Nonsymmorphic Symmetry

Symmetry is the language of order in the universe, and nowhere is this more apparent than in the perfect, repeating lattices of crystals. We are familiar with simple symmetries like rotations and reflections, which define the beautiful facets of a gemstone. When these operations can be performed around a single point, we call the resulting spatial symmetry "symmorphic." However, nature's artistry often involves a more intricate form of order, one that weaves translation and rotation into a single, indivisible action. This is the domain of nonsymmorphic symmetry, a concept whose subtle rules give rise to some of the most robust and exotic phenomena in modern physics.

This article addresses the gap between the simple perception of crystal symmetry and the profound quantum mechanical consequences of these hidden operations. We will explore the fundamental principles behind nonsymmorphic symmetries, then examine their far-reaching applications and interdisciplinary impact. The following chapters will guide you through this journey. In "Principles and Mechanisms," we will define these symmetries, uncover how they are detected, and explain the beautiful paradox that forces electronic bands to stick together. Subsequently, in "Applications and Interdisciplinary Connections," we will see how these principles lead to real-world phenomena, from predictable patterns in diffraction experiments to the engineering of novel topological materials.

In our journey so far, we've come to appreciate that a crystal is a universe of order, a repeating pattern of atoms governed by the laws of symmetry. We are used to thinking about simple symmetries, the kind you can see with your own eyes: rotations that leave a snowflake unchanged, or reflections in a mirror. In a crystal, we add the symmetry of translation—shifting the entire pattern by a certain distance brings it back onto itself. When all the rotational and reflectional symmetries of a crystal can be performed while keeping one point fixed, as if the entire crystal is built by placing a block (the "motif") at each point of a grid (the "lattice"), we call the symmetry ​​symmorphic​​. It's a straightforward, satisfying kind of order.

But nature, it turns out, is a more subtle artist. It employs symmetries that are not immediately obvious, symmetries that weave together two different actions into one indivisible operation. These are the ​​nonsymmorphic​​ symmetries, and they are the secret behind some of the most fascinating and robust phenomena in the world of materials.

Imagine a spiral staircase. You can't get from one step to the one directly above it by a simple vertical shift. Nor can you get there by a simple rotation around the central column. To move from one step to the next, you must both rotate and translate upwards. The two actions are fused. This is the essence of a ​​screw axis​​. A screw axis operation, often denoted as $N_k$, involves a rotation by $2\pi/N$ followed by a fractional translation along the axis of rotation. For example, a ​​$2_1$ screw axis​​ rotates a point by $180^\circ$ and then slides it by half a lattice vector parallel to the axis. You have to perform the operation twice to get back to a position equivalent to a pure lattice translation.

Another wonderful example is the ​​glide plane​​. Imagine walking in the snow, leaving a trail of footprints. Left foot, slide forward; right foot, slide forward. A simple reflection across the center line of your path won't map your footprints onto themselves—a left print would become a right print. To map the pattern, you must reflect a left footprint to a right one, and then slide it forward by half a step. This combined operation of reflection plus a fractional translation parallel to the plane is a glide plane.

These operations are "nonsymmorphic" because the fractional translation part is intrinsic to the symmetry; you can't get rid of it by simply shifting your origin. The crystal is not built by placing identical, identically oriented blocks on a lattice. Instead, the blocks themselves are rotated or flipped relative to one another as they are placed. The symmetry is hidden in the relationship between the lattice points.

How can we possibly know these subtle, sub-atomic ballets are taking place? We can't see atoms, let alone a half-a-lattice-vector slide. The answer, as is so often the case in physics, comes from waves. When we shine X-rays or neutrons on a crystal, they diffract, creating a pattern of bright spots that is essentially the Fourier transform of the atomic arrangement. This diffraction pattern is an exquisite map of the crystal's symmetries.

