Miller Indices, Reciprocal Space & Atomic-Scale Diffraction Mechanics

The quantitative analysis of crystalline diffraction, surface chemistry, and anisotropic mechanical deformation requires a mathematically rigorous coordinate notation to identify planes of lattice points and spatial directions within periodic lattices. This system was formulated by William Hallowes Miller (1839) and is universally denoted as Miller indices.

Direction Indices $[uvw]$

A crystallographic direction is a vector passing through lattice points. Let the direct lattice basis vectors be $\mathbf{a}, \mathbf{b}, \mathbf{c}$. Any lattice translation vector $\mathbf{T}$ can be written as:

where $u, v, w \in \mathbb{Z}$. To determine the direction indices $[uvw]$ of an arbitrary vector passing through the origin:

1. Express the vector components along the crystal axes $\mathbf{a}, \mathbf{b}, \mathbf{c}$ in terms of the lattice parameters $a, b, c$.
2. Reduce these components to the smallest integers having the same ratio:

1. Negative components are written with an overbar, e.g., $[\bar{1}10]$ represents $-1\mathbf{a} + 1\mathbf{b} + 0\mathbf{c}$.
2. A set of symmetry-equivalent directions (dictated by the point group of the crystal) is enclosed in angle brackets: $\langle uvw \rangle$. For example, in a cubic crystal, $\langle 100 \rangle$ encompasses the six equivalent directions $[100]$, $[\bar{1}00]$, $[010]$, $[0\bar{1}0]$, $[001]$, and $[00\bar{1}]$.

Miller Indices of Lattice Planes $(hkl)$

A family of parallel, equally spaced lattice planes is defined by three integers $(hkl)$ obtained through the following algorithm:

1. Find the intercepts of the plane with the crystallographic axes $\mathbf{a}, \mathbf{b}, \mathbf{c}$ in units of the lattice parameters $a, b, c$. Let these intercepts be $x_1, x_2, x_3$. If a plane is parallel to an axis, its intercept is at infinity ($\infty$).
2. Take the reciprocals of the fractional intercepts:

1. Multiply or divide by a common factor to clear fractions, reducing $h', k', l'$ to the smallest coprime integers $(hkl)$:

1. Negative intercepts are designated with overbars: $(\bar{h}kl)$.
2. A family of symmetry-equivalent crystallographic planes is enclosed in braces: $\{hkl\}$. In cubic symmetry, $\{100\} = (100), (\bar{1}00), (010), (0\bar{1}0), (001), (00\bar{1})$.

Hexagonal Miller-Bravais Indices $(hkil)$

In hexagonal and trigonal systems where axes $\mathbf{a}_1, \mathbf{a}_2, \mathbf{a}_3$ lie in the basal plane at $120^\circ$ angles with vertical axis $\mathbf{c}$, standard $(hkl)$ indices do not immediately reveal hexagonal rotational equivalence. The Miller-Bravais 4-index notation $(hkil)$ resolves this by introducing a redundant third index $i$:

Because $\mathbf{a}_1 + \mathbf{a}_2 + \mathbf{a}_3 = \mathbf{0}$, the condition $h + k + i = 0$ is strictly preserved. In this system, prism planes $(10\bar{1}0)$, $(01\bar{1}0)$, and $(\bar{1}100)$ are immediately seen to belong to the identical family $\{10\bar{1}0\}$ by permuting indices.

The perpendicular distance between adjacent parallel planes of the family $(hkl)$ is the interplanar spacing $d_{hkl}$. This parameter directly dictates the diffraction angle $\theta$ observed in X-ray and electron scattering experiments.

Derivation for the Orthorhombic System

In an orthogonal coordinate system where $\alpha = \beta = \gamma = 90^\circ$ and unit cell dimensions are $a, b, c$, consider the plane $(hkl)$ passing closest to the origin. Its intercepts on the axes are $a/h, b/k, c/l$. The normal unit vector $\hat{\mathbf{n}}$ from the origin to the plane forms direction cosines:

Since the sum of squared direction cosines for orthogonal axes equals unity ($\cos^2 \alpha_1 + \cos^2 \beta_1 + \cos^2 \gamma_1 = 1$):

Factoring out $d_{hkl}^2$ gives the general orthorhombic formula:

