---
title: Zeta Phases in Materials Systems
url: https://www.emergentmind.com/topics/zeta-phases
type: topic
---

# Zeta Phases in Materials Systems

Searching arXiv for recent and relevant papers on zeta phases across materials systems.
In contemporary condensed-matter, alloy, and ceramic literature, the designation $\zeta$ is applied to several distinct phases whose symmetry, bonding, dimensionality, and stability are strongly system dependent. Reported examples include high-pressure solid oxygen, a bi-layered precipitate in Al–Ag, vacancy-ordered layered carbides and nitrides, a metallic few-layer tellurium allotrope, molecular nitrogen polytypes near megabar pressures, and a non-icosahedral boron allotrope; in colloid science, by contrast, phase behavior has also been analyzed for spheres with a constant zeta-potential, which is an electrostatic boundary condition rather than a crystal phase [2309.13936; 1705.05175; 2508.18455; 1809.00561; 2604.16641; 1702.03804; 1009.6150]. This suggests that “zeta phase” is a conventional, system-specific phase label rather than a universal structural class.

## 1. Taxonomic scope and nomenclature

The literature assigns the $\zeta$ label to phases that are chemically and crystallographically unrelated. In the available reports, the label spans molecular solids under extreme compression, coherent precipitates in alloys, vacancy-ordered layered ceramics, and low-dimensional elemental allotropes. The commonality is terminological rather than structural [2309.13936; 1705.05175; 2508.18455; 1809.00561; 2604.16641; 1702.03804; 1009.6150].

| System | $\zeta$ designation | Defining feature |
|---|---|---|
| Solid oxygen | candidate and predicted room-temperature $\zeta$ phase | molecular O$_2$ structures including Pm, Pnma, P2$_1$/m, P6$_3$/mmc |
| Al–Ag alloy | $\zeta$ precipitate | alternating Ag-rich and Al-rich bilayers on $\{111\}_{\rm Al}$ |
| Layered carbides/nitrides | $\zeta$-like phases | vacancy-ordered $M_{n+1}X_n$ layered hexagonal/trigonal structures |
| Tellurium few-layers | $\zeta$ allotrope | metallic P4/mmm layered phase with square-net sublayers |
| Molecular nitrogen | $\zeta$-N$_2$ and $t\zeta$-N$_2$ | monoclinic C2/c molecular phases near 1 Mbar |
| Boron | $\zeta$-B | orthorhombic Cmce $\alpha$-Ga-type non-icosahedral allotrope |
| Colloidal spheres | constant zeta-potential | electrostatic charging condition governing fluid/bcc/fcc phase diagrams |

A recurrent misconception is that all $\zeta$ phases form a homologous family. The reported data do not support that interpretation. In one case $\zeta$ denotes a metallic molecular oxygen phase; in another it denotes a coherent AgAl precipitate; in another it denotes vacancy-ordered carbides and nitrides with 16 stacking motifs; and in boron it denotes an $\alpha$-Ga-type three-dimensional covalent framework. The terminology therefore functions locally within each phase diagram.

## 2. High-pressure elemental $\zeta$ phases

In solid oxygen, first-principles structure searches and molecular dynamics identify several crystalline candidates for the $\zeta$ phase at 0 K, with space groups Pnma, P2$_1$/m, Pm, and P6$_3$/mmc. The calculations used USPEX coupled to DFT–PBE relaxations and HSE06 enthalpy corrections. Relative to the $\varepsilon$ phase, taken as C2/m O$_8$, the relevant thermodynamic quantity is
$$
\Delta H(P)=H_\zeta(P)-H_\varepsilon(P).
$$
Within HSE06, $\varepsilon$-O$_8$ remains lower in enthalpy for $P\le 111$ GPa, whereas above $P\approx 111$ GPa the Pnma and Pm structures become favored; at 300 K, HSE06-corrected Gibbs free energies place Pm lowest from 111 to 140 GPa. The predicted room-temperature $\zeta$ phase is therefore the 24-atom Pm structure, which retains molecular O$_2$ bonds of about 1.18–1.19 Å and is metallic [2309.13936].

