Structure of Materials and Crystalline Defects

Crystal lattices, the BCC, FCC and HCP structures, how to name positions, directions and planes, atomic densities, and the defects of real crystals.

📖 Source: Universidad Carlos III de Madrid · Session 3 T2 Structure of Materials I_2025.pdf and Session 4 T2 Structure of Materials II_2025.pdf

Central idea: the performance of a component depends on the properties of its material, and those properties depend on its structure, the way its atoms are arranged. This topic describes that arrangement (crystal lattices, the main metallic structures, how to name positions, directions and planes) and the defects that real crystals always contain.

Part I covers the perfect crystal: crystal systems, atomic packing, the BCC, FCC and HCP structures, interstitial sites, and how to describe positions, directions and planes. Part II covers densities, defects (point, linear and planar), solid solutions, and polymorphism.

Part I: Structure of materials

§ 1 Why the structure matters [Session 3 Slide 2]

The choice of a material for a part, such as an aircraft wing, depends on its properties: a ceramic like diamond would be too brittle, copper too weak, a rubber too flexible. Those properties are not random: they come from the structure, the arrangement of the components of the material.

$$\text{performance of a component} = f(\text{properties}), \qquad \text{properties} = f(\text{structure})$$

§ 2 Crystalline and amorphous solids [Session 3 Slides 3-4]

Why atoms order themselves: atoms tend to sit in the positions of minimum energy. A dense, ordered packing has lower energy than a loose, random one, and the result is a uniform, repeating arrangement called the crystal lattice.

§ 3 Lattice and unit cell [Session 3 Slides 5-7]

The idea goes back to Haüy (1781): however small he broke a calcite crystal, the pieces were always rhombohedral and fitted together with no gaps. So the crystal had to be made of identical small units packed perfectly. That raises the question of how many different shapes can fill space completely with no voids.

§ 4 Lattice parameters and the seven crystal systems [Session 3 Slide 8]

A unit cell is defined by three lattice vectors a, b and c that start at a common origin. It needs six lattice parameters:

Classifying cells by these parameters gives seven crystal systems:

SystemEdgesAngles
Cubic$a = b = c$$\alpha = \beta = \gamma = 90^\circ$
Hexagonal$a = b \neq c$$\alpha = \beta = 90^\circ,\ \gamma = 120^\circ$
Tetragonal$a = b \neq c$$\alpha = \beta = \gamma = 90^\circ$
Rhombohedral$a = b = c$$\alpha = \beta = \gamma \neq 90^\circ$
Orthorhombic$a \neq b \neq c$$\alpha = \beta = \gamma = 90^\circ$
Monoclinic$a \neq b \neq c$$\alpha = \gamma = 90^\circ \neq \beta$
Triclinic$a \neq b \neq c$$\alpha \neq \beta \neq \gamma \neq 90^\circ$

§ 5 Bravais lattices [Session 3 Slides 9-14]

The crystal system fixes the shape of the cell, but not where the atoms sit in it. Bravais (1848) showed that atoms can be placed in these seven systems in only 14 different ways: the Bravais lattices.

There are four ways of placing atoms in a cell:

Combining the 7 systems with these 4 types gives the 14 Bravais lattices:

SystemBravais lattices
Cubicsimple, body-centred, face-centred
Tetragonalsimple, body-centred
Orthorhombicsimple, base-centred, body-centred, face-centred
Rhombohedralsimple
Hexagonalsimple
Monoclinicsimple, base-centred
Triclinicsimple

§ 6 Atomic packing [Session 3 Slides 16-22]

Imagine the atoms as hard spheres arranged in flat layers:

Stacking non-compact layers can still give a compact structure: BCC is built this way, with each layer sitting over the gaps of the one below.

Stacking close-packed layers can be done in two ways, because the third layer has two choices of hollows to sit in:

Stacking sequences of close-packed layers
Figure 1. Stacking of close-packed layers: A B A B ... gives hexagonal close packing (HCP, left) and A B C A B C ... gives face-centred cubic packing (FCC, right). Copied from Session 3 Slide 22.

§ 7 Main metallic structures [Session 3 Slides 24-28]

About 90 % of metals crystallize in one of three densely packed structures. They do so because:

StructureExamples
Body-centred cubic (BCC)Fe, Cr, Mo, V, W, K
Face-centred cubic (FCC)Al, Cu, Pb, Ni, Pt, Ag
Hexagonal close-packed (HCP)Ti, Mg, Cd, Co, Zn
BCC, FCC and HCP unit cells
Figure 2. Unit cells of the three main metallic structures: body-centred cubic (BCC), face-centred cubic (FCC) and hexagonal close-packed (HCP). Copied from Session 3 Slide 25.

To compare them we use three numbers: how many atoms belong to one cell, how many neighbours each atom touches (the coordination number), and how much of the cell volume the atoms fill:

$$\text{Atomic Packing Factor (APF)} = \frac{\text{volume of the atoms in a unit cell}}{\text{total volume of the unit cell}}$$

Counting atoms per cell. An atom at a corner is shared by the 8 cells that meet there, so only ⅛ of it belongs to each cell. An atom on a face is shared by 2 cells (½ each). An atom inside the cell belongs to it completely.

