Volume 103, Issue 1

The third isotope of the third most abundant element, 17 O, records indispensible information on the origin and operation of Earth, the third planet. The measured uniformity in fractionation of 16 O, 17 O, and 18 O in rocks and minerals over the whole of geologic time, from Hadean to Quaternary, records the existence of a global magma ocean prior to the formation of continents. New techniques of high-resolution mass spectroscopy and of femtosecond X-ray diffraction are leading toward a deep understanding of the origin of kinetic isotope fractionation effects during metabolism. Analysis for the rare molecule 17 O 18 O, distinguished by the substitution of two heavy isotopes, in combination with data on 18 O 18 O, provides an insight into the mechanism whereby plants produce oxygen. Given the skills of American Mineralogist readers in three-dimensional visualization of complex crystalline and molecular structures and the talents of biogeochemical colleagues in measuring isotope fractionation by organisms in nature, there is every reason to expect extraordinary advances in understanding the cycling of life’s elements, H, C, N, O, and S between the biosphere, atmosphere, hydrosphere, and lithosphere.

The December 1974 flow in the SW rift zone at Kilauea Volcano, Hawaii, has been established as a Mars analog due to its physical, chemical, and morphological properties, as well as its interaction with the outgassing plume from the primary Kilauea caldera. We focus on a solfatara site that consists of hydrothermally altered basalt and alteration products deposited in and around a passively degassing volcanic vent situated directly adjacent to the December 1974 flow on its northwest side. Reflectance spectra are acquired in the visible/near-infrared (VNIR) region and emission spectra in the mid-infrared (MIR) range to better understand the spectral properties of hydrothermally altered materials. The VNIR signatures are consistent with silica, Fe-oxides, and sulfates (Ca, Fe). Primarily silica-dominated spectral signatures are observed in the MIR and changes in spectral features between samples appear to be driven by grain size effects in this wavelength range. The nature of the sample coating and the thermal emission signatures exhibit variations that may be correlated with distance from the vent. Chemical analyses indicate that most surfaces are characterized by silica-rich material, Fe-oxides, and sulfates (Ca, Fe). The silica and Fe-oxide-dominated MIR/VNIR spectral signatures exhibited by the hydrothermally altered material in this study are distinct from the sulfate-dominated spectral signatures exhibited by previously studied low-temperature aqueous acid-sulfate weathered basaltic glass. This likely reflects a difference in open vs. closed system weathering, where mobile cations are removed from the altered surfaces in the fumarolic setting. This work provides a unique infrared spectral library that includes martian analog materials that were altered in an active terrestrial solfatara (hydrothermal) setting. Hydrothermal environments are of particular interest as they potentially indicate habitable conditions. Key constraints on the habitability and astrobiological potential of ancient aqueous environments are provided through detection and interpretation of secondary mineral assemblages; thus, spectral detection of fumarolic alteration assemblages observed from this study on Mars would suggest a region that could have hosted a habitable environment.

Kiruna-type iron oxide–apatite associations occur in a variety of rock types and their origin has remained controversial. Most of the Kiruna-type deposits are associated with intermediate to felsic rocks, and in rare cases with ultramafic rocks. Here we investigate the Seruwila iron oxide–apatite deposit at the contact between the Highland and Vijayan complexes that has been defined as the “eastern suture” in Sri Lanka, which formed during the late Neoproterozoic assembly of the Gondwana supercontinent. The ore deposit is hosted in an ultramafic intrusion and comprises massive and disseminated mineralization. The ore-bearing rocks are mainly composed of low-Ti magnetite and chlor-fluorapatite. Our petrological and geochemical studies suggest a magmatic–hydrothermal model for the mineralization wherein: (1) the Cl-rich magmatic–hydrothermal fluid scavenged iron and P from the ultramafic magma, transported iron to shallower levels in the crust and deposited along the suture zone to form the massive type magnetite and apatite; and (2) the cooling of the hydrothermal fluids resulted in the growth of disseminated magnetite and the precipitation of sulfide minerals, followed by a calcic metasomatism (scapolitization and actinolitization). This model is in conformity with the genetic relation between Kiruna-type deposits and iron oxide–copper–gold (IOCG) deposits. We also report LA-ICP-MS zircon U–Pb ages from the host ultramafic intrusion suggesting its emplacement at ca. 530 Ma, which is younger than the regional high-grade metamorphism associated with the collisional assembly of the crustal blocks in Sri Lanka at ca. 540 Ma. By analogy with the common occurrence of Kiruna-type deposits in extensional tectonic settings, and the geochemical features of the studied rocks including low silica, high Mg, Fe, Ca with high field strength elements (HFSEs such as Nb, Ta, Zr, Hf, Ti) depletion and strong LREE and F enrichment, we theorize that the ultramafic magmatism occurred in a post-collisional extensional setting derived from a volatile- and LREE-rich metasomatized lithospheric mantle.

