The Reciprocal Lattice
Total Page:16
File Type:pdf, Size:1020Kb
Appendix A THE RECIPROCAL LATTICE When the translations of a primitive space lattice are denoted by a, b and c, the vector p to any lattice point is given p = ua + vb + we. The definition of the reciprocal lattice is that the translations a*, b* and c*, which define the reciprocal lattice fulfil the following relationships: (A. I) a*.b = b*.c = c*.a = a.b* = b.c* =c. a*= 0 (A.2) It can then be easily shown that: (1) Ic* I = 1/c-spacing of primitive lattice and similarly forb* and a*. (2) a*= (b 1\ c)ja.(b 1\ c) b* = (c 1\ a)jb.(c 1\ a) c* =(a 1\ b)/c-(a 1\ b) These last three relations are often used as a definition of the reciprocal lattice. Two properties of the reciprocal lattice are particularly important: (a) The vector g* defined by g* = ha* + kb* + lc* (where h, k and I are integers to the point hkl in the reciprocal lattice) is normal to the plane of Miller indices (hkl) in the primary lattice. (b) The magnitude Ig* I of this vector is the reciprocal of the spacing of (hkl) in the primary lattice. AI ALLOWED REFLECTIONS The kinematical structure factor for reflections is given by Fhkl = I;J;exp[- 2ni(hu; + kv; + lw)] (A.3) where ui' v; and W; are the coordinates of the atoms and hkl the Miller indices of the reflection g. If there is only one atom at 0, 0, 0 in the unit cell then the structure factor will be independent of hkl since for all values of h, k and I we have Fhkl =f. Thus if b.c.c. and f.c.c. crystals are referred to their primitive cells then reflections from the simple metals such as niobium and copper, which have only one atom at each lattice point, will all have intensities given by f(0)2 • Since the three basic g vectors derived below, for the b.c.c. structure for 180 ELECTRON BEAM ANALYSIS OF MATERIALS Table A.l Necessary conditions for allowed reflections in terms of the type of unit cells in crystals. Possible Forbidden Unit cell reflections reflections Primitive All values of h, k and I none Body centred (h + k + I) even (h + k +I) odd Face centred h, k, I all odd or even h, k and l mixed Base centred h and k both odd or even hand k mixed example, are [11 OJ*, [0 11 ]* and [10 1]* it is clear that the allowed reflections are those derived by summing or subtracting these vectors, i.e. reflections such that (h + k + l) is always even. A similar conclusion is reached if the cubic structure cell, which contains atoms at 0, 0, 0 and ±. ±. ±. is used since Fhkt = f { exp(- 2ni0) + exp[- 2ni( ~ + ~ + ~) ]} = {f(l + 1) = 2f if (h + k +/)is even !(1-1)=0 if(h+k+/)isodd. The allowed reflections in the various possible unit cells are shown in Table A.1 If the number of atoms in a unit cell is large there is the possibility that some of the allowed reflections will have zero intensity, e.g. the 200 reflection in silicon. Nevertheless all the allowed reflections must conform with the above classifications. It is of course an important part of structure determination to recognize which reflections are absent (see Chapter 4). The absent reflections in metals such as zirconium (i.e. those for which (h + 2k) is a multiple of 3 and I is odd) show that the structure is h.c.p .. A.2 RECIPROCAL LATTICE FOR F. C. C. AND B.C. C. CRYSTALS Referring the primitive translations to cubic axis for a primitive f.c.c. cell we have where i, j and k are unit vectors along the cubic axes. Since a*= b 1\ c a.(b 1\ c) APPENDIX A 181 then * _ !a[j + k] A !a[i + k] a - !a[i + jJ.!a[j + k] A !a[i + k] and 1 z[· · k] I I 4a 1 + J- = -[i + j- k] =-[III] ia3 [i+jJ.[i+j-k] a a Correspondingly, b* =~[Ill] a with a b.c.c. structure and referring translations to cubic axes then a f] a _ a _ a=-[111 b=-[111] c=-[111] 2 2 2 so that * _ !a[- i + j + k] A !a[i-j + k] a -!a[i+j-kJ·!a[-i+j+k] A!a[i-j+k] which reduces to -!a 2 [i + j] I sa1 3['•+J- • k] . ['•+J '] = -[110]a Similarly, I and c* = -[101] a Note that the basic vectors in the reciprocal lattice for these non-primitive cells are twice those of the crystal of the same system. Thus, the basic translations are aj2 (Ill) for a b.c.c. crystal and Ija (Ill), rather than I j2a ( Ill ) , for the reciprocal lattice for an f.c.c. crystal. A.3 DOUBLE DIFFRACTION Because electrons are strongly scattered it is possible that rescattering of a diffracted beam can give rise to a strong diffracted beam where structure factor considerations suggest the beam should be of zero intensity. The conditions under which a forbidden diffraction spot may appear are most easily seen using the Ewald sphere construction (see Chapter 4). Thus, if the reciprocal lattice point corresponding to g 1 for which the structure factor is large, lies on the Ewald sphere then a strong diffracted beam will be produced in the direction A 1 , as shown on Fig. A.l. Similarly if g2 also 182 ELECTRON BEAM ANALYSIS OF MATERIALS Fig. A.l Ewald sphere construction showing the conditions which must be fulfilled for double diffraction to occur. See text. lies on the sphere a diffracted beam would be expected in the direction A2 . However, if the structure factor is zero for g2 the intensity of A2 would be zero but rediffraction of the beam A 1 by planes perpendicular to g2 - g 1 will give rise to a beam A3 which is parallel to and therefore indistinguishable from A2 . Diffraction maxima due to double diffraction can be distinguished by rotating the crystal about the direction which contains the spot in question. The intensity of this spot will be unchanged unless it is due to double diffrac tion when it will disappear when g 1 is no longer excited. A.4 SHAPES OF DIFFRACTION MAXIMA Significant diffracted intensity is observed from thin samples even when the Bragg condition is not precisely satisfied ( [1 ], [2] ). It can be shown that only for an infinite crystal will the diffraction maxima be points and that as the crystal dimensions get smaller so the diffraction maxima get larger. For a parallelepiped crystal the diffracted amplitude is given by F ¢ g = V fA IBIC exp [ - 2:rri(ux + vy + wz) J dxdydz (A.4) c 0 0 0 APPENDIX A 183 I -3/A -2/A -1/A 0 1/A 2/A 3/A Fig. A.2 Predicted variation in intensity of a diffracted beam from a crystal of edge length A as a function of deviation from the Bragg condition where the deviation is measured in units of 1/A. See text. = ~ sin(nAu) sin(nBv) sin(nCw) (A.5) Vc nu nv nw where A, Band Care the edge lengths of the parallelepiped along x, y and z, and u, v, w are the values of s (the deviation from the Bragg condition) along x, y and z. Fig. A.2 shows how the intensity (obtained by multiplying ¢g by¢;, its complex conjugate), varies along u for v and w = 0. The central maximum has a width at half maximum height of 1/A and successive minima occur at intervals of 1/A. The intensity of the successive maxima decreases very rapidly, as shown in Fig. A.2 but because electron diffraction patterns have such a large dynamic range it is possible to detect maxima out to at least the fiftieth maximum for typical thickness samples. For very thin crystals, such as small precipitates, the spikes in reciprocal space give rise to visible intensity for all beam directions. The dimension of the central diffraction maximum along a given direction in reciprocal space is given by 1/d where d is the parallel dimension of the specimen. Thus a thin plate precipitate gives rise to a long spike in reciprocal space normal to the plate and a needle precipitate gives rise to a disc of intensity. REFERENCES I. Hirsch, P.B., Howie, A., Nicholson, R.B., Pashley, D.W. and Whelan, M.J. (1965) Electron Microscopy of Thin Crystals, Butterworths, Sevenoaks. 2. James, R. W. ( 1958) Optical Principles of the Diffraction of X-rays, Bell, London. Appendix B INTERPLANAR DISTANCES AND ANGLES IN CRYSTALS. CELL VOLUMES. DIFFRACTION GROUP SYMMETRIES B.! THE SEVEN SYSTEMS The axial lengths and angles and the symmetries exhibited m the seven crystal systems are shown in Table B.l. Table B.l The axial relationships and symmetries of the seven crystal systems. Axial length Crystal and angles Minimum symmetry elements Cubic Three equal axes Four, threefold rotation axes at right angles a = b = c, IX = f3 = y = 90° Hexagonal Two coplanar axes at One, sixfold rotation (or 120°, third axis at rotation-inversion) axis right angles a= b =/= C, IX= {3 = 90°, y = 120° Trigonal (or Three equal axes One, threefold rotation rhombohedral) equally inclined (or rotation-inversion) axis a = b = c, IX = f3 = y =!= 90° Tetragonal Three axes at right One, fourfold rotation (or angles rotation-inversion) axis a = b =!= c, IX = f3 = y = 90° Orthorhombic Three orthogonal Three, perpendicular twofold unequal axes (or rotation-inversion) axis a =!= b =!= c, IX = f3 = y = 90° Monoclinic Three unequal axes One, twofold rotation (or one pair not orthogonal rotation-inversion) axis a=!= b =f=c, IX= y = 90o =!= f3 APPENDIX B 185 Table 8.1 (Contd.) Triclinic Three unequal axes None none at right angles a +b +c, a +fJ + y +90° 8.2 INTERPLANAR SPACING The value of d, the distance between adjacent planes in the set (hkl), may be found from the following equations.