7.1 Body-centred cubic (BCC) [Session 3 Slide 26]

$$4R = \sqrt3\,a \quad\Rightarrow\quad a = \frac{4R}{\sqrt3}$$
$$APF = \frac{2\cdot\frac43\pi R^3}{\left(\frac{4R}{\sqrt3}\right)^3} = \frac{\frac83\pi R^3}{\frac{64R^3}{3\sqrt3}} = \frac{\sqrt3\,\pi}{8} = 0.68$$

68 % of the cell is occupied by atoms.

BCC reduced-sphere and hard-sphere cells
Figure 3. BCC: reduced-sphere cell with the face diagonal d and the edge a, and hard-sphere cell with the body diagonal along which the atoms touch, 4R = √3·a. Copied from Session 3 Slide 26.
FCC reduced-sphere and hard-sphere cells
Figure 4. FCC: reduced-sphere cell with the face diagonal d and the edge a, and hard-sphere cell where the atoms touch along the face diagonal, 4R = √2·a. Copied from Session 3 Slide 27.

7.2 Face-centred cubic (FCC) [Session 3 Slide 27]

$$4R = \sqrt2\,a \quad\Rightarrow\quad a = 2\sqrt2\,R$$
$$APF = \frac{4\cdot\frac43\pi R^3}{\left(2\sqrt2\,R\right)^3} = \frac{\frac{16}{3}\pi R^3}{16\sqrt2\,R^3} = \frac{\pi}{3\sqrt2} = 0.74$$

74 % of the cell is occupied: FCC is a close-packed structure.

7.3 Hexagonal close-packed (HCP) [Session 3 Slide 28]

$$V_{cell} = B\cdot c = \frac{3\sqrt3}{2}a^2\cdot\frac{2\sqrt2}{\sqrt3}a = 3\sqrt2\,a^3 = 3\sqrt2\,(2R)^3 = 24\sqrt2\,R^3$$
$$APF = \frac{6\cdot\frac43\pi R^3}{24\sqrt2\,R^3} = \frac{8\pi}{24\sqrt2} = \frac{\pi}{3\sqrt2} = 0.74$$

Like FCC, 74 % of the cell is occupied. That is expected: FCC and HCP are the two ways of stacking the same close-packed layers.

§ 8 Interstitial sites [Session 3 Slides 30-35]

Even in the densest packing, 26 % of the volume is empty. The gaps between the regular atom positions are interstitial sites: places that other (usually smaller) atoms or ions can occupy. There are two kinds:

In compact structures, with $n$ atoms per cell:

$$\text{tetrahedral sites} = 2n, \qquad \text{octahedral sites} = n$$

FCC ($n = 4$):

BCC ($n = 2$):

HCP ($n = 6$): 12 tetrahedral and 6 octahedral sites.

StructureAtoms per cellTetrahedral sitesOctahedral sites
FCC484
HCP6126
BCC2126

BCC is not close-packed, so the rule $2n$ / $n$ does not apply to it.

Note on the official material

For BCC, the 3 octahedral sites counted as "6 × ½" are labelled "center"; they are the centres of the 6 faces (each shared by 2 cells), not the centre of the cube.

Tetrahedral and octahedral sites in FCC
Figure 5. Interstitial sites in the FCC cell: tetrahedral sites (left) and octahedral sites (right), with the metal atoms and the interstitial atoms. Copied from Session 3 Slide 31.

§ 9 Atomic positions [Session 3 Slide 37]

In an orthogonal lattice (all angles 90°), the three lattice vectors define every point of the cell without ambiguity, so they can be used to give the position of each atom. In a cubic cell ($a = b = c$) a position is written as three coordinates, each a fraction of the edge between 0 and 1:

$$\text{position} = (x, y, z), \qquad 0 \le x, y, z \le 1$$

The origin is $(0,0,0)$; the corner along $y$ is $(0,1,0)$; the top corner above the origin is $(0,0,1)$; the corner opposite the origin is $(1,1,1)$; the centre of the cube is $(\tfrac12, \tfrac12, \tfrac12)$.

Atomic positions in a cubic cell
Figure 6. Atomic positions in a cubic cell, with the corners (0,0,1), (0,1,1) and (0,1,0) labelled. Copied from Session 3 Slide 37.

§ 10 Directions in the cubic cell [Session 3 Slides 39-42]

10.1 What a direction is

A crystallographic direction is a vector joining two points of the lattice. It is written with square brackets and no commas, $[u\,v\,w]$. The indices $u, v, w$ are the components of the vector along the three axes, reduced to the smallest integers that keep the same ratio.

Two vectors that are parallel have the same indices: a direction does not depend on where it starts, only on where it points. For that reason any direction can be drawn starting at a corner of the cell.

Negative components are written with a bar over the number: $[\bar u\,v\,w]$ means $u$ is negative. ($[\bar1\,1\,0]$ is read "minus one, one, zero".)

10.2 How to draw a direction from its indices

  1. Choose the origin. Normally $O = (0,0,0)$. If an index is negative, use a new origin $O'$ (see 10.3).
  2. Scale the indices so that the largest becomes 1. This keeps the arrow inside the cell. Example: $[0\,1\,2] \to \tfrac12[0\,1\,2] = [0, \tfrac12, 1]$.
  3. Locate the tip: start at the origin and move those amounts along $x$, $y$ and $z$.
  4. Draw the arrow from the origin to the tip.

Example 1: $[1\,0\,0]$, $[1\,1\,0]$, $[1\,1\,1]$. The largest index is already 1, so the tips are directly $(1,0,0)$, $(1,1,0)$ and $(1,1,1)$: a cube edge, a face diagonal and the body diagonal.

