Close Packing and Ionic Crystal Structures
CSIR NET chemistry: ccp and hcp, voids, radius ratio rules, NaCl, CsCl, ZnS, CaF₂, rutile, NiAs, perovskite and spinel types, lattice energy and defects.
- 19 min
On this page · 15 sections
- Prerequisites
- The Hook: Oranges, Cannonballs and Rock Salt
- Close Packing of Equal Spheres
- Voids: Tetrahedral and Octahedral Holes
- Radius Ratio Rules
- The Standard Structure Types
- Lattice Energy in Brief
- Defects in Crystals
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- Unit cells, Z, coordination number and packing fraction from Crystal Lattices, Unit Cells and Miller Indices.
- Ionic radii and their trends from Atomic and Ionic Radii. Radius ratios on this page use Shannon-type radii, which depend on coordination number.
- Crystal field stabilisation energy from Crystal Field, Ligand Field and MO Theory of Complexes, needed for normal and inverse spinels.
The Hook: Oranges, Cannonballs and Rock Salt
Close Packing of Equal Spheres
In one layer, each sphere touches six others (layer A). The next layer (B) sits in half the hollows of layer A. For the third layer there are two choices:
| Stacking | Third layer | Name | Unit cell | Example metals |
|---|---|---|---|---|
| ABAB… | directly over A | hexagonal close packing (hcp) | hexagonal, 2 atoms per primitive cell (6 per hexagonal prism) | Mg, Zn, Cd, Ti, Co |
| ABCABC… | over the hollows not used by B (C) | cubic close packing (ccp) | face-centred cubic, Z = 4 | Cu, Ag, Au, Al, Ni, Pt |
Both have coordination number 12 (6 in-plane, 3 above, 3 below) and packing fraction 74.0%. They differ only from the third layer onwards, so their energies are very close; many metals and the noble gases choose between them on small differences.
- In ccp, the close-packed layers are the {111} planes of the fcc cell, stacked along a body diagonal.
- In hcp with touching spheres, \( c/a = \sqrt{8/3} = 1.633 \). Real hcp metals deviate (Zn and Cd have noticeably larger c/a, for example), but 1.633 is the ideal reference.
Voids: Tetrahedral and Octahedral Holes
Between close-packed layers there are two kinds of hole:
| Hole | Surrounded by | Number per N packed spheres | Largest sphere that fits (radius ratio) | Positions in the fcc cell |
|---|---|---|---|---|
| Tetrahedral (T) | 4 spheres | 2N | \( r/R = 0.225 \) | 8 holes at \( (\tfrac14, \tfrac14, \tfrac14) \) and equivalents, inside the cell |
| Octahedral (O) | 6 spheres | N | \( r/R = 0.414 \) | body centre + 12 edge centres (\( 1 + 12 \times \tfrac14 = 4 \)) |
In the fcc cell, Z = 4, so there are 4 octahedral and 8 tetrahedral holes, consistent with N and 2N. The hole sizes follow from simple geometry:
- Octahedral: across the square of four spheres, \( 2R + 2r = \sqrt2 \times 2R \), so \( r/R = \sqrt2 - 1 = 0.414 \).
- Tetrahedral: the centre of a tetrahedron of spheres of radius R lies \( \sqrt{3/2}\,R \) from each, so \( r/R = \sqrt{3/2} - 1 = 0.225 \).
- Cubic (in a simple cubic array, not a close packing): along the body diagonal \( 2R + 2r = \sqrt3 \times 2R \), so \( r/R = \sqrt3 - 1 = 0.732 \).
- Triangular: \( r/R = 2/\sqrt3 - 1 = 0.155 \).