Special Cases: Cubic and Tetragonal Systems

1. Cubic System ($a = b = c$):

2. Tetragonal System ($a = b \ne c$, $\alpha = \beta = \gamma = 90^\circ$):

Non-Orthogonal Systems

1. Hexagonal System ($a = b \ne c$, $\alpha = \beta = 90^\circ, \gamma = 120^\circ$):

2. Monoclinic System ($a \ne b \ne c$, $\alpha = \gamma = 90^\circ, \beta \ne 90^\circ$):

3. Triclinic System (General case with metric tensor $G_{ij} = \mathbf{a}_i \cdot \mathbf{a}_j$):

where $V = abc\sqrt{1 - \cos^2\alpha - \cos^2\beta - \cos^2\gamma + 2\cos\alpha\cos\beta\cos\gamma}$ is the unit cell volume.

The concept of the reciprocal lattice—introduced by Josiah Willard Gibbs and formalized in crystallography by Paul Ewald (1913)—is the foundational mathematical bridge connecting real-space crystal geometry to diffraction phenomena observed in momentum space ($\mathbf{k}$-space).

Mathematical Definition of Reciprocal Basis Vectors

Let $\mathbf{a}, \mathbf{b}, \mathbf{c}$ be the primitive basis vectors of the direct crystal lattice with unit cell volume $V_c = \mathbf{a} \cdot (\mathbf{b} \times \mathbf{c})$. The reciprocal lattice basis vectors $\mathbf{a}^, \mathbf{b}^, \mathbf{c}^*$ (often written $\mathbf{b}_1, \mathbf{b}_2, \mathbf{b}_3$ in solid-state physics with a $2\pi$ factor) are defined by the orthogonality relations:

These basis vectors satisfy the Kronecker delta relations:

In physics conventions (including band theory), a factor of $2\pi$ is incorporated: $\mathbf{b}_i = 2\pi \mathbf{a}_i^*$, ensuring $\mathbf{a}_i \cdot \mathbf{b}_j = 2\pi \delta_{ij}$.

Fundamental Theorems of the Reciprocal Lattice

1. Normality Theorem: Every reciprocal lattice vector

is perpendicular to the real-space crystallographic plane $(hkl)$. Proof: Consider two vectors $\mathbf{u}_1$ and $\mathbf{u}_2$ lying in the plane $(hkl)$:

Taking the scalar product:

Similarly, $\mathbf{G}_{hkl} \cdot \mathbf{u}_2 = 0$. Because $\mathbf{G}_{hkl}$ is orthogonal to two non-parallel vectors spanning the plane $(hkl)$, it is strictly normal to $(hkl)$.

2. Spacing Theorem: The magnitude of $\mathbf{G}_{hkl}$ is inversely proportional to the interplanar spacing $d_{hkl}$:

Reciprocal Lattices of Common Bravais Lattices

• Simple Cubic (SC): Direct lattice parameter $a$. The reciprocal lattice is another simple cubic lattice with parameter $a^* = 1/a$ (or $2\pi/a$).
• Face-Centered Cubic (FCC): Direct lattice primitive vectors form an FCC cell with volume $a^3/4$. Its reciprocal lattice is Body-Centered Cubic (BCC) with cubic reciprocal cell parameter $a^*_{\text{cubic}} = 2/a$ (or $4\pi/a$).
• Body-Centered Cubic (BCC): Direct lattice primitive vectors form a BCC cell with volume $a^3/2$. Its reciprocal lattice is Face-Centered Cubic (FCC).

The Ewald Sphere Construction

Paul Ewald introduced an elegant geometric construction to represent the diffraction condition:

1. Draw the reciprocal lattice of the crystal.
2. Direct the incident wavevector $\mathbf{k}_0$ ($|\mathbf{k}_0| = 1/\lambda$) terminating at the reciprocal lattice origin $(000)$.
3. Construct a sphere of radius $R = |\mathbf{k}_0| = 1/\lambda$ centered at the tip of $\mathbf{k}_0$ (point $C$).
4. Diffraction occurs if and only if any reciprocal lattice point $(hkl)$ lies exactly on the surface of the Ewald sphere.

When a point $P = (hkl)$ intersects the sphere surface, the scattered wavevector $\mathbf{k}_s = \vec{CP}$ has magnitude $1/\lambda$, ensuring elastic scattering ($|\mathbf{k}_s| = |\mathbf{k}_0|$), and satisfies the Laue condition $\mathbf{k}_s - \mathbf{k}_0 = \mathbf{G}_{hkl}$.

