 1.14: X-ray powder patterns (a) KCl and (b) NaCl

The first thing that strikes us is that the angles (given by the position of the lines) due to the same set of Miller planes are slightly larger in NaCl. This is because the NaCl unit cell is slightly smaller than that of KCl. The more striking difference is the absence of certain lines like (111), (311), (511) etc. in KCl although these lines are present in powder pattern of NaCl. From systematic absence rules for fcc crystals, we know that the lines with mixed indices should be absent for both NaCl and KCl. This is indeed so. Further, for NaCl we see that the successive orders of (111) planes (these are 111, 222, 333, 444 etc.) are alternately weak and strong. For example, the reflection from (111) is weak and that from (222) is strong and so on. We know that in the binary fcc compounds like NaCl or KCl, the (111) planes are alternately occupied by Na+ (or K+) and Cl ions. If the scattered waves from two or more such planes containing only sodium are in phase and intensify the reflection, the planes containing sodium will be out of phase with planes containing chlorine and will interfere destructively thus diminishing the intensity. This is the reason why (111) reflection in NaCl is weak and (222) reflection is strong. For KCl, the alternate (111) planes have potassium and chlorine. These two ions have the same number of electrons and identical scattering power. So far as X-ray is concerned, K+ and Cl are identical. Since the radiation scattered by these planes are out of phase, the reflected waves will get completely annihilated and the reflections from (111), (333) etc. will not be seen. X-ray scattering power of an atom f is proportional to the number of electrons in the atom. One need not consider the scattering power of all the infinite number of atoms in a crystal. It is enough to consider the atoms of one unit cell. We must know the location of the atoms in a unit cell. This will enable us to determine what is called the structure factor (Fhkl) for a reflection hkl. Fhkl = f1 e 2i( hx1 + ky1 + lz1 ) +. . . + f n e 2i( hxn + kyn + lzn ) (1.10)

where N is the number of atoms in the unit cell and xnynzn defines the coordinate of an atom. Taking the position of a cation as the origin, NaCl crystal will have in its unit cell Na+ at 000,

11 1 11 1 1 111 1 1 1 0 , 0 and 0 and Cl ions at , 00 , 0 and 00 . Putting these coordinates 22 2 22 2 2 222 2 2 2 and simplifying, we get
Fhkl = [ fNa+ + fCl e i( h + k + l ) ][ e 0 + e i( h + k ) + e i( h + l) + e i( k + l ) ] (1.11)

642

The Crystal Structure

15

Since ein is +1 when n is even and 1 when n is odd, the Miller indices hkl with all even or all odd will make the terms within the second square brackets in equation (1.11) equal to 4. With mixed hkl, this part of the equation becomes zero. So the structure factor for NaCl is
+ + f ] e i( h + k + l ) Fhkl = 4 [ fNa Cl

(1.12)

which can be further simplified as

+ + f ) when h, k, l are all even, and Fhkl = 4(fNa Cl + f ) when h, k, l are all odd. Fhkl = 4(fNa Cl
2 . This means that if the cation Intensity of the reflected X-ray beam is proportional to Fhkl

and the anion had the same scattering power, all reflections with h, k, l all odd will vanish as it happens in the case of KCl. These lines appear as weak reflections for NaCl because of the unequal scattering power of sodium and chloride ions. The powder method is inadequate for locating the coordinates of every single atom in a complex solid that needs single crystal X-ray structure determination. This is beyond the scope of this book.

1.5.4 Use of X-ray Powder Patterns

Despite all its limitations, X-ray powder method has become the most important tool for chemists who study solid materials, metals and inorganic solids in particular. It is most commonly used to find out if the desired phase is present in a synthesized or natural material by comparing its X-ray pattern with those of the known materials of similar composition. The presence of any impurity phase can be detected by this method. In a solid state reaction to synthesize some ceramic material, it can be used to check the completion of the reaction. For simple structures, it allows to determine the unit cell type and size, the number of formula units per unit cell and in some simple cases, even intensity analysis is possible. Its one limitation is that the lines become broad and finally disappear as the particle size decreases. The most appropriate size of the particles should be between 100 nm to 103 nm. Another shortcoming is its inability to identify hydrogen atoms thus severely restricting its application to organic compounds.

