Tetratic Phase in 2D Systems

• Tetratic phase is a two-dimensional mesophase characterized by fourfold bond orientational order and short-range positional order.
• It is observed in hard-particle models, vibrated granular media, and colloidal systems, often exhibiting an orientational decay exponent of approximately 1/4.
• The phase’s stability is influenced by particle shape, polydispersity, external fields, and confinement, which affect transitions among isotropic, nematic, and smectic states.

The tetratic phase is a two-dimensional orientationally ordered fluid or mesophase with four-fold symmetry. In its standard form, it combines short-range positional or translational order with quasi-long-range four-fold bond-orientational order, so that the system retains square-like orientational correlations without developing a fully ordered square lattice. Across the literature, tetratic order appears as the four-fold analog of the hexatic phase in defect-mediated melting, but also as a stable liquid-crystalline state in hard-particle models, colloidal monolayers, vibrated granular media, and antiferromagnetic melting problems (Schindler et al., 2018, Dertli et al., 2024, Löffler et al., 2024, Abutbul et al., 2021).

1. Symmetry and order-parameter structure

At the continuum level, the tetratic is associated with a four-fold orientational symmetry. In the formulation of Manyuhina and coauthors, the elementary planar object is a “square” described by two orthonormal head-tail-symmetric directors $n\leftrightarrow -n$ and $m\leftrightarrow -m$ with $n\cdot m=0$, where $m$ is obtained from $n$ by a $\pi/2$ rotation; the corresponding invariant order parameter is fourth-rank and traceless, encoding $D_{4h}$ symmetry (Manyuhina et al., 2014).

Microscopic diagnostics are usually expressed through four-fold orientational or bond-orientational fields. A common local bond-order definition is

$$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$

with $n_j$ the number of bonded neighbors and $\theta_{jk}$ the bond angle relative to a fixed axis. The corresponding global order parameter is

$m\leftrightarrow -m$0

In this language, the tetratic mesophase is defined as a state with only short-range positional order and quasi-long-range four-fold bond-orientational order (Prajwal et al., 2020).

Several papers use directly analogous orientational moments. For hard rounded rectangles, $m\leftrightarrow -m$1 measures nematic order and $m\leftrightarrow -m$2 measures tetratic order; a perfect tetratic has $m\leftrightarrow -m$3 and $m\leftrightarrow -m$4 (Martinez-Raton et al., 2022). In vibrated hard squares, the molecular order variable is $m\leftrightarrow -m$5, where $m\leftrightarrow -m$6 is the orientation angle of square $m\leftrightarrow -m$7, while the bond-order field is obtained from the four nearest neighbors through $m\leftrightarrow -m$8 (Walsh et al., 2015). Schindler and Kapfer used a Voronoi-weighted complex bond-orientational parameter,

$m\leftrightarrow -m$9

for which $n\cdot m=0$0 ranges from $n\cdot m=0$1 to $n\cdot m=0$2 between no local fourfold order and perfect square coordination (Schindler et al., 2018).

2. Correlations, criticality, and defect-mediated melting

The defining correlation function is the four-fold orientational correlator,

$n\cdot m=0$3

or, in the notation of Schindler and Kapfer,

$n\cdot m=0$4

In the tetratic phase, $n\cdot m=0$5 decays algebraically,

$n\cdot m=0$6

whereas in the isotropic phase it decays exponentially. The onset of tetratic order is repeatedly associated with $n\cdot m=0$7, consistent with KTHNY-type or BKTHNY-type expectations for a fourfold “atic” transition (Schindler et al., 2018, Prajwal et al., 2020, Löffler et al., 2024).

For driven quasi-two-dimensional granular matter, Schindler & Kapfer analyzed the fluid–tetratic transition by fitting the four-fold static structure factor to an Ornstein–Zernike form,

$n\cdot m=0$8

and found

$n\cdot m=0$9

together with

$m$0

Their best-fit values, averaged over driving amplitudes $m$1, were $m$2, $m$3, $m$4, and $m$5, while the orientational exponent satisfied $m$6 (Schindler et al., 2018).