A perfectly periodic arrangement gives a perfectly periodic set of diffraction spots. But the fractional translations in nonsymmorphic symmetries introduce a crucial twist. They cause a specific kind of destructive interference. The wave scattered from an atom at position $\mathbf{r}$ and the wave scattered from its glide-partner at a position like (reflect $\mathbf{r}$ + slide) can arrive perfectly out of phase for certain directions, cancelling each other out completely.

This leads to a remarkable and directly observable phenomenon: ​​systematic absences​​. Certain diffraction spots that you would expect to see based on the lattice geometry are mysteriously missing. For example, in a crystal with a glide plane involving a reflection across the $x=0$ plane and a slide by $\frac{1}{2}\mathbf{b}$ along the $y$-axis, the laws of diffraction predict that all reflections of the type $(0, k, l)$ will be completely absent if $k$ is an odd number. This happens regardless of where the atoms are actually located in the unit cell; it is a direct consequence of the geometry of the symmetry itself. Finding these systematic patterns of missing spots in an experiment is like finding the fossilized footprints of the glide plane's dance—it is the irrefutable evidence that a nonsymmorphic symmetry is at play.

The consequences of nonsymmorphic symmetries run much deeper than just the static structure of the crystal. They have a profound and beautiful effect on the electrons that move within it. According to Bloch's theorem, electrons in a crystal behave like waves, each characterized by a crystal momentum vector $\mathbf{k}$. The allowed energies for these electrons form continuous bands, a landscape known as the electronic band structure.

The space of all possible $\mathbf{k}$ vectors forms the ​​Brillouin zone (BZ)​​, which is the fundamental "unit cell" of momentum space. Symmetries of the crystal impose strict rules on the shape of the energy bands. And it is at the boundaries of the Brillouin zone where nonsymmorphic symmetries perform their most startling magic.

A point on the BZ boundary, for instance at $k_x = \pi/a$, represents an electron wave whose wavelength is perfectly "out of sync" with the lattice, with a phase that flips sign from one unit cell to the next. Let's see what happens when we combine this with a nonsymmorphic symmetry, like a $2_1$ screw axis along $x$ whose operation is $S = \{ \text{rotation by } 180^{\circ} \text{ about } x \mid \frac{a}{2}\hat{\mathbf{x}} \}$.

Let's apply this operation twice. The first rotation is cancelled by the second, and the two half-translations add up to a full lattice vector translation, $T_a = a\hat{\mathbf{x}}$. So, $S^2 = T_a$. Now, consider an electron state $|\psi\rangle$ at the BZ boundary, $k_x = \pi/a$. The action of a full lattice translation $T_a$ on this state gives it a phase factor of $\exp(-i k_x a) = \exp(-i\pi) = -1$. So, for any state at this momentum, the operator $S^2$ must act as multiplication by $-1$:

This seems simple enough, but it has a staggering consequence. If $|\psi\rangle$ is an eigenstate of the symmetry operator $S$ with eigenvalue $\lambda$ (i.e. $S|\psi\rangle = \lambda|\psi\rangle$), then applying $S$ twice gives $S^2|\psi\rangle = \lambda^2|\psi\rangle$. Comparing our two results, we are forced to conclude that:

This means the eigenvalues of the symmetry operator $S$ are not $\pm 1$ as one might expect for a rotation, but are the imaginary numbers $\pm i$! The symmetry operator, when acting on electron waves at the BZ boundary, behaves like the mathematical imaginary unit $i$.

This peculiar result is not just a mathematical curiosity; it is the key to understanding why nonsymmorphic symmetries force energy bands to stick together. This phenomenon is known as ​​band sticking​​ or enforced degeneracy. We can understand this through a beautiful argument that reveals a deep contradiction.

The crystal momentum $\mathbf{k}$ is periodic. Just as the real-space lattice repeats, the momentum-space lattice repeats. A point $\mathbf{k}$ is physically indistinguishable from a point $\mathbf{k}+\mathbf{G}$, where $\mathbf{G}$ is any vector of the reciprocal lattice. For our 1D example, if $\mathbf{k} = (\pi/a, 0, 0)$, we can choose $\mathbf{G} = (2\pi/a, 0, 0)$. The point $\mathbf{k}+\mathbf{G} = (3\pi/a, 0, 0)$ describes the exact same physical state as $\mathbf{k}$.