1.6

CRYSTAL DEFECTS

So far in this chapter we have considered the crystal lattice in which all the sites meant for some atoms are occupied appropriately. In the NaCl structure, for example, every site meant for sodium are occupied by sodium and the same is true for chlorine atoms. It did not happen that a chlorine atom occupied a site that is meant for a sodium atom. Such a crystal is called a perfect or an ideal crystal. A perfect arrangement gives the crystal the highest possible negative value of lattice energy and hence apparently such a structure should be most stable. Stability, however, is determined by free energy G which for a solid may be taken as A, the Helmholtz free energy expressed as A = U TS Further, we can write lattice energy UL in place of internal energy U. This equation tells us that A = UL will be minimum only when T = 0 K, i.e., when the contribution of entropy towards free energy is zero. Entropy of an disordered (imperfect) crystal is higher than that of an ordered crystal. Thus, at any temperature other than 0 K, two factors will contribute to free energy.

16

Solid State Chemistry

Perfection in the crystal will minimize the free energy by making the internal energy UL more negative, and the contribution from entropy ST will try to minimize the same by making the crystal more imperfect thus increasing S. The equilibrium structure at any temperature other than 0 K will have some disorder or imperfection. We see that the presence of defects in a crystal is a thermodynamic requirement for stability. The main defects in a crystal are of three types: point defects, dislocations and grain boundaries.

1.6.1 Point Defects

The simplest type of point defect in a crystal is a vacancy. When a lattice site that normally is to be occupied by an atom or an ion is left vacant, we have a point defect known as vacancy and it has the symbol V. If the vacant site was supposed to be occupied by an atom X in the crystal, the symbol is VX meaning the vacancy occurs at a site that is meant for an X atom. If an atom or ion is removed from its normal position in the lattice and is placed at an interstitial site, the point defect is known as an interstitial. If M is the atom that is at an interstitial site, it is given the symobl Mi. Similarly, there may be an interstitial Xi which means an X atom instead of being present at its normal site, occupies an interstitial site. If an atom A in a crystal of composition AB instead of being present at a site meant for it goes to occupy a site which is normally occupied by an atom B, the point defect is called a misplaced atom. The symbol for this is BA. Here the letter B is the site for atom B and the subscript ^ shows the atom that is occupying the site. It goes without saying that AA and BB mean normal lattice sites but AB and BA mean misplaced atoms. A crystal may have in it a variety of point defects but which type will predominate depends on the energetics of defect formation. Many crystals show combination of defects i.e., one type of defect is associated with an equivalent number of defects of another type. Such combination defects are particularly important in ionic crystals for maintaining overall charge neutrality of the crystal. Two such defects of particular importance are the Frenkel and Schottky defects. When an atom or ion is removed from its normal site and placed at an interstitial site, the crystal has simultaneously a vacancy and one interstitial. This combination is known as a Frenkel defect. In a crystal, when a cation vacancy exists with an equivalent number of anion vacancies necessary for its charge neutrality, the combination is known as Schottky defect. Frenkel and Schottky defects are shown in figure 1.15.
+ + + + + + + + ( a) + + + + + + + + + + (b ) + + + + + + + + + + +

Figure 1.15: (a) Frenkel defect and (b) Schottky defect in an ionic crystal

The Crystal Structure

17

The formation of Schottky defect is accompanied by the creation of an equivalent number of new lattice sites or removal of the ions to the gas phase from the lattice. The presence of foreign atoms constitutes another type of point defect. These atoms or ions can be present either in interstitial or in substitutional positions. In an ionic crystal, the introduction of a foreign atom is dependent on its ionic radius and electronic structure. Sometimes, it is possible for the same ion to enter the lattice both interstitially and substitutionally. Substitution of a lattice ion by an impurity ion of different charge will disturb the charge balance. In order to maintain the charge neutrality of the lattice, such substitution may be accompanied either by the creation of a lattice vacancy or by the change of the oxidation state of an ion in the lattice. Substitution of Ag+ by Cd2+ in AgCl leads to the former, whereas substitution of Ni2+ by Li+ in NiO lattice gives rise to the latter. These two cases are illustrated below, one creating a cation vacancy, and the other a valence defect. Here Ni3+ is known as a valence defect. Ag+ Cl Ag+ Cl Cl Cd2+ Cl Ag+ Cl Ag+ Cl Cl Ag+ Cl Ag+ Ni2+ O2 Ni2+ O2 O2 Ni3+ O2 Ni2+ Ni2+ O2 Li+ O2 O2 Ni2+ O2 Ni2+