The square-lattice extension of KTHNY theory developed by Grampel and Podolsky retains the two-step melting scenario but changes the elastic structure. In addition to the Lamé-type elastic constants, an axial anisotropy modulus $m$7 enters the square-lattice free energy, modifying both the logarithmic and angular interactions between defects. The theory predicts a solid-to-tetratic transition driven by dislocation unbinding and a tetratic-to-liquid transition driven by disclination unbinding; it preserves the universal orientational condition $m$8 at tetratic melting, but it does not retain a universal value of the Young’s modulus at the solid-to-tetratic transition (Grampel et al., 22 Jul 2025).

A central point of comparison is that not all four-fold systems follow the same sequence of continuous transitions. In Dertli and Speck’s hard-rectangle simulations, the isotropic-to-tetratic transition is continuous, but the tetratic-to-smectic transition is weakly discontinuous, with hysteresis, a small non-convex region in $m$9, and coexistence over $n$0 (Dertli et al., 2024). This places the tetratic phase within a broader family of four-fold intermediate states rather than a single universal melting route.

3. Equilibrium realizations in hard-particle and colloidal systems

In constant-pressure Monte Carlo simulations of hard rectangles, Dertli and Speck found four phases in the $n$1 plane: isotropic, nematic, tetratic, and smectic. For $n$2, the sequence is $n$3, whereas for $n$4 the nematic is replaced by a tetratic, giving $n$5. At $n$6, the transition points were reported as $n$7 at $n$8 and $n$9 at $\pi/2$0; the tetratic window closes at $\pi/2$1 in a tricritical meeting of the $\pi/2$2–$\pi/2$3 and $\pi/2$4–$\pi/2$5 lines (Dertli et al., 2024).

Colloidal squares provide an experimental realization of the same four-fold phenomenology. In the monolayer experiment on $\pi/2$6 polymer squares, the isotropic regime showed exponential decay of $\pi/2$7 for $\pi/2$8, the intermediate regime at $\pi/2$9 showed algebraic decay with fitted $D_{4h}$0, and the high-density regime at $D_{4h}$1 showed saturation of $D_{4h}$2, consistent with long-range four-fold bond order. The same work states that it provides the first unambiguous experimental demonstration of the tetratic phase (Löffler et al., 2024).

Binary mixtures enrich the phenomenology further. In Monte Carlo simulations of disks with squares at size ratio $D_{4h}$3, the pure-square tetratic mesophase is stable over $D_{4h}$4, while on the square-rich side of the mixture, $D_{4h}$5, the tetratic window is broadened, extending the upper pressure of stability from $D_{4h}$6 to $D_{4h}$7. In the same system, a mosaic phase appears for $D_{4h}$8 and $D_{4h}$9, containing interspersed tetratic, hexatic, and rhombic-like locally ordered clusters (Prajwal et al., 2020).

Density-functional studies of hard kites and related quasi-square shapes likewise stabilize tetratic order. Martínez-Ratón and Velasco found that perfect squares maximize tetratic stability, that rhombuses support tetratic order for $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$0, and that $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$1-kites have the broadest tetratic interval, with $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$2 in SPT or $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$3 in the improved third-virial theory (Martínez-Ratón et al., 2020). Taken together, these results show that the tetratic is not restricted to a single particle geometry but is most robust for quasi-square excluded-area landscapes.

4. Shape, roundness, polydispersity, and external constraints

For freely rotating hard rectangles, Martínez-Ratón & de las Heras used scaled-particle theory to study continuously length-polydisperse systems. They found that the tetratic phase remains confined to small mean aspect ratio $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$4, that the isotropic–nematic transition becomes strongly first order as polydispersity increases, and that the tetratic phase is slightly destabilized with respect to the nematic one. In their physical interpretation, the tetratic phase is formed predominantly by nearly square rods, and the recommended regime for observing a robust tetratic fluid is small mean aspect ratio $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$5–$$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$6, moderate packing $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$7–$$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$8, and low polydispersity $$\psi_4(\mathbf{r}_j)=\frac{1}{n_j}\sum_{k=1}^{n_j}\exp\bigl[i\,4\,\theta_{jk}\bigr],$$9–$n_j$0 (Armas et al., 2017).