Radius Ratio Rules
For an ionic solid built of cations (radius \( r_+ \)) and anions (\( r_- \)), the cation wants as many anion neighbours as possible while still touching them. A cation that rattles in its hole is unstable, so the cation takes the highest coordination number whose limiting ratio it exceeds.
| \( r_+/r_- \) | Coordination number | Geometry | Typical structure |
|---|---|---|---|
| < 0.155 | 2 | linear | — |
| 0.155–0.225 | 3 | trigonal planar | B₂O₃ (B in BO₃ units) |
| 0.225–0.414 | 4 | tetrahedral | ZnS (zinc blende, wurtzite) |
| 0.414–0.732 | 6 | octahedral | NaCl, rutile |
| 0.732–1.000 | 8 | cubic | CsCl, CaF₂ |
| ≥ 1 | 12 | cuboctahedral | (close-packed, equal-sized ions) |
Worked check. With radii for the appropriate coordination numbers: NaCl, \( 102/181 = 0.56 \) (octahedral, correct); CsCl, \( 174/181 = 0.96 \) (cubic, correct); ZnS, \( 60/184 = 0.33 \) (tetrahedral, correct).
The Standard Structure Types
AB structures
| Type | Anion array | Cations in | CN (cation:anion) | Z | Examples |
|---|---|---|---|---|---|
| Rock salt (NaCl) | ccp | all O holes | 6:6 | 4 | most alkali halides, MgO, CaO, FeO, NiO, AgCl, AgBr |
| Caesium chloride | simple cubic | the cube centre | 8:8 | 1 | CsCl, CsBr, CsI, NH₄Cl (one form) |
| Zinc blende (sphalerite) | ccp | half the T holes (alternate) | 4:4 | 4 | ZnS, CuCl, CdS (one form), GaAs, β-SiC |
| Wurtzite | hcp | half the T holes | 4:4 | 2 (hexagonal cell) | ZnS (other form), ZnO, BeO, AlN, GaN |
| Nickel arsenide (NiAs) | hcp | all O holes | 6:6 (Ni octahedral; As trigonal prismatic) | 2 | NiAs, FeS, CoS, NiS |
Two relationships worth remembering:
- Rock salt is two interpenetrating fcc lattices, offset by \( a/2 \) along an edge. Zinc blende is two interpenetrating fcc lattices offset by a quarter of the body diagonal; replace both kinds of atom by carbon and you have diamond.
- NaCl and NiAs both fill all octahedral holes; the difference is ccp versus hcp anions. In NiAs the filled octahedra share faces along c, so metal atoms are close together and the compounds are often metallic.
AB₂ structures
| Type | Description | CN | Z | Examples |
|---|---|---|---|---|
| Fluorite (CaF₂) | Ca²⁺ ccp, F⁻ in all T holes | 8:4 | 4 | CaF₂, SrF₂, BaF₂, UO₂, ThO₂, CeO₂ |
| Antifluorite | anions ccp, cations in all T holes | 4:8 | 4 | Li₂O, Na₂O, K₂O, Na₂S |
| Rutile (TiO₂) | Ti in half the O holes of a distorted hcp O array; tetragonal | 6:3 | 2 | TiO₂, SnO₂, MnO₂, PbO₂, MgF₂ |
| Cadmium iodide (CdI₂) | I⁻ hcp, Cd²⁺ in all O holes of alternate layers | 6:3 | 1 | CdI₂, Mg(OH)₂, many MX₂ with polarisable anions |
| Cadmium chloride (CdCl₂) | the ccp analogue of CdI₂ | 6:3 | — | CdCl₂, MgCl₂ |
Fluorite can also be seen the other way: F⁻ ions form a simple cubic array, and Ca²⁺ occupy alternate cube centres. That is why the fluorite cation is 8-coordinate, like Cs⁺ in CsCl but with half the cubes empty. Those empty cubes give the fluorite structure room for interstitial anions (the anti-Frenkel defects below), and the open anion sublattice lets anions move easily. Yttria-doped ZrO₂ and gadolinia-doped CeO₂ are oxide-ion conductors: the aliovalent dopant creates oxide-ion vacancies, and the oxide ions hop into them.
Layered CdI₂ and CdCl₂ structures (van der Waals gaps between MX₂ sheets) appear when the anion is large and polarisable and the bonding has substantial covalent character.
