“Without compass and square, neither circle nor square can be formed.”Mencius · Stepping up from 2D planar tilings to 3D crystal lattices
Triclinic has zero right angles, whereas monoclinic jumps straight to two (\(\alpha=\gamma=90^\circ\))—why is there no intermediate “diclinic” crystal system with just one \(90^\circ\) angle?
Combining 7 crystal systems with 4 centerings (primitive P, body I, face F, base S) gives \(7\times 4 = 28\) formal candidates—why do only 14 independent Bravais lattices survive?
| System | Characteristic Symmetry | DoF & Parameter Constraints | Bravais Lattices | Fate of Eliminated Centerings |
|---|---|---|---|---|
| Triclinic | None (only \(1\) or \(\bar{1}\)) | 6: \(a\ne b\ne c,\ \alpha\ne\beta\ne\gamma\) | P (1) | I, S, F \(\Rightarrow\) reselect basis to triclinic P |
| Monoclinic | One 2-fold axis (or \(\mathrm{m}\)) | 4: \(a\ne b\ne c,\ \alpha=\gamma=90^\circ\ne\beta\) | P, S (2) | I, F \(\Rightarrow\) reselect \(ac\)-plane basis to monoclinic S |
| Orthorhombic | Three \(\perp\) 2-fold axes (or \(\mathrm{mm2}\)) | 3: \(a\ne b\ne c,\ \alpha=\beta=\gamma=90^\circ\) | P, S, I, F (4) | All four survive independently! |
| Tetragonal | One 4-fold axis (or \(\bar{4}\)) | 2: \(a=b\ne c,\ \alpha=\beta=\gamma=90^\circ\) | P, I (2) | S \(\Rightarrow\) tetragonal P; F \(\Rightarrow\) tetragonal I (\(45^\circ\)) |
| Trigonal | One 3-fold axis (or \(\bar{3}\)) | 2: \(a=b=c,\ \alpha=\beta=\gamma\ne 90^\circ\) | R (1, rhomb.) | I, F \(\Rightarrow\) R; S \(\Rightarrow\) breaks \(C_3\), degrades to monoclinic |
| Hexagonal | One 6-fold axis (or \(\bar{6}\)) | 2: \(a=b\ne c,\ \alpha=\beta=90^\circ,\ \gamma=120^\circ\) | P (1) | I, S, F \(\Rightarrow\) break \(C_6\), degrade to orthorhombic |
| Cubic | Four body-diagonal 3-fold axes | 1: \(a=b=c,\ \alpha=\beta=\gamma=90^\circ\) | P, I, F (3) | S \(\Rightarrow\) breaks \(C_3\), reduces to tetragonal P (14 total) |
Crystal systems are classified by characteristic symmetry elements, not by counting \(90^\circ\) angles: in 3D, a single 2-fold axis (or perpendicular mirror) forces orthogonality to two axes simultaneously (\(\alpha=\gamma=90^\circ\), monoclinic). An accidental single \(90^\circ\) angle carries no new symmetry and remains triclinic.
Out of \(7\times 4 = 28\) centering candidates, half are eliminated either by breaking characteristic symmetry (†, e.g., C-centered cubic loses its 3-fold axes and degrades to tetragonal P) or by same-system basis reselection (§, e.g., tetragonal F rotated \(45^\circ\) = tetragonal I; two base centers close into all-face F), leaving 14 Bravais lattices.
Next lecture: With unit-cell axes established, how do we index arbitrary directions and planes? Miller indices \([uvw]/(hkl)\), hexagonal Miller–Bravais indices, and the Weiss zone law.
1. A crystal happens to have \(a=b=c\) and \(\alpha=\beta=\gamma=90^\circ\), yet lacks four body-diagonal 3-fold axes. Does it belong to the cubic system?
Crystal systems are defined by characteristic symmetry, not accidental metric equality; four body-diagonal 3-fold (or \(\bar{3}\)) axes are mandatory for cubic.
2. Why is there no independent “face-centered tetragonal (F-tetragonal)” lattice among the 14 Bravais lattices?
F-tetragonal preserves the 4-fold axis, but choosing \(a/\sqrt{2}\) basal half-diagonals halves the cell volume into standard body-centered tetragonal (I).
3. In space group \(\mathrm{P6_3/mmc}\) of hcp Mg and Ti, what do \(\mathrm{6_3}\) and the trailing \(\mathrm{c}\) denote?
\(6_3\) rotates by \(360^\circ/6=60^\circ\) with translation \(\tfrac{3}{6}c=c/2\) along \(c\); trailing \(\mathrm{c}\) denotes a mirror reflection followed by a \(c/2\) axial glide.