High-Pressure Oxygen Annealing

• High-pressure oxygen annealing is an advanced post-synthesis treatment that uses elevated oxygen pressures to fill vacancies and stabilize fully oxygenated oxide materials.
• This technique precisely controls oxygen content and defect chemistry, thereby tuning the electronic, magnetic, and superconducting properties in compounds like CaCoO3 and YBa2Cu3-xMoxO7-δ.
• Optimal process parameters, including pressure, temperature, and cooling rate, are crucial to prevent unwanted oxygen superstructures or structural damage.

In the studies considered here, high-pressure oxygen annealing denotes post-synthesis oxygen treatment under elevated oxygen pressure, or an equivalent sealed oxidizing environment under very high external pressure, used to increase oxygen content, eliminate oxygen deficiency, or stabilize highly oxidized structures that are not obtained under ordinary oxygen partial pressure. Its documented roles include converting vacancy-ordered precursors into fully oxygenated perovskites, tuning oxygen-related pinning landscapes in cuprates, and inserting interstitial oxygen into layered nickelates; the same method can also over-oxygenate a material, create competing oxygen superstructures, or damage a target structural motif (Osaka et al., 2017, Los et al., 2021, Zhang et al., 3 Feb 2025, Dong et al., 5 Aug 2025).

1. Scope and definition

Within the present literature set, high-pressure oxygen annealing is not a single apparatus-specific protocol but a family of strongly oxidizing post-treatments. One implementation uses a gold capsule containing sample and oxidizer under $8\ \mathrm{GPa}$ in a cubic-anvil apparatus to oxygenate an oxygen-deficient cobaltite precursor into stoichiometric $\mathrm{CaCoO_3}$ (Osaka et al., 2017). Another uses pure oxygen at $255$–$270$ bar and $500\,^\circ\mathrm{C}$ to modify oxygen defect chemistry and pinning in $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$ single crystals (Los et al., 2021). In layered nickelates, post-annealing at $19$–$100$ bar or $200$ bar oxygen is used to vary oxygen content, but the outcome depends strongly on whether the inserted oxygen preserves or reconstructs the host lattice (Zhang et al., 3 Feb 2025, Dong et al., 5 Aug 2025).

A persistent terminological issue is that oxygen annealing and high-pressure oxygen annealing are not interchangeable. In $\mathrm{CaWO_4}$, the reported treatment is annealing under constant flow of pure oxygen at $\mathrm{CaCoO_3}$0 for $\mathrm{CaCoO_3}$1, with no stated elevated pressure; it is therefore oxygen-flow annealing rather than high-pressure oxygen annealing in the strict sense (Sivers et al., 2012). In $\mathrm{CaCoO_3}$2, the key control variable is a fixed total oxygen dose in a sealed evacuated quartz tube, defined by $\mathrm{CaCoO_3}$3, rather than a pressure-controlled equilibrium oxygen anneal (Sun et al., 2013). In top-gate oxide semiconductor transistors, the relevant sequence is conventional $\mathrm{CaCoO_3}$4 post-deposition annealing followed by oxygen-free post-metal annealing, not high-pressure oxygen annealing (Jiang et al., 30 Jun 2026).

2. Process architectures and representative conditions

The experimental realizations reported in the literature are diverse, but the operative commonality is enhanced oxygen chemical potential during a post-synthesis treatment.

System	Oxygen treatment	Reported outcome
$\mathrm{CaCoO_3}$5	$\mathrm{CaCoO_3}$6, $\mathrm{CaCoO_3}$7, $\mathrm{CaCoO_3}$8, Au capsule with $\mathrm{CaCoO_3}$9	Vacancy-filling conversion to orthorhombic perovskite
$255$0	Pure $255$1, $255$2–$255$3 bar, $255$4, slow cooling	Modified pinning landscape; strong benefit in Mo-substituted crystals
$255$5	$255$6, $255$7–$255$8 bar, $255$9 days or $270$0, cooling-path dependent	Either bilayer retention or bilayer damage
$270$1	$270$2 bar $270$3, $270$4–$270$5, $270$6	Stage-1 striped interstitial oxygen order

