Why can X-rays "see" the internal structure of crystals, while visible light — far brighter — merely glances off the surface? The answer has nothing to do with intensity or color, but with a single number: wavelength. Visible light has wavelengths around 5000 Å — far too large to sense interatomic spacings of a few angströms. X-rays, with wavelengths from 0.5 to a few Å, are precisely matched to the scale of the crystalline world. And when the size of the wave is comparable to the repeating distance of the structure, diffraction occurs — carrying with it all the information about the atomic arrangement inside.
Chapter 2 of Kittel's Introduction to Solid State Physics builds the entire mathematical language to describe this phenomenon: from the intuitive Bragg's law, through Fourier analysis, to the reciprocal lattice — the most powerful tool solid state physics has at its disposal. This post retraces that journey, prioritizing physical intuition over formalism.
1. Bragg's Law: When Imperfect Mirrors Work Best
W. L. Bragg proposed the first explanation in 1912 using an extremely intuitive picture: imagine each atomic plane in a crystal as a semi-transparent mirror — reflecting only a tiny fraction of the incident radiation (around \(10^{-3}\) to \(10^{-5}\)), while the rest passes through to the next plane. When reflections from many parallel planes add up in phase — that is, when the path difference equals an integer number of wavelengths — we get a strong diffracted beam.
Here \(d\) is the spacing between parallel atomic planes, \(\theta\) is the angle of incidence measured from the plane (not from the normal, as in conventional optics), \(\lambda\) is the wavelength, and \(n\) is the order of diffraction — a positive integer. The condition \(\lambda \leq 2d\) immediately shows why visible light cannot diffract from atomic planes: with \(d \sim 2\text{ Å}\), we need \(\lambda \leq 4\text{ Å}\), which means X-rays, thermal neutrons, or high-energy electrons.
Bragg's law tells us when diffraction occurs — the condition on angle and wavelength. But it does not explain why this condition is equivalent to the periodicity of the crystal lattice. For that, we need Fourier analysis.
A subtle point often overlooked: the "reflecting planes" in Bragg's law are not necessarily the physical surfaces of the sample. They are any family of atomic planes inside the crystal, specified by Miller indices (hkl). When we rotate a crystal in an X-ray beam, we successively bring different sets of planes into the Bragg condition.
2. The Reciprocal Lattice: When Fourier Meets Crystallography
The great leap from Bragg's law to a full scattering theory begins with a simple observation: the electron density \(n(\mathbf{r})\) in a crystal is periodic with respect to the lattice basis vectors \(\mathbf{a}_1, \mathbf{a}_2, \mathbf{a}_3\). Any periodic function can be expanded in a Fourier series. In three dimensions, the expansion reads:
The key question: what is the set of allowed wavevectors \(\mathbf{G}\) in this expansion? The requirement is that \(n(\mathbf{r} + \mathbf{T}) = n(\mathbf{r})\) for every lattice translation \(\mathbf{T} = u_1\mathbf{a}_1 + u_2\mathbf{a}_2 + u_3\mathbf{a}_3\). This demands \(\exp(i\mathbf{G} \cdot \mathbf{T}) = 1\), meaning \(\mathbf{G} \cdot \mathbf{T} = 2\pi \times \text{(integer)}\) for all lattice vectors \(\mathbf{T}\).
The set of all vectors \(\mathbf{G}\) satisfying this condition forms a lattice of its own — the reciprocal lattice. Its basis vectors \(\mathbf{b}_1, \mathbf{b}_2, \mathbf{b}_3\) are defined as:
The defining property is \(\mathbf{b}_i \cdot \mathbf{a}_j = 2\pi\delta_{ij}\) — each reciprocal basis vector is orthogonal to the other two real-space basis vectors. Every reciprocal lattice vector has the form \(\mathbf{G} = v_1\mathbf{b}_1 + v_2\mathbf{b}_2 + v_3\mathbf{b}_3\), where \(v_1, v_2, v_3\) are integers (these correspond to the Miller indices \(hkl\) in standard crystallographic notation).