Directions drawn in the cubic cell
Figure 7. Directions drawn from the origin of the cubic cell: [100] and [110] (first cube), [111], [210] and [-110] (bar over the first 1). Copied from Session 3 Slide 39.

Example 2: $[0\,1\,2]$.

  1. Origin $O = (0,0,0)$ (no negative indices).
  2. Largest index 2, divide by 2: $[0, \tfrac12, 1]$.
  3. Tip at $x = 0$, $y = \tfrac12$, $z = 1$: the middle of the top edge on the back face.
  4. Arrow from $(0,0,0)$ to $(0, \tfrac12, 1)$.

Example 3: $[2\,1\,0]$. Divide by 2: $[1, \tfrac12, 0]$. The arrow goes from the origin to the middle of the bottom edge on the front face, $(1, \tfrac12, 0)$. It lies in the bottom face because $w = 0$. It is the third small cube of the Slide 39 figure shown under Example 1.

Example 4: $[1\,2\,2]$. Divide by 2: $[\tfrac12, 1, 1]$. The tip is at the middle of the top edge on the right face, $(\tfrac12, 1, 1)$.

Direction [012]
Figure 8. Example 1 of the slide: the direction [012] reduced to ½[0 1 2] = [0, ½, 1], drawn from the origin; the red arrow is the direction [110]. Copied from Session 3 Slide 39.
x y z [122] ½[1 2 2] → tip at (½, 1, 1)
Figure 9. The direction [1 2 2] reduced to ½[1 2 2] = [½, 1, 1] and drawn from the origin. Portal drawing, not in the slides (EXTRA)..

10.3 Directions with negative indices

If an index is negative, the arrow would leave the cell when drawn from $(0,0,0)$. To keep it inside, move the origin: the new origin $O'$ has a 1 where the index is negative and a 0 where it is positive (or zero). Then draw the arrow from $O'$ as usual.

DirectionNew origin $O'$Tip ($O'$ + scaled indices)
$[\bar1\,1\,1]$$(1,0,0)$$(1-1,\ 0+1,\ 0+1) = (0,1,1)$
$[1\,\bar1\,0]$$(0,1,0)$$(0+1,\ 1-1,\ 0) = (1,0,0)$
$[0\,2\,\bar1]$$(0,0,1)$scaled $[0, 1, -\tfrac12]$: $(0,\ 1,\ 1-\tfrac12) = (0,1,\tfrac12)$

Example 5: $[\bar1\,2\,2]$. The $u$ index is negative, so $O' = (1,0,0)$. Scale by ½: $[-\tfrac12, 1, 1]$. Tip $= (1-\tfrac12,\ 0+1,\ 0+1) = (\tfrac12, 1, 1)$. The arrow goes from the front-bottom-left corner $(1,0,0)$ up to the middle of the top edge on the right face.

Directions with negative indices
Figure 10. The directions [-111], [1-10] and [02-1] (bar over the negative index) drawn in one cube, each from its new origin O' = (1,0,0), (0,1,0) and (0,0,1). Copied from Session 3 Slide 40.

10.4 How to read the indices of a drawn direction

  1. Take a vector along the direction that passes through the origin (or any parallel copy of it: parallel vectors have the same indices).
  2. Measure its projections on the three axes, in units of the cell edges $a$, $b$, $c$. In practice: coordinates of the head minus coordinates of the tail.
  3. Multiply or divide by a common factor to reach the smallest integers.
  4. Write them in square brackets without commas, with a bar over any negative number.

Example 6 [Session 3 Slide 41]. A vector goes from the tail $(0,1,0)$ to the head $(1,0,1)$. Head − tail $= (1-0,\ 0-1,\ 1-0) = (1, -1, 1)$. They are already the smallest integers, so the direction is $[1\,\bar1\,1]$.

Example 7 [Session 3 Slide 41]. A vector goes from $(0,1,1)$ to $(0,0,0)$. Head − tail $= (0, -1, -1)$: the direction is $[0\,\bar1\,\bar1]$. Its opposite, from $(0,0,0)$ to $(0,1,1)$, is $[0\,1\,1]$: reversing an arrow changes the sign of every index.

Example 8. A vector goes from $(0,0,0)$ to $(\tfrac12, 1, 0)$. Head − tail $= (\tfrac12, 1, 0)$. Multiply by 2 to clear the fraction: $[1\,2\,0]$.

Reading the indices of a drawn direction: [1-11]
Figure 11. A vector from (0,1,0) to (1,0,1): head minus tail gives (1, -1, 1), the direction [1-11]. Copied from Session 3 Slide 41.
Reading the indices of a drawn direction: [0-1-1]
Figure 12. A vector from (0,1,1) to the origin: head minus tail gives (0, -1, -1), the direction [0-1-1]. Copied from Session 3 Slide 41.

10.5 Families of directions

Directions that are equivalent in the lattice (the atoms are spaced the same way along them) form a family, written with angle brackets $\langle u\,v\,w\rangle$. In a cube:

The family <111>
Figure 13. The four body diagonals of the cube, the family <111>. Copied from Session 3 Slide 42.
x y z the cube edges <1 0 0>
Figure 14. The cube edges [100], [010] and [001] of the family <100>. Portal drawing, not in the slides (EXTRA)..