In $270$7, the route is explicitly two-step: a Brownmillerite-like oxygen-deficient precursor is first synthesized under ambient-pressure conditions, then pulverized, sealed with $270$8 in a gold capsule, and annealed at $270$9 for $500\,^\circ\mathrm{C}$0 under $500\,^\circ\mathrm{C}$1 (Osaka et al., 2017). In YBCO-based crystals, the oxygen treatment is performed directly in pure oxygen at $500\,^\circ\mathrm{C}$2–$500\,^\circ\mathrm{C}$3 bar and $500\,^\circ\mathrm{C}$4–$500\,^\circ\mathrm{C}$5, followed by slow cooling to room temperature at $500\,^\circ\mathrm{C}$6 (Los et al., 2021). In $500\,^\circ\mathrm{C}$7, high-$500\,^\circ\mathrm{C}$8 annealing at $500\,^\circ\mathrm{C}$9 becomes a cooling-path problem as much as a pressure problem: slow cooling for $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$0 day from $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$1 to $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$2 produces a new phase, whereas quick pressure release and cooling in $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$3 from $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$4 to $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$5 followed by furnace cooling preserves the 2222 bilayer phase much more effectively (Zhang et al., 3 Feb 2025).

These process variations indicate that “high-pressure oxygen annealing” is best understood operationally rather than by a single canonical pressure range. The decisive variables are the oxygen source, the oxygen activity achieved during annealing, and the structural response of the host lattice under the full thermal-pressure-cooling path.

3. Chemical functions and structural transformations

The clearest vacancy-filling example is $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$6. Ambient-pressure synthesis yields an oxygen-deficient perovskite described as Brownmillerite-like $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$7, even after sintering in flowing oxygen at $\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$8. High-pressure oxygen annealing then removes oxygen deficiency and stabilizes the fully oxygenated GdFeO$\mathrm{YBa_2Cu_{3-x}Mo_xO_{7-\delta}}$9-type perovskite $19$0, with Co explicitly identified as $19$1 and the sample described as free from oxygen deficiency (Osaka et al., 2017). The structural product is orthorhombic $19$2, not cubic, with strong octahedral tilting and Co–O–Co angles of $19$3 and $19$4. This demonstrates that high-pressure oxygen annealing can simultaneously fill vacancies and stabilize a highly oxidized but strongly distorted lattice.

In $19$5, the chemical role is subtler. In pure Y123, high-pressure oxygen annealing is used to alter oxygen defect structure by increasing oxygen content, introducing interstitial oxygen, and modifying oxygen ordering and vacancy clustering (Los et al., 2021). In Mo-substituted crystals, the same treatment is described as essential for proper oxygenation: it enables formation of fully oxygenated stoichiometric $19$6 dimers in the CuO chains, which then act as effective pinning centers. The paper therefore separates two oxygen-pressure effects: oxygen-defect tuning in pure Y123 and realization of the intended Mo-based defect chemistry in substituted YBCO.

The nickelate case shows that high oxygen activity can shift from vacancy healing to structural reconstruction. In $19$7, high-$19$8 annealing with slow cooling produces a new phase that can be modeled either as a hybrid single-layer–trilayer $19$9 or as tetragonal bilayer $100$0, while STEM directly shows significant single layers and trilayers and describes the process as damage of the bilayer structure (Zhang et al., 3 Feb 2025). In $100$1, high-pressure oxygen annealing does not primarily fill inner-apical vacancies, because the as-grown Pr-substituted material is already nearly stoichiometric; instead it inserts interstitial oxygen into the rock-salt-like interlayer region and produces stage-1 striped interstitial oxygen order with period-$100$2 and period-$100$3 variants (Dong et al., 5 Aug 2025). The formal relation reported for this hyperstoichiometric state is

$100$4

A plausible implication is that high-pressure oxygen annealing is not merely vacancy compensation. Depending on the starting defect chemistry, it can either restore an intended stoichiometric framework, as in $100$5, or create new oxygen sublattices and competing ordered states, as in bilayer nickelates.