That identification is supported, but not made trivial, by comparison to experiment. PowderCell-simulated X-ray diffraction at 116 GPa gives the best match for Pm to the broad, overlapping peaks near $14.9^\circ$. Raman/IR mode counting disfavors the ideal P6$_3$/mmc model because it yields only three Raman-active modes, whereas experiment in $\zeta$ oxygen shows at least seven modes in the 400–900 cm$^{-1}$ interval. Superconductivity estimates from the McMillan formula give $T_c\approx 1.6$ K for Pnma, $6.2$ K for Pm, $0.02$ K for C2/m, $9.4$ K for P6$_3$/mmc, and $30$ K for P2$_1$/m, to be compared with the measured 0.6 K; none is exact, but Pm remains compatible in order of magnitude [2309.13936].

Molecular nitrogen exhibits a different use of the same label. At 98 GPa and 293 K, $\zeta$-N$_2$ is monoclinic C2/c with $Z=16$ N$_2$ per cell and lattice parameters $a=7.101(2)$ Å, $b=6.520(1)$ Å, $c=4.871(2)$ Å, $\beta=100.05(4)^\circ$. Laser heating of $\zeta$-N$_2$ to 1800–2500 K at 78–98 GPa yields a new polytype, $t\zeta$-N$_2$, also C2/c but with a tripled $c$ axis, $c=14.614(7)$ Å, and $Z=48$. DFT/PBE enthalpies place $t\zeta$ within 2–3 meV/atom of $\zeta$ above 35 GPa and slightly more stable for $P>35$ GPa; Raman spectroscopy reveals additional low-frequency and vibron features, and the phase likely corresponds to the previously reported $\kappa$-N$_2$ [2604.16641].

In boron, $\zeta$ denotes a non-molecular, non-icosahedral allotrope synthesized at 115 GPa and 2100 K. $\zeta$-B adopts the orthorhombic $\alpha$-Ga-type structure, space group Cmce, with $Z=8$ and lattice parameters $a=2.7039(10)$ Å, $b=4.8703(32)$ Å, $c=2.9697(6)$ Å at 115(2) GPa. Each B atom is seven-fold coordinated, with in-plane bond lengths 1.66(1), 1.72(1), and 1.75(1) Å and short interlayer bonds 1.59(1) Å, generating a fully three-dimensional covalent network. The transformation from $\beta$-B is accompanied by a $\sim 7.5\%$ volume collapse, and the second-order Birch–Murnaghan fit at 115 GPa gives $K_{115}=575(65)$ GPa with $V_{115}=39.20(8)$ Å$^3$ [1702.03804].

These three elemental cases already show that the $\zeta$ label covers qualitatively different bonding regimes: molecular metallic oxygen, molecular nitrogen polytypes, and a covalent non-icosahedral boron framework.

## 3. Precipitate and intermediate $\zeta$ phases in alloys

In the Al–Ag system, Zhang et al. reported a new precipitate phase $\zeta$ by scanning transmission electron microscopy. It is a modulated, shear-free layered structure composed of alternating bilayers enriched in Ag or Al. The repeat consists of six Ag-rich and six Al-rich $\{111\}_{\rm Al}$ planes, giving twelve $\{111\}_{\rm Al}$ layers per repeat with the chemical modulation
\[
\ldots({\rm Ag\,Ag})({\rm Al\,Al})({\rm Ag\,Ag})({\rm Al\,Al})({\rm Ag\,Ag})({\rm Al\,Al})\ldots
\]
and periodicity
\[
c_\zeta=N\,d_{111}^{\rm Al}, \qquad N=12.
\]
In a hexagonal description the phase is R$\bar 3$m with $c_\zeta\approx 27.35$ Å experimentally and $a_\zeta\approx 2.88$ Å, while the simplest trigonal unit cell is described as P3 and contains 12 sites with overall AgAl composition in the pure-layer model [1705.05175].