In 1912, William Henry Bragg and William Lawrence Bragg formulated an intuitive physical model of X-ray diffraction by treating crystal planes as specular semi-transparent mirrors.

Geometrical Derivation of Bragg's Law

Consider a monochromatic beam of X-rays with wavelength $\lambda$ incident at a glancing angle $\theta$ (the Bragg angle) on a set of parallel crystallographic planes with interplanar spacing $d_{hkl}$.

1. Ray 1 reflects specularly from the uppermost plane at atom $A$.
2. Ray 2 reflects specularly from the adjacent parallel plane at atom $B$, separated by perpendicular distance $d_{hkl}$.
3. Drop perpendiculars from atom $A$ onto the incident ray ($AD$) and the reflected ray ($AE$).
4. The extra path length $\Delta L$ traveled by Ray 2 relative to Ray 1 is:

1. For constructive interference between the scattered waves, the path difference must be an integer multiple of the incident wavelength $\lambda$:

In modern crystallographic notation, the order of reflection $n$ is absorbed directly into the Miller indices $(hkl)$: a second-order reflection $n=2$ from $(100)$ is denoted as the first-order reflection from $(200)$, where $d_{200} = d_{100}/2$. Hence, Bragg's Law is written simply:

Physical Limitations: Why $\sin\theta \le 1$

Because $|\sin\theta| \le 1$, constructive diffraction can only occur if:

If the incident radiation has wavelength $\lambda > 2d_{\max}$, no diffraction peaks can exist at any angle. For typical crystal lattice parameters ($a \approx 2 - 6\text{ Å}$), $d_{hkl} \approx 0.5 - 4\text{ Å}$. Therefore, diffraction requires radiation with $\lambda \sim 0.5 - 2.5\text{ Å}$, which corresponds precisely to the X-ray regime (e.g., $\text{Cu } K_\alpha_1 = 1.54056\text{ Å}$, $\text{Mo } K_\alpha = 0.71073\text{ Å}$), thermal neutrons ($\lambda \sim 1 - 2\text{ Å}$), and high-energy electrons ($100 - 300\text{ keV}$, $\lambda \sim 0.02 - 0.04\text{ Å}$).

Kinematic vs Dynamical Diffraction Theories

• Kinematic Theory (Darwin-Staverman): Assumes single scattering only. The incident beam passes through the crystal without significant attenuation, and scattered beams do not rescatter. Valid for imperfect, mosaic crystals, powders, and thin specimens ($\text{thickness} < 100\text{ nm}$).
• Dynamical Theory (Ewald-von Laue): Accounts for multiple coherent scattering between incident and diffracted beams inside a perfect, macroscopic single crystal. Gives rise to anomalous transmission (the Borrmann effect), extinction effects (primary and secondary extinction), and Darwin reflection curves with finite plateau widths.

Max von Laue (1912) formulated diffraction from a three-dimensional periodic array of scatterers without invoking the planar specular mirror analogy.

The Three Laue Equations

Consider a 1D row of identical atoms separated by translation vector $\mathbf{a}$. Let unit vectors in the direction of the incident and scattered beams be $\hat{\mathbf{s}}_0$ and $\hat{\mathbf{s}}$. The path difference between rays scattered by adjacent atoms is $\mathbf{a} \cdot (\hat{\mathbf{s}} - \hat{\mathbf{s}}_0)$. For constructive interference across the 1D lattice:

Extending this requirement to an infinite three-dimensional crystal with basis vectors $\mathbf{a}, \mathbf{b}, \mathbf{c}$ requires simultaneous constructive interference along all three crystal dimensions:

These are the three Laue equations. Geometrically, each equation defines a family of coaxial cones around the corresponding crystal axis. Diffraction beams emerge only along the lines of intersection of three mutually compatible cones.

Vector Equivalence to the Reciprocal Lattice

Define the wavevector of the incident beam as $\mathbf{k}_0 = \frac{1}{\lambda}\hat{\mathbf{s}}_0$ and the scattered beam as $\mathbf{k} = \frac{1}{\lambda}\hat{\mathbf{s}}$. The scattering vector $\Delta \mathbf{k}$ is:

Dividing the Laue equations by $\lambda$:

Recalling the definition of the reciprocal lattice basis vectors where $\mathbf{a}_i \cdot \mathbf{a}_j^* = \delta_{ij}$, the unique solution for $\Delta \mathbf{k}$ satisfying all three equations simultaneously is:

In physics notation (where $|\mathbf{k}| = 2\pi/\lambda$):

This elegant result proves that the Laue condition is mathematically identical to Bragg's Law: diffraction occurs if and only if the change in photon wavevector equals a reciprocal lattice vector $\mathbf{G}_{hkl}$.