§ 11 Planes in the cubic cell [Session 3 Slides 44-50]

11.1 Miller indices of a plane

A crystallographic plane is described by its Miller indices $(h\,k\,l)$, written in round brackets and no commas. They are the reciprocals of the points where the plane cuts the axes $x$, $y$ and $z$ (measured in units of $a$, $b$, $c$):

$$h = \frac{1}{\text{intercept on } x}, \qquad k = \frac{1}{\text{intercept on } y}, \qquad l = \frac{1}{\text{intercept on } z}$$

A plane parallel to an axis never cuts it: its intercept is $\infty$, and $1/\infty = 0$. So a zero index always means "parallel to that axis".

Example: a plane that cuts $x$ at 1 and $y$ at 1 and is parallel to $z$ has intercepts $(1, 1, \infty)$, reciprocals $(1, 1, 0)$: it is the plane $(1\,1\,0)$.

11.2 How to find the indices of a drawn plane

  1. Choose the origin $O$. The plane must not pass through it; if it does, use the nearest other corner.
  2. Find the intercepts of the plane with the three axes.
  3. Take their reciprocals.
  4. Reduce them to the smallest integers with the same ratio.
  5. Write them in parentheses without commas: $(h\,k\,l)$.
Intercepts $x, y, z$ReciprocalsPlane
$1,\ 1,\ 1$$1,\ 1,\ 1$$(1\,1\,1)$
$1,\ 1,\ \infty$$1,\ 1,\ 0$$(1\,1\,0)$
$1,\ \infty,\ \infty$$1,\ 0,\ 0$$(1\,0\,0)$
$\tfrac12,\ \infty,\ \infty$$2,\ 0,\ 0$$(2\,0\,0)$
$\tfrac12,\ 1,\ \infty$$2,\ 1,\ 0$$(2\,1\,0)$
$\infty,\ 1,\ 1$$0,\ 1,\ 1$$(0\,1\,1)$

Example 1: $(1\,0\,0)$, $(1\,1\,0)$, $(1\,1\,1)$.

x y z (100) intercepts x = 1, y = ∞, z = ∞
Figure 15. Plane (100): intercept x = 1, parallel to y and z, the front face of the cube. Portal drawing, not in the slides (EXTRA)..
Plane (110)
Figure 16. Plane (110): intersections a = 1, b = 1, c = ∞. Copied from Session 3 Slide 45.
Plane (111)
Figure 17. Plane (111): intersections a = 1, b = 1, c = 1. Copied from Session 3 Slide 45.

Example 2: $(2\,0\,0)$, $(2\,1\,0)$, $(0\,1\,1)$.

Plane (200)
Figure 18. Plane (200): intersections a = ½, b = ∞, c = ∞. Copied from Session 3 Slide 45.
x y z (210) x = ½, y = 1, z = ∞ x y z (011) x = ∞, y = 1, z = 1
Figure 19. Planes (210) and (011) in the cubic cell, with their intercepts. Portal drawing, not in the slides (EXTRA)..

11.3 How to draw a plane from its indices

  1. Choose the origin: $O = (0,0,0)$, or $O'$ if any index is negative (see 11.4).
  2. Take the reciprocals of the indices, $1/h$, $1/k$, $1/l$: these are the intercepts.
  3. Mark the intercepts on the axes ($1/0 = \infty$ means the plane does not cut that axis: it is parallel to it).
  4. Join the intercepts to draw the plane.

Example 3: $(0\,1\,1)$.

  1. No negative indices: $O = (0,0,0)$.
  2. Intercepts: $x = 1/0 = \infty$, $y = 1/1 = 1$, $z = 1/1 = 1$.
  3. The plane is parallel to $x$ and passes through $(0,1,0)$ and $(0,0,1)$.
  4. Because it is parallel to $x$, slide that line along $x$: the plane is the rectangle with corners $(0,1,0)$, $(0,0,1)$, $(1,0,1)$, $(1,1,0)$.

11.4 Planes with negative indices

To draw a plane with a negative index inside the cell, move the origin in the direction of that index: as with directions, $O'$ has a 1 in the position of each negative index and a 0 elsewhere. The intercepts are then measured from $O'$, and a negative intercept means going backwards along that axis.

PlaneNew origin $O'$Intercepts from $O'$Result
$(\bar1\,0\,0)$$(1,0,0)$$x = -1$, $y = \infty$, $z = \infty$the back face $x = 0$
$(1\,\bar1\,0)$$(0,1,0)$$x = 1$, $y = -1$, $z = \infty$vertical diagonal plane through the edges at $(0,0)$ and $(1,1)$
$(1\,\bar1\,\bar1)$$(0,1,1)$$x = 1$, $y = -1$, $z = -1$triangle through $(1,1,1)$, $(0,0,1)$ and $(0,1,0)$
$(\bar2\,1\,1)$$(1,0,0)$$x = -\tfrac12$, $y = 1$, $z = 1$triangle through $(\tfrac12,0,0)$, $(1,1,0)$ and $(1,0,1)$

Example 4: $(1\,\bar1\,\bar1)$, step by step.

  1. Negative indices in $y$ and $z$: $O' = (0,1,1)$.
  2. Intercepts from $O'$: $x = 1/1 = 1$, $y = 1/(-1) = -1$, $z = 1/(-1) = -1$.
  3. From $O' = (0,1,1)$: one unit forward in $x$ gives $(1,1,1)$; one unit back in $y$ gives $(0,0,1)$; one unit back in $z$ gives $(0,1,0)$.
  4. The plane is the triangle joining those three points.

Example 5: $(\bar2\,1\,1)$, step by step.