Its formation pathway is explicitly intermediate. Samples oil-quenched and aged at 200 $^\circ$C for 7 days develop large $\varepsilon$–GP zones of about 20–25 nm. In situ heating at 150–200 $^\circ$C inside the TEM shows that after as little as 3 min at 200 $^\circ$C, $\varepsilon \rightarrow \zeta$ ordering appears within the GP zone. On longer times, typically more than 10–20 min or at higher temperature, newly nucleated $\gamma'$ (Al$_2$Ag, HCP) plates form at the edge of $\zeta/\varepsilon$ regions and grow at the expense of $\zeta$, completing the sequence $\varepsilon \rightarrow \zeta \rightarrow \gamma'$ [1705.05175].

The thermodynamic interpretation is that $\zeta$ is a local energy minimum. DFT formation energies per Ag atom give $+89$ meV/Ag for isolated substitutional Ag, $-65$ meV/Ag for an infinite Ag monolayer on $\{111\}_{\rm Al}$, and a minimum separation energy of about $-50$ meV/Ag when two Ag layers are separated by two Al layers along $\langle 111\rangle$. Periodic repetition of that Ag$_2$Al$_2$ motif yields the $\zeta$ structure, with $E_F(\zeta)\approx -89$ meV/Ag. The broader energy sequence is solid solution $\rightarrow$ small Ag clusters $\rightarrow$ planar clusters $\rightarrow \varepsilon$ $(E_F\approx -72\ldots -81$ meV) $\rightarrow \zeta$ $(E_F\approx -89$ meV) $\rightarrow \gamma'$ $(E_F\approx -100\ldots -120$ meV), so $\zeta$ is more stable than $\varepsilon$ but less stable than $\gamma'$ [1705.05175].

The metastability has a direct thermodynamic basis. The strain-energy estimate
\[
E_e=\mu\,\delta^2\,V
\]
gives about 3 meV/atom using $\mu\approx 26$ GPa, $\delta\approx 0.07$, and $V\approx 16$ Å$^3$. Configurational entropy favors the more disordered $\varepsilon$ state, with a Bragg–Williams upper bound
\[
\Delta S_{\rm mix}=-k_B\bigl[X\ln X+(1-X)\ln(1-X)\bigr]\approx 0.67\,k_B
\]
per atom for Al–40 at% Ag, corresponding to $T\Delta S\sim 35$ meV/atom at 200 $^\circ$C. Because this is comparable in magnitude to the enthalpic gain $\Delta H(\varepsilon\rightarrow\zeta)\sim 10$ meV, $\zeta$ remains metastable rather than terminally stable [1705.05175].

## 4. Layered and low-dimensional $\zeta$ phases

A different and more systematic use of the $\zeta$ designation appears in hexagonal, layered carbides and nitrides studied as ultra-high temperature ceramics. In that setting, $\zeta$- and $\eta$-type structures have general formula $M_{n+1}X_n$ with $n=1,2,3$, where $M$ is a transition metal and $X$ is C or N. The defining structural feature of the vacancy-ordered $\zeta$ phase is that every second M–M layer contains a plane of X vacancies. The reported prototype space groups are P6$_3$mc, R$\bar 3$m, P$\bar 3$m1, and P3m1, and stacking is indexed by the Jagodzinski–Wyckoff descriptors $h$ and $c$; for example, $\zeta$-Ta$_4$C$_3$ is described by hcch-hcch-hcch [2508.18455].

High-throughput DFT finds substantial chemical breadth for this motif. Across 11 transition metals and both C and N, 67 previously unreported hexagonal, layered materials lie within 0.1 eV/atom of the convex hull, and 9 lie exactly on the hull. Selected examples on or near the hull are Ta$_4$C$_3$, Nb$_4$C$_3$, and Zr$_4$C$_3$ with $\Delta E_{\rm hull}=0.00$ eV/atom, Nb$_3$C$_2$ at $+0.02$ eV/atom, Mo$_2$C at $+0.03$ eV/atom, and V$_4$C$_3$ at $+0.10$ eV/atom. The authors classify phases as stable if $\Delta E_{\rm hull}=0$, near-stable if $\Delta E_{\rm hull}<0.05$ eV/atom, and synthesizable if $\Delta E_{\rm hull}<0.1$ eV/atom [2508.18455].