Roundness suppresses the tetratic even more directly. In the rounded-hard-rectangle model with mean roundness $n_j$1 and mean aspect ratio $n_j$2, the tetratic phase is locally favored for $n_j$3, where

$n_j$4

Thus, at $n_j$5, tetratic order is stable upon increasing density from the isotropic fluid only for $n_j$6; for $n_j$7 one recovers $n_j$8, whereas for $n_j$9 one finds $\theta_{jk}$0, and the tetratic region vanishes. The same study reports that polydispersity raises both $\theta_{jk}$1 and $\theta_{jk}$2 spinodal densities slightly but does not qualitatively extend tetratic stability (Martinez-Raton et al., 2022).

Mixtures can either suppress or enhance the tetratic. In binary mixtures of hard superellipses, pure hard squares undergo a second-order $\theta_{jk}$3 transition at $\theta_{jk}$4, but adding as little as $\theta_{jk}$5–$\theta_{jk}$6 long rectangles lowers the $\theta_{jk}$7 bifurcation of the squares to $\theta_{jk}$8–$\theta_{jk}$9, with a stable mixed twofold phase region in $m\leftrightarrow -m$00 where a mixed $m\leftrightarrow -m$01–$m\leftrightarrow -m$02 state coexists with isotropic fluid (Mizani et al., 2020).

External constraints also alter the status of four-fold order. In the density-functional theory and Monte Carlo study of hard rectangles on flat and cylindrical manifolds, the field-free bulk diagram contains stable isotropic, nematic, tetratic, and smectic phases; the tetratic appears for $m\leftrightarrow -m$03 in DFT or $m\leftrightarrow -m$04 in Monte Carlo, typically at $m\leftrightarrow -m$05–$m\leftrightarrow -m$06. By contrast, an aligning field $m\leftrightarrow -m$07 eliminates the true tetratic and generates a binematic phase, with two dominant peaks at $m\leftrightarrow -m$08 and two smaller residual peaks at $m\leftrightarrow -m$09 (Sitta et al., 2017).

Gravity produces a different kind of spatial restructuring. In the local-density DFT for polydisperse hard rounded rectangles, the bulk fluid exhibits isotropic, nematic, and tetratic phases, while sedimentation leads to stacking sequences containing pure tetratic stacks, tetratic layers sandwiched between isotropic and/or nematic layers, and reentrant floating tetratic stacks such as $m\leftrightarrow -m$10–$m\leftrightarrow -m$11–$m\leftrightarrow -m$12 (Eckert et al., 10 Nov 2025). This suggests that tetratic order can be selected not only by bulk density and shape but also by gravity-driven fractionation.

5. Nonequilibrium granular matter, confinement, and disorder

The tetratic phase is not restricted to equilibrium systems. In quasi-two-dimensional vibrated inelastic frictionless spheres, Schindler & Kapfer found a density–amplitude phase diagram with a fluidlike disordered phase, an ordered phase with threefold symmetry, phase coexistence, metastable traveling clusters, anisotropic stable clusters at low amplitude, and a square bilayer state connected to the fluid by BKTHNY-type two-step melting with an intermediate tetratic phase. For moderate amplitudes $m\leftrightarrow -m$13, increasing density yields the sequence fluid $m\leftrightarrow -m$14 tetratic $m\leftrightarrow -m$15 square bilayer solid, while at still higher density or higher amplitude a first-order transition to a three-fold hexagonal bilayer phase intervenes (Schindler et al., 2018).

In the experimental system of vibrated hard squares, the tetratic occupies the density window $m\leftrightarrow -m$16. In this regime, $m\leftrightarrow -m$17 and $m\leftrightarrow -m$18 cross over to algebraic decay, translational order remains short-ranged with $m\leftrightarrow -m$19 growing from $m\leftrightarrow -m$20 at $m\leftrightarrow -m$21 to $m\leftrightarrow -m$22 at $m\leftrightarrow -m$23, and rotational dynamics slows dramatically: $m\leftrightarrow -m$24 with $m\leftrightarrow -m$25 and $m\leftrightarrow -m$26, while translational diffusion remains finite until $m\leftrightarrow -m$27 (Walsh et al., 2015).