Vacancy produced by Cd2+ substitution

Valence defect produced by Li+ substitution

Unlike the inherent thermodynamic defects, the impurity defects can be deliberately added in a controlled way that is known as doping. This helps in modifying the properties of the solid. There is another type of defect that very much influences the electronic properties of the solid. These are the electronic defects. Let us assume that a solid has an anion vacancy. The anion vacancy is the site that normally should have been occupied by an anion. Hence this site (surrounded by cations) has a strong positive potential and can trap electrons. It will need some energy to free the trapped electron. When sodium chloride is heated in sodium vapour, we see that is acquires a yellow colour. The heating in excess sodium creates new lattice sites for sodium. To maintain the structure, an equal number of chloride sites are also created, but the latter are vacant. The electron of the sodium atom that is ionized in the NaCl matrix is trapped by the chloride ion vacancy. This trapped electron can be freed into the crystal by absorbing visible light and hence the colour. A trapped electron at an anion vacancy is an electronic defect and it is known as F centre (from German Farbe meaning colour). In fact, many of the point defects may have a charge different from the normal charge of the original site due to trapped electrons or holes. These electrons and holes can be released to the conduction band or to the valence band of the solid and this will modify its electronic properties. The symbols of the point defects then should also include the information of the charge that they carry. It is more convenient to talk of an effective charge. The effective charge of a point defect is taken as the difference of its charge from the charge of the ion that normally occupies that site. Let us illustrate this with some examples. Let us say that there is a cation vacancy in the MgO lattice. This vacant site was originally occupied by a Mg2+ ion that had a charge +2. Now it has zero charge and hence its effective charge is 2. We write the symbol VMg , where the superscript stands for two negative effective charges. If the cation vacancy captures a hole, its effective charge becomes uninegative and the symbol used is VMg . If it captures another hole, it will have a real charge +2, but a zero effective
X charge and will be denoted as VMg . An oxygen vacancy with zero charge is having two effective

18

Solid State Chemistry

will be unipositive and the symbol for it is VO . If it captures a second electron, its effective charge

positive charges and will be represented as VO . If it captures one electron, its effective charge X , although its true charge is 2. Similarly, the charge of other will be zero and the symbol is VO types of point defects should be included in the symbol. Thus, if a bivalent metal ion goes to a normally unoccupied interstitial site, the effective charge of the interstitial site becomes +2 because originally this site did not have any charge. In this case, however, the real charge too is +2. The inclusion of the effective charge in the notation of the point defects is important as complete balance of effecticve charge will have to be maintained in writing balanced equations for defect reactions.

1.6.2 Defect Equations

It should be remembered that the total effective positive and negative charges should be equal in a crystal. In writing an equation involving point defects, the following rules should be followed. 1. The ratio of the number of regular cationic and anionic sites in a solid is constant. If the crystal has a NaCl structure, this ratio is unity. If the solid crystallizes with the fluorite structure, this ratio is 1: 2. This will be so even if the solid is non-stoichiometric. This means that if we create a Na lattice site, we are obliged also to create a Cl lattice site even if that site is to be kept vacant. 2. Defect equation may include creation of new lattice sites as well as elimination of the existing lattice sites i.e., the total number of lattice sites may change. 3. The total mass balance should be maintained which means the total number of each type of atoms on both sides of the equation should be the same. 4. The total effective charge should be same on both sides, i.e., a charge balance should be maintained. Next, let us try to write some defect equations following the rules stated above. Here only stoichiometric compounds will be considered. Non-stoichiometeric oxides will be considered later. Let us first consider Schottky defects in a stoichiometric oxide MO. The equation for the formation of a Schottky defect can be written as
MM + OO l VM + VO + MM + OO

The regular lattice sites on the right hand side means that the formation of Schottky defect requires that new lattice sites are created. Cancelling common terms, we write
O l VM + VO

where O stands for a perfect lattice. Here we have assumed that the doubly charged vacancies predominate. It is obvious that the doubly charged vacancies can capture or release electrons and can either become singly charged or neutral. If these species are to be included in the equation, the corresponding electrons or holes are also to be added. A Frenkel defect is created by transferring a regular ion to an interstitial site. Suppose the interstitial ion is a cation as is generally the case, we can write
MM l VM + Mi

assuming that the doubly charged interstitials and vacancies predominate.