Structures with two kinds of cation
| Type | Description | Examples |
|---|---|---|
| Corundum (α-Al₂O₃) | O²⁻ hcp; Al³⁺ in 2/3 of the O holes | Al₂O₃, Cr₂O₃, α-Fe₂O₃ |
| Ilmenite (FeTiO₃) | corundum with Fe and Ti in alternate layers | FeTiO₃ |
| Perovskite (ABX₃) | B at cube corners (octahedrally coordinated by X at edge centres), large A at the body centre (12-coordinate); equivalently, A and X together form a ccp array with B in ¼ of the O holes | CaTiO₃, SrTiO₃, BaTiO₃, KNbO₃ |
| Spinel (AB₂O₄) | O²⁻ ccp; A in 1/8 of the T holes, B in 1/2 of the O holes | MgAl₂O₄, Co₃O₄, Mn₃O₄, Fe₃O₄ (inverse) |
Perovskite geometry. In cubic SrTiO₃ (\( a \approx 390.5 \) pm), Ti–O \( = a/2 \approx 195 \) pm and Sr–O \( = a/\sqrt2 \approx 276 \) pm. For the ideal cubic structure \( r_A + r_X = \sqrt2(r_B + r_X) \), which defines the Goldschmidt tolerance factor
$$t = \frac{r_A + r_X}{\sqrt2\,(r_B + r_X)}$$Values near 1 give cubic perovskites; smaller values lead to tilted, lower-symmetry variants. BaTiO₃ distorts at room temperature (the Ti⁴⁺ ion is off-centre), which is the origin of its ferroelectricity.
Spinels: normal versus inverse
In a cubic close-packed array of 32 oxide ions (the cubic spinel cell, \( Z = 8 \)) there are 64 T and 32 O holes. Spinel fills 8 T and 16 O.
| Spinel | T holes hold | O holes hold | Written as | Examples |
|---|---|---|---|---|
| Normal | all A²⁺ | all B³⁺ | \( \text{A}[\text{B}_2]\text{O}_4 \) | MgAl₂O₄, Mn₃O₄, Co₃O₄, ZnFe₂O₄ |
| Inverse | half the B³⁺ | A²⁺ + other half of B³⁺ | \( \text{B}[\text{AB}]\text{O}_4 \) | Fe₃O₄, NiFe₂O₄, CoFe₂O₄ |
The choice is driven largely by octahedral site preference energy (OSPE), the difference between octahedral and tetrahedral CFSE (with \( \Delta_t \approx \tfrac49\Delta_o \)):
- Fe₃O₄: Fe³⁺ is high-spin d⁵ and has zero CFSE anywhere; Fe²⁺ (d⁶) gains CFSE in an octahedron. So Fe²⁺ takes O sites and pushes half the Fe³⁺ to T sites: inverse.
- Mn₃O₄: Mn³⁺ (high-spin d⁴) has a large octahedral preference (and is Jahn–Teller distorted); Mn²⁺ (d⁵) has none. Mn³⁺ stays octahedral: normal.
- Co₃O₄: Co³⁺ is low-spin d⁶ in oxide, with a very large octahedral preference: normal.
- NiFe₂O₄: Ni²⁺ (d⁸) has a strong octahedral preference; Fe³⁺ has none: inverse.
Lattice Energy in Brief
The lattice energy U is the energy released when gaseous ions come together to form one mole of the solid. The Born–Landé equation sums the Coulomb energy over the whole lattice (through the Madelung constant M) and adds a short-range repulsion (through the Born exponent n):
$$\boxed{U = \frac{N_A M |z_+ z_-| e^2}{4\pi\varepsilon_0 r_0}\left(1 - \frac{1}{n}\right)}$$With \( N_Ae^2/4\pi\varepsilon_0 = 1.389 \times 10^5 \) kJ mol⁻¹ pm and \( r_0 \) in pm:
| Structure type | Madelung constant M |
|---|---|
| Rock salt | 1.748 |
| Caesium chloride | 1.763 |
| Zinc blende | 1.638 |
| Wurtzite | 1.641 |
| Fluorite | 2.519 |
Born exponents rise with the size of the ion’s noble-gas core: He 5, Ne 7, Ar 9, Kr 10, Xe 12; for a salt take the average of the two ions.