  1. Negative index in $x$: $O' = (1,0,0)$.
  2. Intercepts from $O'$: $x = 1/(-2) = -\tfrac12$, $y = 1$, $z = 1$.
  3. From $(1,0,0)$: half a unit back in $x$ gives $(\tfrac12,0,0)$; one unit in $y$ gives $(1,1,0)$; one unit in $z$ gives $(1,0,1)$.
  4. Join the three points.
Planes with negative indices
Figure 20. Planes (-100), (1-10), (1-1-1) and (-211) with their new origins O' and intersections; the slide prints the last one as (2 1 1), but its intersection a = −½ shows the bar over the 2. Copied from Session 3 Slide 49.

11.5 Families of planes

$$(100)(010)(001)(\bar100)(0\bar10)(00\bar1) = \{100\}\quad\text{(the six cube faces)}$$
$$(111)(\bar111)(1\bar11)(11\bar1)(\bar1\bar11)(\bar11\bar1)(1\bar1\bar1)(\bar1\bar1\bar1) = \{111\}$$
The family {100}
Figure 21. The six cube faces, one shaded in each cube, form the family {100}. Copied from Session 3 Slide 50.
x y z O′ two members of {1 1 1}
Figure 22. The planes (111) and (-111), two of the eight members of the family {111}. Portal drawing, not in the slides (EXTRA)..

11.6 Directions and planes with the same indices

In cubic lattices, a direction is perpendicular to the plane with the same indices: $[1\,1\,1] \perp (1\,1\,1)$, $[1\,0\,0] \perp (1\,0\,0)$, $[1\,1\,0] \perp (1\,1\,0)$. This only holds for cubic cells.

x y z [111] (111)
Figure 23. In a cubic cell the direction [111] is perpendicular to the plane (111). Portal drawing of the rule stated on Session 3 Slide 45, not in the slides (EXTRA)..

§ 12 Directions and planes in the hexagonal cell [Session 3 Slides 51-52]

The hexagonal cell is described with four axes: three in the basal plane, $a_1$, $a_2$, $a_3$, at 120° from each other, and the vertical axis $c$. Indices therefore have four numbers: $h$, $k$, $i$ for the three basal axes and $l$ for $c$.

Planes. Only three of the four indices are independent, because the three basal axes are related:

$$i = -(h + k)$$

So a plane written with three indices $(h'\,k'\,l')$ becomes $(h\,k\,i\,l)$ by adding $i$. Example: $(0\,1\,0) \equiv (0\,1\,\bar1\,0)$, since $i = -(0 + 1) = -1$. This plane is a side face of the prism, parallel to $a_1$ and to $c$.

Axes of the hexagonal cell
Figure 24. Hexagonal cell with the axes a1, a2, a3 (120° apart in the base) and c, and the shaded side face (01-10) ≡ (010). Copied from Session 3 Slide 51.

Directions. A direction $[h'\,k'\,l']$ written in three indices becomes $[h\,k\,i\,l]$ with

$$h = \tfrac13(2h' - k'), \qquad k = \tfrac13(2k' - h'), \qquad i = -\tfrac13(h' + k'), \qquad l = l'$$

and the result is then reduced to integers.

Example: $[1\,0\,0]$. $h = \tfrac13(2\cdot1 - 0) = \tfrac23$, $k = \tfrac13(2\cdot0 - 1) = -\tfrac13$, $i = -\tfrac13(1 + 0) = -\tfrac13$, $l = 0$. Multiplying by 3: $[1\,0\,0] \equiv [2\,\bar1\,\bar1\,0]$, along $a_1$.

Example: $[1\,1\,0]$. $h = \tfrac13(2 - 1) = \tfrac13$, $k = \tfrac13(2 - 1) = \tfrac13$, $i = -\tfrac13(1 + 1) = -\tfrac23$, $l = 0$. Multiplying by 3: $[1\,1\,0] \equiv [1\,1\,\bar2\,0]$.

Example: $[0\,0\,1]$. $h = k = i = 0$ and $l = 1$: $[0\,0\,0\,1]$, along $c$.

Directions in the hexagonal cell
Figure 25. Directions in the hexagonal cell: [2-1-10] along a1 (= [100] in three indices), [11-20] (= [110]), [0001] along c, and [1-211]; the short arrows labelled [100] and [110] are the three-index vectors. Copied from Session 3 Slide 52.
Note on the official material

The conversion rule for the fourth index is printed as $l = -l'$. With that sign, the vertical direction $[0\,0\,1]$ would become $[0\,0\,0\,\bar1]$, pointing down, while the same material labels it $[0\,0\,0\,1]$. The rule is $l = l'$.

§ 13 Spacing between parallel planes [Session 3 Slide 53]

The interplanar spacing $d_{hkl}$ is the distance between two neighbouring parallel planes $(h\,k\,l)$. It depends on the Miller indices and on the lattice parameters.

Cubic crystals:

$$d_{hkl} = \frac{a}{\sqrt{h^2 + k^2 + l^2}}$$

Example: for $(1\,1\,0)$, $d_{110} = a/\sqrt{1 + 1 + 0} = a/\sqrt2$.

Orthogonal crystals in general (all angles 90°, edges $a$, $b$, $c$):

$$\frac{1}{d_{hkl}^2} = \frac{h^2}{a^2} + \frac{k^2}{b^2} + \frac{l^2}{c^2}$$

With $a = b = c$ this reduces to the cubic formula.