4. Verification of oxygen stoichiometry and defect state

High-pressure oxygen annealing is unusually dependent on post-anneal verification, because diffraction alone rarely settles oxygen stoichiometry. The $100$6 study provides a textbook multi-probe example. Synchrotron X-ray diffraction yields a single orthorhombic perovskite phase with $100$7, $100$8, $100$9, and $200$0, and Rietveld refinement with occupancies fixed to $200$1 gives a satisfactory structural model. Thermogravimetric analysis in $200$2 then gives an observed weight loss of $200$3, compared with $200$4 expected for stoichiometric $200$5, leading to the explicit conclusion that the sample is free from oxygen deficiency. Bond-valence analysis further yields $200$6 for Ca and $200$7 for Co using

$200$8

supporting the oxidation state $200$9 (Osaka et al., 2017).

In nickelates, oxygen-sensitive real-space probes are decisive. For $\mathrm{CaWO_4}$0, PXRD and Rietveld refinement show the emergence of extra peaks after $\mathrm{CaWO_4}$1 bar $\mathrm{CaWO_4}$2 annealing with slow cooling, but STEM establishes the more physically transparent picture: abundant short-range intergrowths of single-layer, bilayer, and trilayer Ruddlesden-Popper blocks (Zhang et al., 3 Feb 2025). For $\mathrm{CaWO_4}$3, multislice electron ptychography directly resolves near-full occupancy of the inner-apical oxygen sites in the as-grown material, detects interstitial oxygen columns in the high-pressure-oxygen-annealed state, and maps the associated octahedral tilt reversal and phase separation. EELS complements this by showing a normalized oxygen K-edge prepeak intensity of $\mathrm{CaWO_4}$4 in the as-grown phase and $\mathrm{CaWO_4}$5 after high-pressure oxygen annealing, together with a prepeak redshift, which the authors interpret as substantial hole doping into $\mathrm{CaWO_4}$6-$\mathrm{CaWO_4}$7 hybridized states (Dong et al., 5 Aug 2025).

In YBCO, the annealing effect is inferred mainly from superconducting electrodynamics rather than direct oxygen-site imaging. Magnetization loops, Bean-model current extraction, Kramer analysis, and Dew-Hughes scaling are used to discriminate point-like, surface-like, and volume-like pinning contributions after different oxygen treatments (Los et al., 2021). This suggests that, in oxygen-pressure studies, the relevant verification protocol depends on the material class: formal stoichiometry and cation valence for high-valence perovskites, local oxygen topology for layered nickelates, and vortex-pinning phenomenology for cuprates.

5. Effects on electronic, magnetic, and superconducting properties

The value of high-pressure oxygen annealing is often that it unlocks a property regime inaccessible from oxygen-deficient precursors. In $\mathrm{CaWO_4}$8, the fully oxygenated phase is an antiferromagnetic and incoherent metal. It orders antiferromagnetically at $\mathrm{CaWO_4}$9, has $\mathrm{CaCoO_3}$00, shows a room-temperature resistivity of about $\mathrm{CaCoO_3}$01, and exhibits an electronic specific-heat coefficient $\mathrm{CaCoO_3}$02 (Osaka et al., 2017). The target property set therefore depends directly on the successful stabilization of the Co$\mathrm{CaCoO_3}$03 perovskite.

In YBCO, the effect is conditional on composition. For pure Y123, high-pressure oxygen annealing improves $\mathrm{CaCoO_3}$04 at low temperatures, roughly up to $\mathrm{CaCoO_3}$05, by a factor of $\mathrm{CaCoO_3}$06–$\mathrm{CaCoO_3}$07, but above about $\mathrm{CaCoO_3}$08 the current density becomes smaller than in the lower-pressure-treated crystal, and at $\mathrm{CaCoO_3}$09 $\mathrm{CaCoO_3}$10 is not measurable (Los et al., 2021). In Mo-substituted YBCO, the result is much more favorable: after additional annealing at $\mathrm{CaCoO_3}$11 bar and $\mathrm{CaCoO_3}$12, $\mathrm{CaCoO_3}$13 increases at all measured temperatures over the whole measured field range up to $\mathrm{CaCoO_3}$14, with gains of about $\mathrm{CaCoO_3}$15 at $\mathrm{CaCoO_3}$16 and $\mathrm{CaCoO_3}$17, about $\mathrm{CaCoO_3}$18 at $\mathrm{CaCoO_3}$19 and $\mathrm{CaCoO_3}$20, and an overall summary increase of $\mathrm{CaCoO_3}$21–$\mathrm{CaCoO_3}$22. The paper interprets this as the combined effect of properly oxygenated $\mathrm{CaCoO_3}$23 dimers, additional interstitial oxygen, and a mixed pinning landscape.