So, any physical property of the state, including its eigenvalue under the symmetry $S$, must be the same at $\mathbf{k}$ and $\mathbf{k}+\mathbf{G}$. But is it? The mathematics of group representations tells us how the eigenvalue transforms. The eigenvalue $\lambda(\mathbf{k}+\mathbf{G})$ is related to the eigenvalue at $\lambda(\mathbf{k})$ by a phase factor that depends on the fractional translation $\mathbf{t} = \frac{a}{2}\hat{\mathbf{x}}$:

Let's compute this phase factor: $\mathbf{G} \cdot \mathbf{t} = (2\pi/a)(a/2) = \pi$. The phase factor is $\exp(-i\pi)=-1$. So we find:

Here lies the paradox! Physics demands that the eigenvalue be the same, $\lambda(\mathbf{k}+\mathbf{G}) = \lambda(\mathbf{k})$, because the points describe the same state. But the mathematics of the symmetry demands that the eigenvalue flip its sign! The only way to satisfy $\lambda = -\lambda$ is if $\lambda=0$, but we already proved the eigenvalues must be $\pm i$.

The only way out of this logical impasse is to abandon our initial assumption: that the state $|\psi\rangle$ is a single, non-degenerate eigenstate. It cannot be. The state at this momentum must belong to a multi-dimensional space. The symmetry operator $S$ cannot act on a single state; it must act on a set of at least two states, mixing them together. Its representation cannot be a single number ($\pm i$), it must be a matrix, for instance, a $2 \times 2$ matrix like $\begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix}$, which squares to $-\mathbb{1}$. A $2 \times 2$ matrix acts on a two-component vector, which means there must be at least two degenerate states at that energy. The bands must touch.

This is a profoundly beautiful result. The subtle, screw-like symmetry of the atomic positions creates a mathematical paradox for the quantum waves moving through them, and the only way nature can resolve this paradox is by forcing two energy levels to become one. The bands are "stuck" together, not by accident, but by the fundamental laws of symmetry and quantum mechanics. A similar logic applies when you have two nonsymmorphic symmetries whose representations anticommute, which also forbids one-dimensional solutions and forces degeneracy.

One must be careful. Does a single screw axis always force degeneracy? The surprising answer is no. Our argument about eigenvalues had a hidden participant: time-reversal symmetry. In many common physical systems, the combination of the nonsymmorphic spatial symmetry and time-reversal symmetry is what creates the unavoidable degeneracy. If one explicitly breaks time-reversal symmetry (for example, in a magnetic material), it is possible to find a loophole in the argument. A single, non-degenerate state with an eigenvalue of $i$ can exist, because there is no time-reversal symmetry to demand the existence of its $-i$ partner. Even more exotic scenarios are possible, where time-reversal itself is combined with a fractional translation, creating an anti-unitary nonsymmorphic symmetry which can enforce degeneracy on its own, even for spinless particles.

These enforced degeneracies are more than just a curiosity. They are the "pins" that hold the fabric of the electronic band structure in place. The way the bands must connect between these unavoidable sticking points can dictate the overall "topology" of the band structure, much like the number of holes in a donut dictates its fundamental nature. These nonsymmorphic symmetries are often the breeding ground for exotic topological phases of matter, such as Dirac and Weyl semimetals, materials that host strange, massless electronic states with robust properties. What begins with a simple, elegant twist in the crystal lattice culminates in some of the most exciting and promising materials in modern physics.

Now that we have grappled with the definition of a nonsymmorphic symmetry—this peculiar marriage of a rotation or reflection with a fractional trip through the crystal lattice—a fair question to ask is: "So what?" Does nature truly care about this seemingly esoteric rule? Is it a mere crystallographic curiosity, or does it have profound, observable consequences?