Spacing between (110) planes
Figure 26. Parallel (110) planes seen along z, with their spacing d110. Copied from Session 3 Slide 53.

Part II: Densities and crystalline defects

§ 14 Atomic densities [Session 4 Slides 7-9]

14.1 Volume density [Session 4 Slide 7]

$$\rho = \frac{\text{mass of the unit cell}}{\text{volume of the unit cell}} = \frac{n\cdot M}{V_{cell}\cdot N_A}$$

$n$ = atoms per cell, $M$ = atomic weight, $V_{cell}$ = volume of the cell, $N_A$ = Avogadro's number ($6.023\times10^{23}$ atoms/mol). Dividing only the number of atoms by the volume gives the number of atoms per unit volume, $\rho_V = n/V_{cell}$.

Example: copper (FCC), $M = 63.5$ g/mol, $R = 1.28$ Å $= 1.28\times10^{-8}$ cm.

14.2 Linear density [Session 4 Slide 8]

$$\rho_{[uvw]} = \frac{\text{number of atoms whose centre lies on the line}}{\text{length of the line}}\quad[\text{atoms/m}]$$

Only atoms cut through their centre by the line count, and an atom at the end of the segment counts ½ (it is shared with the next segment).

Example: Cu (FCC), direction $[1\,0\,0]$. The cube edge has an atom at each end: 2 × ½ = 1 atom, over the length $a = 2\sqrt2\,R$:

$$\rho_{[100]} = \frac{1\ \text{atom}}{2\sqrt2\cdot1.28\times10^{-8}\ \text{cm}} = 2.76\times10^{7}\ \text{atoms/cm}$$

Example: Cu (FCC), direction $[1\,1\,1]$. The body diagonal has length $\sqrt3\,a$. In an FCC cell no atom sits at the centre of the cube, so the only atoms on this line are the two corners: 2 × ½ = 1 atom:

$$\rho_{[111]} = \frac{1\ \text{atom}}{\sqrt3\,a}$$
Note on the official material

The official example counts "1 central atom + 2 · ½ vertices" along $[1\,1\,1]$ in FCC copper, giving $2/(\sqrt3\,a) = 2\sqrt3/(3a)$. A centre atom only exists in BCC. In FCC the face-centre atoms (such as $(\tfrac12, \tfrac12, 0)$) do not lie on the line $(t, t, t)$, so the count is 1 atom and $\rho_{[111]} = 1/(\sqrt3\,a)$.

14.3 Planar density [Session 4 Slide 9]

$$\rho_{(hkl)} = \frac{\text{number of atoms whose centre lies on the plane}}{\text{area of the plane}}\quad[\text{atoms/m}^2]$$

Example: Cu (FCC), plane $(1\,0\,0)$. The plane is a cube face: 1 atom at its centre plus 4 corners, each shared by 4 faces (¼ each): 1 + 4 · ¼ = 2 atoms, over the area $a^2$:

$$\rho_{(100)} = \frac{2\ \text{atoms}}{a^2} = \frac{2}{(2\sqrt2\cdot1.28\times10^{-8}\ \text{cm})^2} = 1.5\times10^{15}\ \text{atoms/cm}^2$$
The (100) plane of FCC copper
Figure 27. Planar density in FCC copper: the (100) face holds one central atom and four corner atoms shared by four faces, 2 atoms over the area a². Copied from Session 4 Slide 9.
[1 0 0] in FCC 2 × ½ = 1 atom over length a [1 1 1] in FCC only 2 × ½ = 1 atom over √3·a
Figure 28. Atoms whose centres lie on the [100] edge and on the [111] body diagonal of an FCC cell: 2 × ½ = 1 atom in both cases (no atom at the cube centre in FCC, see the note on the official material). Portal drawing, not in the slides (EXTRA)..

14.4 Summary of the three structures [Session 4 Slide 10]

FCCBCCHCP
Atoms per cell$8\cdot\tfrac18 + 6\cdot\tfrac12 = 4$$8\cdot\tfrac18 + 1 = 2$$12\cdot\tfrac16 + 2\cdot\tfrac12 + 3 = 6$
Atomic radius$r = \dfrac{a}{2\sqrt2}$$r = \dfrac{a\sqrt3}{4}$$r = \tfrac12\sqrt{\dfrac{a^2}{3} + \dfrac{c^2}{4}}$
Cell volume$a^3$$a^3$$a^2 c\,\dfrac{3\sqrt3}{2}$
Empty volume $V_h = V_c - V_{atoms}$$5.872\,r^3$$3.94\,r^3$$8.808\,r^3$
Empty fraction $V_h/V_c$0.260.320.26
Octahedral sites4, radius $\tfrac12(a - 2r) = 0.41\,r$6, radius $\tfrac12(a - 2r) = 0.15\,r$6, radius $0.41\,r$
Tetrahedral sites8, radius $\left(\sqrt{\tfrac32} - 1\right)r = 0.225\,r$12, radius $\left(\sqrt{\tfrac53} - 1\right)r = 0.291\,r$12, radius $0.225\,r$
Close-packed planes$\{111\}$: 4$\{110\}$: 6$\{0001\}$: 1
Close-packed directions$\langle110\rangle$: 6$\langle111\rangle$: 4$\langle2\bar1\bar10\rangle$: 3

The empty fractions match the packing factors: $1 - 0.74 = 0.26$ and $1 - 0.68 = 0.32$.