Elastic and thermal indicators further motivate the ceramic interpretation. Representative values include $C_{11}=650$ GPa, $C_{33}=750$ GPa, $C_{44}=220$ GPa, $B_{\rm VRH}=345$ GPa, $G_{\rm VRH}=180$ GPa, $E=460$ GPa, and $\Theta_D=780$ K for $\zeta$-Ta$_4$C$_3$. The final melting temperature estimates average a Lindemann model with parsimonious neural networks, and five newly stable candidates are predicted above 2500 K: Nb$_3$C$_2$ at $\sim 2600$ K, several Mo$_2$C stackings at $\sim 2680$–2750 K, and Zr$_4$C$_3$ at $\sim 2550$ K. Additional reported values are $T_m\approx 3500$ K for V$_4$C$_3$, $3400$ K for Nb$_4$C$_3$, and $3110$ K for Ta$_4$C$_3$ [2508.18455].

Few-layer tellurium supplies yet another low-dimensional $\zeta$ allotrope. Wang et al. identified a $\zeta$ phase consisting, at the monolayer level, of three Te sublayers stacked along $z$, each sublayer forming a perfect square net. The symmetry is tetragonal P4/mmm, with $a=b\approx 3.02$–3.15 Å and vacuum spacing chosen above 15 Å in slab calculations. The monolayer is metallic, and the phase remains more stable than the monolayer $\gamma$ phase by about 29 meV/Te; the bilayer is about 35 meV/Te more stable than the corresponding $\alpha$ bilayer. The energy difference with $\alpha$ decreases with thickness and vanishes at four layers, i.e. 12 sublayers, so the crossover thickness is approximately 4 layers [1809.00561].

The electronic and interlayer characteristics of Te $\zeta$ are likewise distinctive. The phase is metallic in PBE+SOC, with bands crossing $E_F$ along $\Gamma$–X, X–M, and M–$\Gamma$, and SOC does not open a gap. Strong interlayer coupling gives layer-dependent quantum-well states; the average interlayer stabilization relative to the monolayer is about $-6$ meV/Te for the bilayer, $-11$ meV/Te for the trilayer, and $-14$ meV/Te for the four-layer system. Charge doping further stabilizes $\zeta$, and for $|\Delta q|\gtrsim 0.05$ e/Te the $\zeta$–$\alpha$ crossover thickness increases beyond 4 layers [1809.00561].

## 5. Thermodynamic and structural criteria of identification

Across these systems, identification of a $\zeta$ phase is primarily a problem of relative stability under constrained thermodynamic variables. In high-pressure oxygen, the central quantities are enthalpy and Gibbs free energy relative to $\varepsilon$-O$_8$,
\[
\Delta H(P)=H_{\rm candidate}(P)-H_{\varepsilon\text{-}O_8}(P),
\]
\[
G(P,T)\simeq [U+PV]_{\rm MD}+\langle E_{\rm HSE}-E_{\rm PBE}\rangle-T\,S_{\rm vib},
\]
with $S$ obtained from the MD-derived vibrational density of states. In layered carbides and nitrides, the corresponding stability measure is the formation enthalpy
\[
\Delta H_f(M_{n+1}X_n)=E_{\rm tot}(M_{n+1}X_n)-[(n+1)\mu_M+n\mu_X],
\]
followed by convex-hull analysis against all known competing compounds for each $M$–$X$ pair. In nitrogen, enthalpies are again referenced directly to $\zeta$-N$_2$ through
\[
\Delta H_{\alpha}(P)=H_\alpha(P)-H_\zeta(P).
\]
These are formally different constructions, but all are used to decide whether a $\zeta$ phase is equilibrium, near-stable, or metastable under specified conditions [2309.13936; 2508.18455; 2604.16641].