The Crystal Structure

19

We have considered the important types of point defect. Next we shall look into the concentration of the thermodynamic defects under specific conditions. The equilibrium concentration of the inherent thermodynamic defects is dependent on temperature. It is possible to calculate this concentration at any given temperature provided the energy required for their formation is known. Here we shall derive the equations that relate the defect concentration to temperature.

1.6.3 Vacancy Concentration in a Metal

The change of an ideal crystal at constant volume and temperature to a defect crystal with vacant lattice sites will be accompanied by a change in Helmholtz free energy such that A = U TS (1.13) where A, U and S are the respective changes in free energy, internal energy and entropy. At equilibrium, the change in free energy with the change in the number of vacancies n will be zero. We can write
a Af I F G H n JK

= 0
T ,V

(1.14)

If is the energy needed to create a vacancy, U = n. (1.15) Neglecting the contribution to entropy by the vibration of the atoms which in any case is small, the entropy of the defect structure will be essentially its configurational entropy ( Sconf. = S). Hence, S = k ln w (1.16) where k is the Boltzmann constant and w the thermodynamic probability which is equal to the number of ways n vacant sites and (N n) occupied sites can be arranged among the N available lattice sites. This will give N! w = (1.17) N n !n!

Substituting this in equation (1.16) and applying Stirlings approximation, we get S = k N ln N N n ln N n n ln n It is now possible to write

f a

(1.18)

A = n kT N ln N N n ln N n n ln n

f a

(1.19)

Differentiating A with respect to n and equating to zero (condition of equilibrium concentration of defects), we get = kT N n ln N n n ln n n This will give N n = kT ln n and n = exp (1.20) kT N n

f a

FG H

IJ K

20

Solid State Chemistry

Since N >> n,

n = exp N

FG IJ H kT K FG IJ H K
(1.21)

If EV is the energy necessary to create one mole of vacancy, then

EV = Avogadros number and

n E = exp V N RT

Thus, knowing the energy of fromation of vacancies, it is possible to calculate the number of vacancies per mole of metal atoms at a given temperature.

1.6.4 Schottky Defect Concentration in an Ionic Crystal

Let us next calculate the equilibrium concentration of Schottky defects in an ionic crystal of stoichiometry MX. Let s be the energy necessary to create a Schottky defect. This means that this is the energy needed to produce one cation and one anion vacancy in MX. If the total cationic site is N and the number of action vacancy is ns, the number of occupied cationic sites will be (N ns). The number of ways in which ns cation vacancies and (N ns) cationic sites can be arranged in N available cationic sites is w =

n! N ns ! ns !

(1.22)

There will be ns anion vacancy since stoichiometry is MX. The number of ways in which ns anion vacancies and (N ns) anions can be distributed among N anionic sites is also equal to w given by equation (1.22). The total probability W = w.w. The configurational entropy is S = 2k ln which means

bN n g ! n !
s s

N!

(1.23)

S = 2k N ln N N n s ln N n s n s ln ns We can now write A = ns 2kT N ln N N ns ln N ns ns ln ns

g b

(1.24)

g b

(1.25)

In order to minimize A, we differentiate the above equation with respect to ns and equate that to zero. After rearranging the result, we get

ns N

= exp

FG IJ H 2kT K
s

(1.26)

If Es is the energy of formation of one mole of Schottky defect in the crystal MX, then

ns N

= exp

FG E IJ H 2RT K
s

(1.27)

The expression for Schottky defect concentration for crystal of composition MX2 can be similarly deduced.

The Crystal Structure

21

1.6.5 Frenkel Defect Concentration

A crystal with Frenkel defects will have equal number of vacancies and interstitials. Let the number of Frenkel defects be nf. If N is the total number of lattice sites, there will (N nf) occupied sites. The number of ways in which nf vacancies (N nf) occupied sites can be arranged in N available lattice sites gives the probability w =

eN n j ! n !
f f

N!

(1.28)

Let the total number of interstitial sites be N*. The number of ways in which nf interstitial atoms and (N* nf) unoccupied interstitial sites can arrange gives the probability w* =

eN * n j ! n !
f f

N *!

(1.29)

Total configurational probability W = w.w* In this case, S = k [ln

eN n j ! n !
f f f

N!