§ 15 Perfect and real crystals [Session 4 Slides 12-13]

A perfect crystal has every atom in its regular position, the same spacing $d_{hkl}$ everywhere for each set of planes, and no missing or extra atoms. It could only exist at 0 K.

A real crystal is different: its atoms vibrate, some positions are empty (vacancies), some atoms are displaced from their ideal positions, and these defects change the properties.

A crystalline defect is any disruption of the perfect periodic arrangement of atoms or ions.

Why defects always exist. Creating a defect costs energy (its enthalpy $\Delta H$ is positive) but it also increases disorder (its entropy $\Delta S$ is positive). The crystal settles where its Gibbs energy is lowest:

$$\Delta G = \Delta H - T\Delta S$$

A defect forms when it lowers $G$, that is when $\Delta G < 0$, or $\Delta H < T\Delta S$. The equilibrium number of defects is where $G$ stops decreasing ($\Delta G = 0$): beyond it, extra defects would cost more energy than the disorder they add.

The entropy comes from the number of ways $W$ of placing $n$ defects among $N$ possible positions:

$$S = k_B \ln W, \qquad W = \frac{N!}{(N - n)!\,n!}$$

with $k_B = 1.380622\times10^{-23}$ J/K, the Boltzmann constant. In one mole of a crystal there are $N_A$ possible places for a single vacancy, so $W \sim 10^{23}$ even for one defect: the gain in entropy is large. Because the $T\Delta S$ term grows with temperature, the number of defects increases with temperature.

§ 16 Types of defects [Session 4 Slide 14]

Defects are classified by their dimension:

TypeExamples
1. Point (0-D)vacancies, interstitials, substitutional atoms, Schottky and Frenkel defects, order-disorder
2. Linear (1-D): dislocationsedge (Taylor), screw (Burgers), mixed
3. Planar or extended (2-D)external surfaces, grain boundaries, twin boundaries, stacking faults
4. Complex (volume, 3-D)clusters

§ 17 Point defects [Session 4 Slides 15-18]

17.1 What they are [Session 4 Slide 15]

Point defects are local disruptions of an otherwise perfect arrangement. Although they occupy only one or two sites, their effect is felt over much larger distances in the crystal. They appear when atoms move after gaining energy (for example, when the temperature rises). Foreign atoms can also act as point defects:

17.2 Vacancies and interstitials [Session 4 Slide 16]

The number of vacancies at temperature $T$ is

$$n_v = N\exp\left(-\frac{\Delta H_v}{RT}\right), \qquad N = \frac{N_A\,\rho}{M}$$
SymbolMeaning
$n_v$vacancies per cm³
$N$lattice points per cm³
$\Delta H_v$energy to form one mole of vacancies (J/mol)
$n_v/N$fraction of vacant sites
$N_A$, $M$, $\rho$Avogadro's number, atomic (molecular) weight, density

In metals at equilibrium the fraction of vacancies reaches at most about $n_v/N \approx 10^{-4}$.

Vacancies are produced during solidification (impurities, alloys), by bombarding with high-energy particles, during plastic deformation (processing), and simply by heating: higher temperature → stronger thermal vibrations → more vacancies.

17.3 Schottky and Frenkel defects in ionic crystals [Session 4 Slides 17-18]

In an ionic crystal, removing a single ion would leave the crystal electrically charged. Defects therefore come in combinations that keep it neutral:

$$n_S = N\exp\left(-\frac{\Delta H_S}{2RT}\right)$$
$$n_F = \sqrt{N N_i}\,\exp\left(-\frac{\Delta H_F}{2RT}\right)$$

$\Delta H_S$ and $\Delta H_F$ are the energies needed to create each defect, $N$ the number of lattice positions, $N_i$ the number of interstitial positions, $T$ the temperature in kelvin. In a given crystal $\Delta H_S \neq \Delta H_F$, and the defect with the lower energy is the one that forms.

Example: Schottky defects in NaCl and MgO.

Crystal$\Delta H_S$ (J per defect)$n_S$ at 300 K (per mol)$n_S$ at 1000 K (per mol)
NaCl$3.69\times10^{-19}$$2.64\times10^{4}$$9.38\times10^{17}$
MgO$10.57\times10^{-19}$$2.12\times10^{-32}$$1.39\times10^{7}$

Two conclusions: the number of defects rises enormously with temperature, and because $\Delta H_S$(MgO) > $\Delta H_S$(NaCl), vacancies are much harder to create in MgO.

Checking the NaCl value at 300 K

Here the energy is given per defect, so the gas constant $R$ is replaced by the Boltzmann constant $k_B$: $\dfrac{\Delta H_S}{2k_BT} = \dfrac{3.69\times10^{-19}}{2\cdot1.38\times10^{-23}\cdot300} = 44.6$, and $n_S = 6.02\times10^{23}\cdot e^{-44.6} \approx 2.6\times10^{4}$ per mol.

§ 18 Solid solutions [Session 4 Slides 19, 21-24]

18.1 What a solid solution is [Session 4 Slide 21]

When two materials are mixed as liquids and then solidified, the result is either a new phase or a solid solution.