Worked example: NaCl. \( r_0 = 282 \) pm, \( M = 1.748 \), \( n = (7 + 9)/2 = 8 \):
$$U = \frac{1.389 \times 10^5 \times 1.748}{282}\left(1 - \tfrac18\right) \approx 753\ \text{kJ mol}^{-1}$$This is a few per cent below the value from a Born–Haber cycle, which is typical for alkali halides. MgO has the same structure, \( |z_+z_-| = 4 \) and a shorter \( r_0 \approx 210 \) pm, giving about 3960 kJ mol⁻¹: more than five times NaCl. That factor explains the very high melting point and hardness of MgO.
What matters most for Part B: U grows with charge product first and with smaller \( r_0 \) second. The CsCl structure’s slightly larger M does not compensate for the loss of the shorter contacts unless the cation is large, which is another way of reading the radius ratio rules.
Defects in Crystals
A real crystal above 0 K always contains point defects, because creating a few vacancies costs enthalpy but gains configurational entropy.
| Defect | What happens | Effect on density | Favoured when | Examples |
|---|---|---|---|---|
| Schottky | equal numbers of cation and anion vacancies (keeping neutrality) | decreases | ions of similar size, high coordination | NaCl, KCl, KBr, CsCl |
| Frenkel | an ion (usually the smaller cation) leaves its site for an interstitial site | unchanged | large size difference, low coordination, polarisable lattice | AgCl, AgBr, AgI, ZnS |
| Anti-Frenkel | anion moves to an interstitial site | unchanged | fluorite structure (empty cubes available) | CaF₂, UO₂ |
AgBr shows both Schottky and Frenkel defects. The equilibrium number of defects follows a Boltzmann-type law:
$$\frac{n_S}{N} \approx \exp\left(-\frac{\Delta H_S}{2kT}\right), \qquad n_F \approx \sqrt{NN_i}\,\exp\left(-\frac{\Delta H_F}{2kT}\right)$$where \( \Delta H_S \) creates one cation–anion vacancy pair and \( N_i \) is the number of interstitial sites. The factor 2 in the exponent appears because each defect event creates two independent species. For \( \Delta H_S = 2.0 \) eV at 1000 K, \( n_S/N \approx \exp(-11.6) \approx 9 \times 10^{-6} \); at 800 K it is about \( 5 \times 10^{-7} \). The numbers are small, but they are what make ionic conduction and diffusion in solids possible.
Non-stoichiometric defects
| Type | Mechanism | Example | Consequence |
|---|---|---|---|
| Metal deficiency | cation vacancies, charge balanced by higher oxidation states | Fe₁₋ₓO (wüstite), Ni₁₋ₓO, Cu₂₋ₓO | p-type semiconduction |
| Metal excess (anion vacancies) | anion sites hold trapped electrons (F-centres) | NaCl heated in Na vapour turns yellow; KCl turns violet | colour, paramagnetism |
| Metal excess (interstitial cations) | extra cations and electrons in interstitial sites | ZnO turns yellow on heating (loses O) | n-type semiconduction |
Wüstite arithmetic. In Fe₀.₉₅O, let x mol of Fe be Fe³⁺: \( 3x + 2(0.95 - x) = 2 \), so \( x = 0.10 \). The fraction of iron present as Fe³⁺ is \( 0.10/0.95 = 10.5\% \). The link from these defects to n-type and p-type behaviour is on the Band Structure of Solids page.
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| ccp (ABC) = fcc lattice | hcp (ABAB), hexagonal lattice; both 74% and CN 12 | |
| Zinc blende: ccp S, half T | wurtzite: hcp S, half T | |
| NaCl: ccp anions, all O | NiAs: hcp anions, all O | |
| Fluorite CaF₂ (8:4) | antifluorite Na₂O (4:8) | |
| Normal spinel A[B₂]O₄ | inverse spinel B[AB]O₄ | |
| Schottky (vacancy pair, density falls) | Frenkel (ion to interstitial, density same) | |
| F-centre: electron in an anion vacancy | interstitial cation (ZnO): excess metal without vacancy | |
| Madelung constant (geometry only) | lattice energy (geometry, charges, \( r_0 \), n) |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- How many octahedral and tetrahedral holes are there per atom in a close-packed array?