+ ln

eN * n j ! n !
f f

N *!

(1.30)

The change in free energy in forming nf number of Frenkel defects is A = nf kT [ln Solving

eN n j ! n !
f

N!

+ ln

eN * n j ! n !
f f

N *!

(1.31)

F AI G H n JK
f

= 0
V ,T

(1.32)

we get

n2 f ( N n f ) ( N * n f )

= exp

F I GH kT JK
f

Since nf is very much smaller than both N and N*, we get nf f Ef = exp = exp NN * kT kT

F GH

I JK

F GH

I JK

(1.33)

where Ef is the energy needed to form a mole of Frenkel defects. The concentration of the various types of point defects present at a temperature T can be calculated if the energy of defect formation is known.

1.6.6 Dislocations
Dislocations also known as line defects and are of two kinds. They are edge dislocation and screw dislocation. An edge dislocation is what would happen if a half plane is inserted into a crystal. The dislocation extends along a line perpendicular to the plane of the paper. The structure is distorted near the dislocation and several lattice constants are to be covered before this distortion disappears. Edge dislocation can move along the crystal under a shear force. Suppose force is applied in

22

Solid State Chemistry

opposite directions at the two ends of a crystal. Instead of separating it into two parts that would need a large number of bonds to be broken, atoms on one side can just move a short distance. This will make the dislocation move along the crystal in the direction of the shear force. Movement of the edge dislocations in the crystal are responsible for their plastic property. Edge dislocation may form accidentally during crystal growth or may be produced by bending a crystal. Dislocation density in a common crystal is of the order 106/cm2.

Figure 1.16: Edge dislocation

Edge dislocation are given the symbol as shown in figure 1.16 where the dislocation line is perpendicular to the plane of the paper. Dislocations can be identified by taking electron microscope pictures at magnification 105 or higher where the lattice planes appear as straight lines. Dislocations show up as disruption in these straight lines. The other type of line defect is known as screw dislocation. Screw dislocation results from movement of one part of a crystal relative to another. The distance of movement is less than a lattice constant and hence the coordination number of the atoms near the screw dislocation does not change. The screw dislocation only leads to a distortion of the bonds in its vicinity. If one moves from lattice point to lattice point around the dislocation, one will move up as if in a spiral staircase. The name screw dislocation is derived from this (see figure 1.17). Dislocations severely affect the mechanical properties of solids.

Figure 1.17: Screw dislocation

The Crystal Structure

23

1.6.7 Plane Defects or Grain Boundaries

So far we discussed defects that are present within a crystal. Many solids however does not exist as a single crystal but as aggregate of a large number of small crystals or grains. These grains have different orientations. As a result, the environment of the atoms near the boundary of two such grains is bound to be different from what it is within a crystallite of grain. The periodicity of the atoms will be destroyed at and near the grain boundary region and this region will be noncrystalline. If this non-crystalline region is fairly thick, each crystal will try to force the atoms of this region to orient themselves in accordance with its own structure. This will be so because the irregularity in the non-crystalline region will increase the energy of the solid. Grain boundaries may be wide-angle (figure 1.18) or small-angle depending upon the tilt of the angle of the boundary. Small-angle grain boundaries can be treated as a series of dislocations. When the grain boundary becomes a reflection plane, the two crystals on either side of the reflection plane constitutes a twin.

Figure 1.18: Wide-angle grain boundary

Grain boundaries seriously affect the mechanical properties of solids. They can be removed by heat treatment. At high temperature when the mobility of the atoms increases, some grains grow at the expense of others and this process is known as secondary recrystallization.

SUGGESTED READING
1. Azaroff, L.V. Introduction to Solids, TMH edition, Tata McGraw-Hill, New Delhi. 2. Azaroff, L.V. and Buerger, M.J. The Powder Method in X-ray Crystallography, McGraw-Hill, 1958. 3. Epifanov, G.I. Solid State Physics, MIR, Moscow, 1979. 4. Hannay, N.B. Solid State Chemistry, Prentice-Hall, New Jersey, 1967. 5. Keer, H.V. Principles of the Solid State, Wiley Eastern, New Delhi, 1993. 6. Kittel, C. Introduction to Solid State, 3rd edition, John Wiley, 1966. 7. Megaw, H.D. Crystal Structure: A Working Approach, Saunders College Publ., 1973.