The answer, as is so often the case in physics, is that this subtle twist in the blueprint of a crystal leads to a breathtaking array of phenomena. It's a fundamental design principle that leaves its fingerprints everywhere, from the way a crystal reflects X-rays to the very existence of exotic electronic states that may power the next generation of quantum technologies. Let us embark on a journey to see where these hidden symmetries reveal their power.

Perhaps the most direct way to "see" a symmetry is to watch how waves interact with a crystal. When we shine X-rays on a crystal, we are performing a kind of cosmic survey, mapping out its internal atomic arrangement. The resulting diffraction pattern of bright spots is a portrait of the crystal's periodic structure. Simple rules, like those for face-centered or body-centered lattices, tell us which reflections to expect. But nonsymmorphic crystals hold a surprise.

Imagine a wave scattering off one layer of atoms. A glide plane dictates that another identical layer exists, but it's been reflected and then shifted by half a lattice spacing. For a very specific angle of incoming X-ray—a specific Miller index $(h,k,l)$—the wave scattering from this second, shifted layer will travel a path that is exactly half a wavelength different from the first. The two waves arrive at the detector perfectly out of phase, and they annihilate each other. A bright spot we expected to see is mysteriously absent.

This phenomenon, known as a systematic absence or extinction, is a smoking gun for a nonsymmorphic symmetry. The famous diamond structure provides a classic example. While its underlying face-centered cubic lattice dictates that reflections must have Miller indices that are all even or all odd, the nonsymmorphic "diamond glide" adds a stricter condition. It extinguishes certain all-even reflections, such as $(2,0,0)$, because their indices do not sum to a multiple of four. This is not a subtle effect; it is a clear, unambiguous signature in the diffraction data that tells us the crystal's symmetry is more complex than a simple stacking of unit cells. Similarly, in the mineral olivine, a glide plane systematically erases reflections of the type $(hk0)$ unless the index $h$ is an even number. These missing reflections are not flaws; they are vital clues, the crystal's way of telling us about its hidden, nonsymmorphic order.

Electrons, being waves themselves, are also subject to these rules. This leads to even deeper consequences for a material's electronic properties. A powerful modern technique called Angle-Resolved Photoemission Spectroscopy (ARPES) acts like a high-resolution camera for the electronic states in a crystal. It can measure a band's energy and momentum, but it can also be sensitive to the shape, or parity, of the electron's wavefunction. In a nonsymmorphic crystal, it's possible to find two electronic bands that are degenerate—they have the exact same energy at a specific momentum on the Brillouin zone boundary. They appear as a single feature. However, the nonsymmorphic symmetry can demand that these two seemingly identical states have opposite parity; one is "even" and the other is "odd" with respect to the mirror operation of the glide. By tuning the ARPES experiment to be sensitive only to, say, the even states, a remarkable thing happens: only one of the two degenerate bands appears. The other is completely invisible, a "dark" state hiding in plain sight. This beautiful experiment confirms that nonsymmorphic symmetry not only dictates which energies are allowed, but it also sculpts the very quantum mechanical nature of the electronic wavefunctions themselves.

The most profound consequences of nonsymmorphic symmetries lie in their influence on the electronic band structure, the "rules of the road" for electrons traveling through a crystal. Here, the fractional translation becomes a master architect, forcing bands to connect and stick together in ways that are forbidden in simpler, symmorphic crystals.

To grasp this, let's consider a simple thought experiment. Imagine a one-dimensional chain of atoms described by a tight-binding model. In a simple symmorphic crystal with one atom per unit cell, the energy band is generally non-degenerate everywhere. Now, let's build a nonsymmorphic version with two atoms per unit cell, related by a glide symmetry. The glide's fractional translation of half a lattice constant, $a/2$, introduces a crucial momentum-dependent phase into the coupling between the two sublattices. At the boundary of the Brillouin zone, where the crystal momentum is $k=\pi/a$, this phase makes the coupling term between the two atoms mathematically vanish. With no way to interact, the two states become degenerate. This is not an accident; the symmetry absolutely requires the two bands to "stick together" at the zone boundary, doubling the degeneracy from one to two.