Finite-temperature and kinetic effects are often decisive. The oxygen study explicitly combines PBE molecular dynamics, HSE06 thermodynamic perturbation, and vibrational entropy at 300 K, changing the preferred $\zeta$ assignment from a near-degeneracy at 0 K to a specific Pm equilibrium phase above about 111 GPa. In Al–Ag, by contrast, the enthalpic preference for ordering into $\zeta$ is offset by configurational entropy, while formation also depends on vacancy flux and the pre-existence of sufficiently large $\varepsilon$ GP zones. A plausible implication is that the same $\zeta$ label can refer either to an equilibrium phase boundary or to a kinetic waypoint embedded in a transformation sequence [2309.13936; 1705.05175].

Experimental discrimination is correspondingly multimodal. Oxygen relies on agreement among static structure search, X-ray diffraction peak positions, Raman/IR mode multiplicity, and superconducting $T_c$. Nitrogen uses single-crystal X-ray diffraction plus Raman signatures, especially additional low-frequency modes and vibron multiplicity, to distinguish $t\zeta$ from $\zeta$. Boron required single-crystal synchrotron X-ray diffraction under laser-heated diamond-anvil-cell conditions to establish the Cmce $\alpha$-Ga framework. In the Al–Ag alloy, atomic-resolution STEM and in situ annealing establish both structure and transformation path. These cases show that no single signature is sufficient across all materials; the operational meaning of a $\zeta$ phase is fixed by a combination of crystallography, spectroscopy, and free-energy analysis [2309.13936; 2604.16641; 1702.03804; 1705.05175].

## 6. Distinctions, misconceptions, and outstanding problems

An important distinction is between a $\zeta$ phase and a constant zeta-potential. In suspensions of colloidal spheres, Smallenburg et al. studied phase diagrams under the boundary condition
\[
\phi(a)=\phi_0\equiv e\zeta/(k_BT),
\]
within a Poisson–Boltzmann cell model. The resulting renormalized charge $Z^*(\eta,n_s)$ and effective screening $\bar\kappa$ feed a DLVO/Yukawa description whose phase behavior includes fluid, bcc, and fcc domains, as well as re-entrant melting driven by density- and salt-dependent discharge. This is not a crystallographic $\zeta$ phase; it is a charging protocol whose phase consequences are nonetheless mathematically well defined [1009.6150].

Within materials science proper, the literature shows several distinct stability classes. In oxygen, Pm is predicted to be the equilibrium room-temperature $\zeta$ phase above approximately 111 GPa. In Al–Ag, $\zeta$ is a metastable intermediate between $\varepsilon$ GP zones and $\gamma'$. In tellurium, $\zeta$ is the most stable few-layer form up to a crossover thickness of four layers, with the crossover shifted to larger thickness by doping. In carbides and nitrides, $\zeta$ identifies a broad vacancy-ordered structural family with both stable and near-stable members. In nitrogen, $t\zeta$ is a polytype closely competing with $\zeta$ itself. In boron, $\zeta$ is a high-pressure allotrope reached only under extreme conditions [2309.13936; 1705.05175; 1809.00561; 2508.18455; 2604.16641; 1702.03804].

Several unresolved points remain explicit in the reports. For solid oxygen, the Pm assignment still leaves discrepancies such as extra Raman modes and broad XRD peaks, which may arise from mixed-phase coexistence in the 96–124 GPa range and instrumental resolution; higher-pressure, higher-resolution experiments are specifically encouraged. For layered carbides and nitrides, the cited work states that open questions remain about structure, stability, and compositional pervasiveness. For nitrogen, the likely correspondence between $t\zeta$-N$_2$ and the previously reported $\kappa$-N$_2$ reframes earlier phase identification. For Al–Ag, the recognition of $\zeta$ as a vacancy-mediated intermediate implies that controlling vacancy flux or in situ ageing conditions may alter precipitation pathways. Taken together, these issues indicate that the term “zeta phase” often marks regions of dense polymorphism, subtle free-energy competition, and nontrivial kinetic trapping rather than settled crystallographic closure [2309.13936; 2508.18455; 2604.16641; 1705.05175].

Source: https://www.emergentmind.com/topics/zeta-phases