Think of the reciprocal lattice as the frequency spectrum of a periodic signal. Real lattice ↔ reciprocal lattice is exactly like time-domain signal ↔ frequency spectrum in a Fourier transform. A crystal with period \(a\) in real space has reciprocal lattice spacing \(2\pi/a\). The denser the crystal structure (\(a\) small), the sparser its reciprocal lattice (\(b\) large), and vice versa.
3. The Diffraction Condition and the Laue Equations
With the reciprocal lattice in hand, we can state the diffraction condition precisely. The total scattered amplitude is proportional to the integral:
Substituting the Fourier expansion of \(n(\mathbf{r})\) and evaluating the integral, one finds that \(F\) is appreciable only when \(\Delta\mathbf{k} = \mathbf{G}\) — that is, when the scattering vector equals a reciprocal lattice vector. This is the diffraction condition:
The equation \(2\mathbf{k} \cdot \mathbf{G} = G^2\) is the form most commonly encountered in Brillouin zone problems. One can show it is completely equivalent to Bragg's law: the spacing between (hkl) planes is \(d(hkl) = 2\pi/|\mathbf{G}|\), and substituting this recovers \(2d\sin\theta = n\lambda\).
The Laue equations are a third equivalent statement, obtained by taking the dot product of the condition \(\Delta\mathbf{k} = \mathbf{G}\) with each of \(\mathbf{a}_1, \mathbf{a}_2, \mathbf{a}_3\):
The geometric interpretation is elegant: each equation constrains \(\Delta\mathbf{k}\) to lie on a cone around the axis \(\mathbf{a}_i\). For diffraction to occur, \(\Delta\mathbf{k}\) must simultaneously satisfy all three — it must lie at the intersection of three cones. This is a very restrictive condition, explaining why single-crystal diffraction experiments require systematic scanning of angle or wavelength.
To summarize, the three statements of the diffraction condition are physically equivalent:
| Statement | Equation | Emphasis |
|---|---|---|
| Bragg's law | \(2d\sin\theta = n\lambda\) | Geometric, intuitive |
| Reciprocal lattice condition | \(\Delta\mathbf{k} = \mathbf{G}\) | Fourier, reciprocal space |
| Laue equations | \(\mathbf{a}_i \cdot \Delta\mathbf{k} = 2\pi v_i\) | Cone geometry, experiment |
4. The Brillouin Zone: From Diffraction to Solid State Physics
The Brillouin zone is constructed from the reciprocal lattice in a manner completely analogous to the Wigner–Seitz cell in real space: choose a reciprocal lattice point as the origin, draw line segments to all neighboring lattice points, then construct the perpendicular bisector plane of each segment. The smallest enclosed volume around the origin is the first Brillouin zone.
The diffraction condition \(2\mathbf{k} \cdot \mathbf{G} = G^2\) can be rewritten as \(\mathbf{k} \cdot (\tfrac{1}{2}\mathbf{G}) = (\tfrac{1}{2}G)^2\), which means: a wave is diffracted if and only if the tip of its wavevector \(\mathbf{k}\) lies on the perpendicular bisector plane of a reciprocal lattice vector — that is, on the boundary of a Brillouin zone. Every wavevector \(\mathbf{k}\) ending at a zone boundary satisfies the Bragg condition.
| Crystal lattice | First Brillouin zone |
|---|---|
| Simple cubic (sc) | Cube |
| Body-centered cubic (bcc) | Rhombic dodecahedron (12 faces) |
| Face-centered cubic (fcc) | Truncated octahedron (14 faces) |
The Brillouin zone originated from diffraction theory, but its true significance extends far beyond that context: it is the fundamental unit for describing electronic band structure, phonon dispersion, and virtually all elementary excitations in solids. All electrical, optical, and thermal properties of materials are understood through the lens of this space.