This "band sticking" is a universal feature. The mathematical reason is deep and elegant. At the zone boundary, the algebra of the symmetry operations changes. The operators representing two nonsymmorphic symmetries might no longer commute; instead, they might be forced to anticommute ($AB = -BA$). If two operators $A$ and $B$ anticommute, they cannot be simultaneously diagonalized, which forbids a non-degenerate eigenstate and guarantees at least a two-fold degeneracy. This shift in the algebraic structure at the zone boundary is captured by the theory of projective representations, which provides the rigorous framework for why nonsymmorphic symmetries can forbid one-dimensional representations, thereby forcing degeneracies.

This phenomenon of enforced band sticking is not just a theoretical nicety; it is the fundamental ingredient for a new class of materials known as topological materials. These materials exhibit exotic electronic properties that are protected by the underlying symmetries of the crystal lattice.

​​Pinning Down the Dirac Point:​​ In some materials, like graphene, the conduction and valence bands touch at isolated points in the Brillouin zone, known as Dirac points. Near these points, electrons behave as massless relativistic particles. Often, these points are protected by a combination of symmetries, like time-reversal and inversion, but they are not necessarily fixed in momentum space. Nonsymmorphic symmetries provide a different, more rigid protection mechanism. They can create Dirac points that are unmovable, "pinned" to high-symmetry lines or planes on the boundary of the Brillouin zone. This rigidity makes the properties of these nonsymmorphic Dirac semimetals incredibly robust and predictable.

​​The Hourglass Dance:​​ The story becomes even more enchanting when nonsymmorphic glide or screw symmetries combine with time-reversal symmetry, a fundamental constraint for any material with spin-orbit coupling. Let us follow the energy bands along a path from the center of the Brillouin zone ($\Gamma$ point) to the edge ($X$ point).

1. At the $\Gamma$ point, the combined symmetries dictate that a Kramers pair of degenerate electrons must have opposite eigenvalues under the glide symmetry.
2. At the $X$ point, the same symmetries conspire to demand that a Kramers pair must have the same glide eigenvalue.

Now, we have a puzzle. The energy bands must be continuous. How can a pair of states start with opposite symmetry labels at $\Gamma$ and arrive at $X$ with identical labels? It seems impossible, like trying to continuously transform your left hand into your right hand. The resolution is a beautiful "partner switching." A band that starts at $\Gamma$ as part of one Kramers pair must cross another band to connect with a different partner at $X$. This enforced crossing, which is forbidden in simpler crystals, traces a pattern that looks remarkably like an hourglass. This "hourglass fermion" is not an accident; it is a topologically protected feature of the band structure that cannot be removed unless one of the underlying symmetries is broken. When viewed in an extended zone scheme, this connectivity reveals a "Möbius twist" where a continuous band crosses a shifted copy of itself, with the crossing protected because the two branches carry opposite glide eigenvalues.

​​Nodal Lines:​​ The wonder does not stop there. This hourglass crossing was guaranteed along a specific high-symmetry line in the Brillouin zone. If we now allow the momentum to vary in the plane containing this line, the protection mechanism still holds. The crossing point simply shifts, sweeping out a continuous line of degeneracies—a nodal line. The existence of this robust, one-dimensional manifold of massless electronic states is a direct and spectacular consequence of the crystal's nonsymmorphic nature.

From missing reflections in a simple diffraction pattern to the guaranteed existence of exotic topological quasiparticles, nonsymmorphic symmetry reveals itself as a deep and powerful organizing principle in the quantum world. What begins as a subtle rule in crystallography blossoms into a key ingredient for engineering the materials of the future. It is a testament to the profound and often surprising unity of physics, where the abstract language of symmetry writes the concrete laws governing the behavior of matter.