A solid solution is a crystalline solid with two or more elements dispersed as atoms in a structure that has a single phase. There are two kinds:

18.2 Defects created by a solid solution in ionic solids [Session 4 Slide 22]

When MgO and Al₂O₃ form a solid solution, Al³⁺ ions replace Mg²⁺ ions in the MgO crystal. Each Al³⁺ carries one more positive charge than the Mg²⁺ it replaces, so to keep the crystal neutral some cation sites must stay empty. Three Mg²⁺ (total charge +6) are replaced by two Al³⁺ (also +6) and one vacancy:

$$3\,\text{Mg}^{2+} \;\leftrightarrow\; 2\,\text{Al}^{3+} + 1\ \text{vacancy}$$

18.3 Hume-Rothery rules [Session 4 Slide 23]

The conditions for forming a solid solution:

Substitutional solid solutions:

Interstitial solid solutions:

Meeting all the conditions does not guarantee complete solubility; failing one or more of them gives partial solubility.

18.4 Order and disorder [Session 4 Slides 19, 24]

In substitutional solid solutions of elements with similar electronegativities, the two kinds of atoms can be arranged in two ways:

Example: the Cu-Au alloy is disordered above 390 °C and ordered below 390 °C.

§ 19 Linear defects: dislocations [Session 4 Slides 26-37]

19.1 What a dislocation is [Session 4 Slide 26]

A dislocation is a line imperfection in an otherwise perfect crystal. Dislocations can move through the crystal under relatively small forces, and their movement can shift whole crystal planes past each other. They explain:

They form during solidification, during plastic (permanent) deformation, from concentrations of vacancies, and from atomic disorder in solid solutions. There are three types: edge (Taylor), screw (Burgers) and mixed.

19.2 Edge dislocation [Session 4 Slides 29-30]

An edge dislocation is equivalent to inserting an extra half-plane of atoms into the crystal. It distorts the lattice and stores energy.

It is characterized by the Burgers vector b: the magnitude and direction of the lattice distortion caused by the dislocation. To find it, draw a loop around the dislocation taking the same number of atom spacings in each direction; around a perfect crystal the loop would close, but around a dislocation it does not, and b is the vector needed to close it.

For an edge dislocation:

19.3 Screw dislocation [Session 4 Slide 31]

A screw dislocation bends some lines of atoms locally, as if a shear stress had been applied to one part of the crystal. Following the atoms around the dislocation line traces a spiral ramp (a screw). The Burgers vector is again the vector that closes a loop around it, and:

19.4 Mixed dislocation [Session 4 Slide 32]

Real dislocations are usually mixed: they change gradually from edge to screw character, with a transition region between the two. The Burgers vector is the same along the whole dislocation.

19.5 How dislocations move [Session 4 Slides 33-37]

A metal can be deformed plastically only if its dislocations can move; its strength and ductility depend on how easily they move.

Slip is the process by which plastic deformation happens through the motion of dislocations:

Why the densest direction: the energy needed to move a dislocation is $E \propto |\vec b|^2$, and the shortest Burgers vector is found along the close-packed direction. In FCC:

Slip along $[1\,1\,0]$ costs half the energy.

Slip systems of the three structures:

StructureSlip planeSlip directionNumber of slip systems
BCC$\{110\}$$\langle\bar111\rangle$6 × 2 = 12
FCC$\{111\}$$\langle\bar110\rangle$4 × 3 = 12
HCP$(0001)$$\langle2\bar1\bar10\rangle$1 × 3 = 3

Slip happens on the planes of densest packing and along the directions of highest linear density. More slip systems → more ductility: HCP, with only 3 slip systems, has far fewer than BCC or FCC.

§ 20 Planar defects [Session 4 Slides 39-46]

Planar defects are boundaries that separate regions of a material that have the same crystal structure but different orientation. There are four kinds: grain boundaries, twin boundaries, stacking faults and external surfaces.

20.1 Grain boundaries [Session 4 Slides 40-42]

When a liquid cools, its atoms lose kinetic energy, bonds become stable, and atoms start to gather into crystals. Crystals start (nucleate) at many points at the same time and grow until they meet. The solid ends up as a polycrystal: many crystals, or grains, each with its own orientation. The grain boundaries are the surfaces where grains meet.

$$\text{grain size} \downarrow \;\Rightarrow\; \text{mechanical strength} \uparrow$$

20.2 Twin boundaries [Session 4 Slides 43-44]

A twin boundary separates two parts of a crystal whose planes are slightly misoriented (crystal twinning). It is a special grain boundary where one side is the mirror image of the other, with a highly symmetrical interface. Twins form during deformation or heat treatments. They hinder slip, making dislocation movement harder, and so they increase strength. Austenitic stainless steel and brass (70 Cu / 30 Zn) are typical examples.

20.3 Stacking faults [Session 4 Slide 45]

A stacking fault is an interruption of one or two layers in the stacking sequence of close-packed planes. In FCC, the perfect sequence ABC ABC ABC becomes, for example, ABC AB ABC: locally, two layers follow the HCP pattern.

20.4 External surfaces [Session 4 Slide 46]

The external surface is where the crystal or grain ends. Surface atoms have fewer neighbours (a lower coordination number) than atoms inside, so the surface has a higher energy than the interior.

§ 21 Polymorphism and allotropy [Session 4 Slides 49-50]

The same element or compound can take different crystal structures under different conditions of pressure and temperature.

Carbon:

Zirconia (ZrO₂): it changes from monoclinic to tetragonal and then to cubic (CaF₂-type, fcc) as it is heated. The shape change of the transformation consumes energy, and this is used to make the material tougher (tetragonal zirconia polycrystal).

Iron (steels): BCC at room temperature, FCC from 910 °C. This change of structure is what makes it possible to heat treat steels.