While the dimensions of the unit cell determine the positions of diffraction peaks ($\theta$ angles), the arrangement of specific atoms within the unit cell determines the intensity of each reflection through the structure factor $F_{hkl}$.

Mathematical Formulation of the Structure Factor

The unit cell structure factor $F_{hkl}$ represents the total amplitude and phase of radiation scattered by all $N$ atoms in the unit cell relative to a single free electron at the origin:

where:

• $(x_j, y_j, z_j)$ are the fractional coordinates of the $j$-th atom in the unit cell.
• $f_j$ is the atomic scattering factor (or atomic form factor) of atom $j$, defined as the Fourier transform of its electron density $\rho_j(\mathbf{r})$:

At zero scattering angle ($q = 0$), $f_j(0) = Z_j$ (the atomic number, total number of electrons). As $\theta$ increases, destructive interference between rays scattered from different parts of the electron cloud causes $f_j$ to monotonically decrease.

The diffracted intensity $I_{hkl}$ is proportional to the modulus squared:

where $\phi_j = 2\pi (hx_j + ky_j + lz_j)$.

Systematic Absences (Extinctions) from Lattice Centering

Non-primitive Bravais lattices possess translational centering operators that cause destructive interference for entire classes of reflections:

1. Body-Centered (I-Centering):

• Coordinates: $(0,0,0)$ and $\left(\frac{1}{2}, \frac{1}{2}, \frac{1}{2}\right)$.
• Structure factor:

• Selection Rule: Reflections are observed only when $h + k + l = \text{even}$. If $h + k + l = \text{odd}$, $F_{hkl} = 0$ (extinction).

2. Face-Centered (F-Centering):

• Coordinates: $(0,0,0)$, $\left(\frac{1}{2}, \frac{1}{2}, 0\right)$, $\left(\frac{1}{2}, 0, \frac{1}{2}\right)$, $\left(0, \frac{1}{2}, \frac{1}{2}\right)$.
• Structure factor:

• If $h, k, l$ are unmixed (all even or all odd): $(-1)^{h+k} = 1$, so $F_{hkl} = 4f$.
• If $h, k, l$ are mixed (some even, some odd): $F_{hkl} = 0$.
• Selection Rule: Reflections are observed only when $h, k, l$ are all even or all odd.

3. Base-Centered (C-Centering):

• Coordinates: $(0,0,0)$ and $\left(\frac{1}{2}, \frac{1}{2}, 0\right)$.
• Selection Rule: Observed only when $h + k = \text{even}$.

Extinctions from Translational Symmetry Elements

• Glide Planes: Cause systematic absences in 2D zonal reflections (e.g., an $a$-glide perpendicular to $c$ requires $h = \text{even}$ for $(hk0)$ reflections).
• Screw Axes: Cause systematic absences in 1D axial reflections (e.g., a $2_1$ screw axis along $c$ requires $l = \text{even}$ for $(00l)$ reflections).

The practical determination of crystal structures, lattice parameters, and phase purity relies on specialized experimental geometries designed to bring reciprocal lattice points into contact with the Ewald sphere.

The Debye-Scherrer Powder Method

In powder X-ray diffraction (PXRD), the sample is finely ground into millions of randomly oriented crystallites ($1 - 10\text{ }\mu\text{m}$).

• Diffraction Geometry: Because all spatial orientations are statistically populated, every reciprocal lattice vector $\mathbf{G}_{hkl}$ sweeps out a continuous sphere in reciprocal space.
• The intersection of this reciprocal sphere with the Ewald sphere generates a cone of diffracted rays with semi-apex angle $2\theta_{hkl}$ coaxial with the incident beam.
• These cones intersect a planar detector or cylindrical film as concentric circular rings (Debye-Scherrer rings).
• In a modern Bragg-Brentano diffractometer ($\theta - 2\theta$ geometry), a scintillation or silicon strip detector rotates along a goniometer circle, recording diffracted intensity as a function of scattering angle $2\theta$.