Strong confinement changes defect organization. In vertically vibrated monolayers of cylinders with aspect ratio $m\leftrightarrow -m$28, planar anchoring at curved boundaries induces tetratic ordering. In a simply connected circular cavity, topology requires four $m\leftrightarrow -m$29 defects, which appear near the outer wall at the vertices of an approximate square. In an annulus, the Euler characteristic is $m\leftrightarrow -m$30, so topology allows a defect-free distorted tetratic field, but in the experiments the system instead fragments into four azimuthal sectors or “bridges” of smectic order separated by domain walls containing disordered regions that act as residual point defects (Armas et al., 2020).

Quenched disorder can stabilize the tetratic as an intermediate phase. In 2D Hertzian spheres with random pinning, a pinning fraction of $m\leftrightarrow -m$31 broadens the tetratic wedge substantially. At $m\leftrightarrow -m$32, on the left branch of square-crystal melting the tetratic region expands from roughly $m\leftrightarrow -m$33–$m\leftrightarrow -m$34 without pinning to roughly $m\leftrightarrow -m$35 with pinning; on the right branch, where the unpinned system melts directly in a first-order jump, pinning opens a tetratic wedge from about $m\leftrightarrow -m$36 to $m\leftrightarrow -m$37. In the pinned system the square $m\leftrightarrow -m$38 tetratic transition is BKT-type continuous and the tetratic $m\leftrightarrow -m$39 liquid transition is first order (Tsiok et al., 2020).

6. Topology, antiferromagnetic variants, and theoretical limits

Topological constraints make the tetratic especially sensitive to curvature. On the sphere, topology requires total charge $m\leftrightarrow -m$40, so eight $m\leftrightarrow -m$41 disclinations are a natural tetratic defect content. Within the one-constant approximation, Manyuhina and collaborators found that these eight defects minimize their pair interaction by occupying the vertices of a straight cube inscribed in the sphere. When the surface is deformable, the competition between tetratic elastic energy and bending energy favors an intermediate superspheroidal shape, and the instability from the spherical shape occurs for $m\leftrightarrow -m$42 (Manyuhina et al., 2014).

A distinct variant is the antiferromagnetic tetratic. In the hard-sphere-between-plates realization studied by Podolsky and coauthors, strong antiferromagnetic coupling suppresses fundamental dislocations because they generate strings of frustrated bonds, while double dislocations with $m\leftrightarrow -m$43 do not. The resulting tetratic is characterized by exponentially decaying translational correlations and algebraically decaying $m\leftrightarrow -m$44, and it occupies the interval $m\leftrightarrow -m$45 between the solid and the liquid (Abutbul et al., 2021). In the computational treatment of antiferromagnetic square-lattice melting, the same distinction becomes more subtle: because elementary dislocations carry $m\leftrightarrow -m$46 gauge flux, no local thermodynamic order parameter separates antiferromagnetic and paramagnetic tetratic regimes, although algorithmically constructed staggered magnetizations can do so through a nontrivial decoding problem (Weinstein et al., 2024).

Theoretical descriptions do not always capture tetratic order reliably. In hard right isosceles triangles, Monte Carlo simulations show a liquid-crystalline regime with strong four- and eight-fold correlations between $m\leftrightarrow -m$47 and $m\leftrightarrow -m$48, but standard density-functional theory, even when extended to include the exact third virial coefficient, predicts only a uniaxial nematic as the stable liquid crystal. The missing ingredient is identified as higher-order body correlations and, more fundamentally, clustering or self-assembly correlations such as dimers and tetramers (Martinez-Raton et al., 2021).

The broader implication is that the tetratic phase is both a symmetry class and a defect-mediated regime. In some systems it appears in the canonical two-step square-lattice melting sequence; in others it competes with nematic, smectic, hexatic, octatic, or topologically constrained states; and in still others its existence depends sensitively on anisotropy, clustering, confinement, or quenched disorder (Grampel et al., 22 Jul 2025, Dertli et al., 2024, Martinez-Raton et al., 2021).