- What is the limiting radius ratio for octahedral coordination?
- What are the coordination numbers of Ca²⁺ and F⁻ in fluorite?
Answers
- One octahedral and two tetrahedral holes per packed sphere.
- 0.414 (\( \sqrt2 - 1 \)).
- Ca²⁺ 8, F⁻ 4.
Level 2: Part B Style
Practice Q1. In an oxide, O²⁻ ions are cubic close-packed, cations A occupy one-eighth of the tetrahedral holes and cations B occupy half of the octahedral holes. The formula is:
(a) AB₂O₃ (b) A₂BO₄ (c) AB₂O₄ (d) A₂B₂O₅
Answer: (c) AB₂O₄
Per N oxide ions: T holes 2N, so A = N/4; O holes N, so B = N/2. Ratio A : B : O = 1/4 : 1/2 : 1 = 1 : 2 : 4. This is the spinel description.
Practice Q2. A salt has \( r_+/r_- = 0.55 \). The radius ratio rules predict:
(a) CN 4, zinc blende (b) CN 6, rock salt (c) CN 8, caesium chloride (d) CN 3, trigonal
Answer: (b)
0.55 lies between 0.414 and 0.732, the range for octahedral coordination.
Practice Q3. Which of the following is an inverse spinel?
(a) MgAl₂O₄ (b) Mn₃O₄ (c) Co₃O₄ (d) NiFe₂O₄
Answer: (d) NiFe₂O₄
Ni²⁺ (d⁸) has a strong octahedral site preference while high-spin Fe³⁺ (d⁵) has none, so Ni²⁺ goes octahedral and half the Fe³⁺ move to tetrahedral sites: Fe³⁺[Ni²⁺Fe³⁺]O₄. The other three are normal.
Practice Q4. A crystal shows a measured density lower than that calculated from its unit cell and formula. The defect most consistent with this is:
(a) Frenkel (b) Schottky (c) interstitial metal excess (d) none; a lower density is impossible
Answer: (b) Schottky
Schottky defects remove ions from the crystal, so mass per cell falls while the cell volume barely changes. Frenkel defects only move ions; interstitial excess adds mass.
Level 3: Part C Style
Practice Q5. A sample of iron oxide has composition Fe₀.₉₂O, with the rock-salt structure. Which statements are correct?
- About 17% of the iron is present as Fe³⁺.
- The oxide contains cation vacancies.
- The oxide would be expected to be an n-type semiconductor.
- The density is lower than that of an ideal FeO crystal with the same cell edge.
(a) 1 and 2 only (b) 1, 2 and 4 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4
Answer: (b) 1, 2 and 4 only
- \( 3x + 2(0.92 - x) = 2 \) gives \( x = 0.16 \); fraction of Fe as Fe³⁺ \( = 0.16/0.92 = 17.4\% \). Correct.
- Metal deficiency in a rock-salt oxide means vacant cation sites. Correct.
- Wrong. Fe³⁺ next to Fe²⁺ acts as a positive hole (electron can hop from Fe²⁺ to Fe³⁺); metal-deficient oxides are p-type.
- Fewer Fe per cell with the same O content and nearly the same a: lower density. Correct.
Practice Q6. CsCl has a cubic cell with \( a = 412 \) pm (M = 168.36 g mol⁻¹). Taking \( r(\text{Cl}^-) = 181 \) pm, find the Cs–Cl distance, the radius of Cs⁺, the radius ratio and the density. Is the structure consistent with the radius ratio rules?