5. The Structure Factor \(S_{\mathbf{G}}\): Bragg Says "Where," \(S_{\mathbf{G}}\) Says "How Bright"
Bragg's law determines which angles produce diffraction — but says nothing about the intensity of the diffracted beam. That depends on the electron distribution within each unit cell, encoded in the structure factor:
Here the sum runs over all atoms \(j\) in the unit cell, \(f_j\) is the atomic form factor — the ratio of the scattering amplitude of the real atom to that of a point electron — and \((x_j, y_j, z_j)\) are the fractional coordinates of the atom within the unit cell. The structure factor can vanish for certain reflections, giving rise to systematic absences (selection rules).
Systematic Absences in the bcc Lattice
The bcc lattice has two atoms per conventional cubic cell: one at \((0,0,0)\) and one at \((\tfrac{1}{2},\tfrac{1}{2},\tfrac{1}{2})\). The structure factor becomes:
When \(h + k + l\) is odd, the exponential equals \(-1\) and \(S = 0\): the reflection is completely forbidden. When \(h + k + l\) is even, \(S = 2f\). Thus the diffraction pattern of sodium (bcc structure) lacks the (100), (300), (111), and (221) reflections — but shows (200), (110), (222).
The physical explanation: the (100) reflection corresponds to a phase difference of \(2\pi\) between waves reflected from the two boundary planes of the cubic cell. But the bcc lattice has an atomic plane halfway between them — exactly half a lattice spacing away — so reflections from this midplane are shifted by \(\pi\) and perfectly cancel the boundary reflections.
Systematic Absences in fcc and the Classic KCl vs. KBr Example
The fcc lattice has four atoms per conventional cell: \((0,0,0)\), \((0,\tfrac{1}{2},\tfrac{1}{2})\), \((\tfrac{1}{2},0,\tfrac{1}{2})\), \((\tfrac{1}{2},\tfrac{1}{2},0)\). The structure factor vanishes whenever the Miller indices are mixed even and odd — reflections appear only when all indices are even or all are odd.
Both KCl and KBr have the fcc rock-salt structure, yet their X-ray diffraction patterns look strikingly different. In KCl, K⁺ has 18 electrons and Cl⁻ also has 18 electrons — their atomic form factors are nearly identical (\(f_K \approx f_{Cl}\)), making the crystal "look like" a simple cubic lattice with lattice constant \(a/2\). All reflections with mixed even/odd indices (which should be forbidden in fcc) are nearly extinguished. In KBr, Br⁻ has 36 electrons — its form factor is very different from K⁺, so all fcc-allowed reflections appear clearly.
6. The Unified Picture
After the journey from simple Bragg's law to the rich structure factor, we can step back and see the big picture: the periodicity of the crystal lattice generates a reciprocal lattice in Fourier space. Every diffraction condition — whether stated in the language of Bragg, Laue, or Brillouin zones — is a different expression of the same truth: diffraction occurs when the scattering vector coincides with a reciprocal lattice vector. A diffraction pattern does not map atoms — it maps the reciprocal lattice, with the intensity at each point reflecting the electron distribution within the unit cell through the structure factor.
This is the foundation of all X-ray crystallography — and also the gateway to the chapters that follow in Kittel: the electronic band structure (Chapter 7) is built entirely in reciprocal space, phonons (Chapter 4) are described by wavevectors \(\mathbf{q}\) in the same space, and virtually all physical properties of solids are analyzed through the lens of the Brillouin zone.
7. Further Reading and Exercises
For those who wish to go deeper, Chapter 4 (Phonons I) directly applies the concepts of reciprocal lattice and Brillouin zone to lattice vibrations, while Chapter 7 (Energy Bands) extends them to electronic band structure. From the problems in Chapter 2, I especially recommend: Problem 1 (interplanar spacing — practice connecting \(\mathbf{G}\) and \(d\)), Problem 4 (width of diffraction maxima — understanding the limits of finite crystals), and Problem 5 (diamond structure factor — applying selection rules to a more complex structure).