The Rotating Single-Crystal Method

To solve unknown crystal structures without high-resolution single-crystal area detectors, the rotating crystal method (or Weissenberg/precession technique) aligns a single crystal along a known crystallographic axis (e.g., $\mathbf{c}$):

• The crystal is rotated around $\mathbf{c}$ inside a monochromatic X-ray beam.
• The reciprocal lattice consists of planar layers perpendicular to the rotation axis spaced by $c^* = 1/c$.
• As the crystal rotates, reciprocal lattice points within each layer intersect the Ewald sphere, producing horizontal parallel rows of spots on a cylindrical film called layer lines.
• The spacing between layer lines directly gives the lattice parameter along the rotation axis:

where $\mu_n$ is the angle subtended by the $n$-th layer line from the equatorial line ($n=0$).

Peak Broadening & The Scherrer Formula

In real nanomaterials, diffraction peaks possess finite angular widths $\beta$ (full width at half maximum, FWHM) due to crystallite size and lattice strain:

where $K \approx 0.9$ is the Scherrer shape factor, $L$ is the volume-weighted mean crystallite column length, and $\varepsilon = \Delta d/d$ is root-mean-square microstrain. Plotting $\beta\cos\theta$ versus $\sin\theta$ (Williamson-Hall plot) separates the size contribution (intercept) from microstrain (slope).

While X-rays scatter from atomic electron clouds, structural chemistry utilizes two complementary quantum probes: electrons and neutrons.

Comparison of Quantum Scattering Probes

| Characteristic | X-Ray Diffraction (XRD) | Electron Diffraction (ED/TEM) | Neutron Diffraction (ND) | | :--- | :--- | :--- | :--- | | Primary Interaction | Electromagnetic with electron cloud $\rho(\mathbf{r})$ | Coulombic with electrostatic potential $\phi(\mathbf{r})$ | Strong nuclear force with atomic nuclei | | Scattering Cross-Section | $\sim 10^{-24}\text{ cm}^2$ (Moderate) | $\sim 10^{-20}\text{ cm}^2$ ($10^4 - 10^5\times$ larger) | $\sim 10^{-24}\text{ cm}^2$ (Moderate/Weak) | | Penetration Depth | $10 - 100\text{ }\mu\text{m}$ (Bulk) | $10 - 100\text{ nm}$ (Thin specimens only) | Several $\text{cm}$ (Deep bulk probe) | | Sensitivity to Light Elements | Proportional to $Z^2$; insensitive to $\text{H}, \text{Li}$ | Good sensitivity to light atoms | High sensitivity (nuclear scattering length $b$) | | Isotopic Sensitivity | None ($Z$ identical) | None | Exceptional (e.g., $^1\text{H}$ vs $^2\text{D}$) | | Magnetic Moment | Negligible magnetic scattering | Negligible | Large ($\mu_n = -1.913\mu_N$); maps spin order |

Electron Diffraction in Transmission Electron Microscopy (TEM)

High-energy electrons ($100 - 300\text{ keV}$) possess extremely short de Broglie wavelengths:

For $200\text{ kV}$ electrons, relativistic correction yields $\lambda = 0.0251\text{ Å}$. Because $\lambda$ is two orders of magnitude smaller than X-ray wavelengths, the Ewald sphere radius $R = 1/\lambda \approx 40\text{ Å}^{-1}$ is virtually planar on the scale of the reciprocal lattice. Consequently, an entire 2D plane of the reciprocal lattice intersects the Ewald sphere simultaneously, yielding instantaneous 2D spot patterns (Selected Area Electron Diffraction, SAED).

Neutron Diffraction & Magnetic Crystallography

Neutrons produced in spallation sources or nuclear reactors are thermalized to room temperature ($E \approx 25\text{ meV}$), giving de Broglie wavelengths $\lambda \approx 1.8\text{ Å}$, ideal for crystal diffraction:

Because neutrons interact via the strong nuclear force, the nuclear scattering length $b$ varies erratically across the periodic table and between isotopes (e.g., $b(^1\text{H}) = -3.74\text{ fm}$ vs $b(^2\text{D}) = +6.67\text{ fm}$), allowing precise localization of hydrogen in proteins and hydrates via deuteration. Furthermore, the neutron carries an intrinsic magnetic dipole moment $\mu_n = -1.913\mu_N$. The magnetic interaction with unpaired electron spins enables direct determination of antiferromagnetic, ferrimagnetic, and helical spin structures (e.g., the historical discovery of antiferromagnetism in $\text{MnO}$ by Clifford Shull, 1949).