Answer: 357 pm; 176 pm; 0.97; 4.00 g cm⁻³; yes
Cs⁺ at the body centre, Cl⁻ at the corners: contact along the body diagonal, \( d = \sqrt3a/2 = 356.8 \) pm. Then \( r(\text{Cs}^+) = 356.8 - 181 = 176 \) pm and \( r_+/r_- = 0.97 \), above 0.732, so cubic (8:8) coordination is predicted, as observed.
Z = 1: \( \rho = 168.36/[6.022 \times 10^{23} \times (4.12 \times 10^{-8})^3] = 4.00 \) g cm⁻³.
Practice Q7. Using the Born–Landé equation (\( N_Ae^2/4\pi\varepsilon_0 = 1.389 \times 10^5 \) kJ mol⁻¹ pm), compare the lattice energies of NaCl (\( r_0 = 282 \) pm, n = 8) and MgO (\( r_0 = 210 \) pm, n = 7), both rock salt (M = 1.748). Which statement is incorrect?
(a) U(NaCl) is about 750 kJ mol⁻¹. (b) U(MgO) is about 3960 kJ mol⁻¹. (c) Most of the difference comes from the charge product. (d) If MgO adopted the CsCl structure at the same \( r_0 \), its lattice energy would fall because M is smaller.
Answer: (d) is incorrect
(a) \( 1.389 \times 10^5 \times 1.748/282 \times 0.875 = 753 \) kJ mol⁻¹. Correct.
(b) \( 1.389 \times 10^5 \times 1.748 \times 4/210 \times (6/7) = 3964 \) kJ mol⁻¹. Correct.
(c) The charge product contributes a factor of 4; the distance contributes \( 282/210 = 1.34 \); the Born term about 0.98. Correct.
(d) Incorrect. The CsCl Madelung constant (1.763) is slightly larger than rock salt’s (1.748). The reason MgO does not take the CsCl structure is that \( r_0 \) would have to increase for eight-fold contact with a small cation; at fixed \( r_0 \) the energy would be marginally higher, not lower.
Quick Revision Table
| Idea | Formula or fact |
|---|---|
| Close packing | ccp ABC (= fcc), hcp AB; both CN 12, 74% |
| Ideal hcp | \( c/a = \sqrt{8/3} = 1.633 \) |
| Holes | N octahedral, 2N tetrahedral; fcc cell 4 O + 8 T |
| Radius ratio limits | 0.155 (3), 0.225 (4), 0.414 (6), 0.732 (8) |
| NaCl | ccp Cl⁻, Na⁺ all O; 6:6; Z = 4 |
| CsCl | simple cubic; 8:8; Z = 1 |
| Zinc blende / wurtzite | ccp / hcp S, half T; 4:4 |
| Fluorite / antifluorite | 8:4 / 4:8; Z = 4 |
| Rutile, CdI₂ | 6:3 |
| NiAs | hcp As, Ni all O |
| Corundum | hcp O, 2/3 O holes |
| Perovskite ABX₃ | A 12-coordinate, B 6; \( t = (r_A + r_X)/\sqrt2(r_B + r_X) \) |
| Spinel AB₂O₄ | ccp O, A 1/8 T, B 1/2 O; Fe₃O₄, NiFe₂O₄ inverse |
| Born–Landé | \( U = 1.389 \times 10^5 M\vert z_+z_-\vert (1 - 1/n)/r_0 \) kJ mol⁻¹ |
| Madelung | NaCl 1.748, CsCl 1.763, ZnS 1.638, CaF₂ 2.519 |
| Defects | Schottky (density falls), Frenkel (unchanged); \( n/N \approx e^{-\Delta H/2kT} \) |
| Non-stoichiometry | Fe₁₋ₓO p-type; ZnO excess metal n-type; F-centres colour |
Key Takeaways
What’s Next
- Bragg’s Law and X-ray Diffraction: how diffraction tells NaCl from CsCl and KCl, and fcc from bcc.
- Band Structure of Solids: why ZnS is an insulator, GaAs a semiconductor and NiAs metallic.
- Related structures of the elements: Allotropy of Main Group Elements for diamond and graphite.
- Back to the chapter overview.