In bilayer nickelates, the dominant message is negative. In $\mathrm{CaCoO_3}$24, high-$\mathrm{CaCoO_3}$25 annealing generally makes ambient-pressure transport more metallic and reduces low-temperature upturns, but slow-cooled high-$\mathrm{CaCoO_3}$26 samples show a weaker superconducting transition under external pressure, consistent with damage to the bilayer structure (Zhang et al., 3 Feb 2025). In $\mathrm{CaCoO_3}$27, the contrast is more severe: the as-grown sample is superconducting under $\mathrm{CaCoO_3}$28 with $\mathrm{CaCoO_3}$29 and zero resistance at $\mathrm{CaCoO_3}$30, whereas the high-pressure-oxygen-annealed sample shows no superconductivity up to $\mathrm{CaCoO_3}$31 (Dong et al., 5 Aug 2025). The reported explanation is not simply oxygen disorder. Rather, the ordered interstitial oxygen stripes create quasi-1D lattice potentials, preserve orthorhombic tilt distortions under pressure, and inject excess holes into $\mathrm{CaCoO_3}$32-$\mathrm{CaCoO_3}$33 hybridized orbitals, thereby competing with superconductivity.

6. Tradeoffs, misconceptions, and process limits

The most common misconception is that stronger oxygenation is automatically beneficial. The literature summarized here shows the opposite. In $\mathrm{CaCoO_3}$34, ordinary-pressure oxygen processing cannot stabilize the fully oxygenated Co$\mathrm{CaCoO_3}$35 perovskite, whereas an $\mathrm{CaCoO_3}$36 oxidizing treatment succeeds (Osaka et al., 2017). In Pr-substituted bilayer nickelates, moderate oxygen or ozone annealing is described as improving superconductivity by mitigating oxygen vacancies, but high-pressure oxygen annealing drives the system into a trivial, non-superconducting metallic state by creating ordered interstitial oxygen superstructures (Dong et al., 5 Aug 2025). In $\mathrm{CaCoO_3}$37, the cooling path is decisive: high oxygen pressure with slow cooling damages the bilayer structure, while rapid pressure release and quenching suppress the problematic new phase (Zhang et al., 3 Feb 2025).

A second misconception is that oxygen-rich post-treatment should be classified solely by oxygen atmosphere. Several studies show that pressure, total oxygen budget, and post-oxygen relaxation must be distinguished. In top-gate indium-rich oxide transistors, conventional $\mathrm{CaCoO_3}$38 post-deposition annealing suppresses oxygen vacancies, but excessive oxygen treatment creates donor-like oxygen-rich defects such as oxygen dimers or O–O bonds; an oxygen-free anneal in $\mathrm{CaCoO_3}$39 or vacuum is then used to suppress those O-rich defects (Jiang et al., 30 Jun 2026). In $\mathrm{CaCoO_3}$40, too much oxygen during sealed-tube annealing completely oxidizes the sample and turns it into an insulator (Sun et al., 2013). In $\mathrm{CaCoO_3}$41, flowing-oxygen annealing reduces absorption and scattering but also introduces an additional absorption band around $\mathrm{CaCoO_3}$42, plausibly due to $\mathrm{CaCoO_3}$43 hole centers, while leaving intrinsic light yield essentially unchanged (Sivers et al., 2012).

These examples suggest a general process principle: oxygen activity must be matched to the defect chemistry that is actually present. When the dominant defect is oxygen deficiency, high-pressure oxygen annealing can be indispensable. When the material is already near stoichiometric, the same treatment can create new oxygen-rich defects, interstitial order, or structural reconstruction. In that sense, high-pressure oxygen annealing is best regarded not as a universally improving post-treatment, but as a defect-engineering tool whose success depends on whether it removes the relevant oxygen defect without generating a more consequential one.