Many studies have proved the usefulness of Li-mica and chlorite geochemistry as indicators of the chemical and thermal evolution of magmatic systems. This study highlights the suitability of Li-micas as tracers of hydrothermal mineralizing events in world-class W-Sn deposits associated with Jurassic (190–150 Ma) granites in China through the complex magmatic–hydrothermal evolution of the Piaotang deposit (South Jiangxi). A paragenetic sequence has been established for the Piaotang deposit comprising (1) a first “silicate-oxide” stage that hosts abundant W-Sn mineralization (wolframite and cassiterite), (2) a “calcic” stage with scheelite and wolframite, (3) a “base metal sulfides” stage with cassiterite and wolframite, and (4) a late “sulfide” stage, involving for the first time a polyphase emplacement of the mineralization. Li-micas from the underlying granite, greisen, and the different stages represented in the veins, were studied. The chemistry of the micas (characterized by intermediate compositions between phlogopite-zinnwaldite-muscovite poles) demonstrates the presence of end-members representing three different fluids that were involved in the emplacement of the Piaotang deposit. These end-members can be linked to previous fluid inclusion studies conducted on this deposit. The three fluids are identified to be magmatic, meteoric (as previously reported in the literature), and also metamorphic, and are shown to have mixed throughout the different stages. Moreover, it appears that the magmatic fluids could not have been derived from the Piaotang biotite granite but instead must have originated from a more evolved rare metal granite that is presently unidentified. These fluids were responsible for the greisenization. Finally, chlorite geochemistry reveals the occurrence of a heating process (from 200 °C in stage II to 300 °C in stage III) during the post-mineralizing stages, which was responsible for the precipitation of new generations of ore-bearing minerals (cassiterite and wolframite) concomitant with a continuous gain of metals during the emplacement of the Piaotang deposit.

Minor amounts of markedly Cl-rich amphibole and biotite are found in the Archean (2.8–2.9 Ga) iron-formation lithologies from the eastern Beartooth Mountains, Montana, U.S.A. These rocks are typified by mineral assemblages of quartz + magnetite + orthopyroxene + garnet ± clinopyroxene ± plagioclase having equilibrated during granulite facies conditions of ~775–800 °C and 6–8 kbar. The metamorphic Cl-rich amphibole and biotite are prograde and occur as inclusions in orthopyroxene and garnet as well as in the matrix. The high-grade Fe-rich amphiboles (mostly Cl-rich potassic-hastingsite and magnesio-ferri-hornblende) and biotites contain concentrations of Cl reaching up to 2.9 and 3.4 wt%, respectively. Biotites contain up to 10.5 wt% BaO and 6.9 wt% TiO 2 . Substitution of Cl into amphibole and biotite is more likely where Cl-bearing anion sites are enlarged. In amphibole threshold values of the X Fe 2+ , A K, and T Al appear to be a precondition before significant amounts of Cl are incorporated into the structure. In biotite, in addition to X Fe 2+ , substitution of T Al also is positively correlated with Cl. The high Cl content in the minerals is suggestive of a coexisting high salinity aqueous fluid. Based on calculations using mineral chemistry, estimations of the aqueous fluid composition indicate a Cl-rich aqueous fluid (~25 wt% NaCl)with f H 2 O / f HCl ratios of ~0.68–0.82. There is evidence for brine-CO 2 immiscibility during peak metamorphism. In amphibole once the threshold values for significant Cl incorporation are attained, the Fe 2+ -Mg partitioning of the amphibole and the coexisting mafic silicates changes such that the amphibole more favorably partitions Fe 2+ . A feedback mechanism can be generated such that the more Cl available from a fluid the more Fe 2+ -rich the amphibole can become, and this produces a crystal structure that can accommodate more Cl, which makes this amphibole more favorable for Fe 2+ incorporation, and the cycle continues until Cl saturation in the crystal structures is reached or aqueous Cl – is fully equilibrated with the amphibole.

“Super-deep” diamonds are thought to have a sub-lithospheric origin (i.e., below ~300 km depth) because some of the mineral phases entrapped within them as inclusions are considered to be the products of retrograde transformation from lower-mantle or transition-zone precursors. CaSiO 3 -walstromite, the most abundant Ca-bearing mineral inclusion found in super-deep diamonds, is believed to derive from CaSiO 3 -perovskite, which is stable only below ~600 km depth, although its real depth of origin is controversial. The remnant pressure ( P inc ) retained by an inclusion, combined with the thermoelastic parameters of the mineral inclusion and the diamond host, allows calculation of the entrapment pressure of the diamond-inclusion pair. Raman spectroscopy, together with X-ray diffraction, is the most commonly used method for measuring the P inc without damaging the diamond host. In the present study we provide, for the first time, a calibration curve to determine the P inc of a CaSiO 3 -walstromite inclusion by means of Raman spectroscopy without breaking the diamond. To do so, we performed high-pressure micro-Raman investigations on a CaSiO 3 -walstromite crystal under hydrostatic stress conditions within a diamond-anvil cell. We additionally calculated the Raman spectrum of CaSiO 3 -walstromite by ab initio methods both under hydrostatic and non-hydrostatic stress conditions to avoid misinterpretation of the results caused by the possible presence of deviatoric stresses causing anomalous shift of CaSiO 3 -walstromite Raman peaks. Last, we applied single-inclusion elastic barometry to estimate the minimum entrapment pressure of a CaSiO 3 -walstromite inclusion trapped in a natural diamond, which is ~9 GPa (~260 km) at 1800 K. These results suggest that the diamond investigated is certainly sub-lithospheric and endorse the hypothesis that the presence of CaSiO 3 -walstromite is a strong indication of super-deep origin.

Amblygonite–montebrasite series and lacroixite from Nagatare Li–Cs–Ta (LCT) pegmatite, Fukuoka Prefecture, Japan, were investigated by powder X-ray diffraction (XRD), electron microprobe analyses, and transmission electron microscope (TEM)/scanning transmission electron microscope (STEM) analyses. Scattered patchy or lamellar lacroixite was contained in montebrasite and amblygonite in all observed specimens. TEM/STEM observations revealed that the patchy and lamellar texture comprised lacroixite and low-fluorine montebrasite having same crystal orientations as that of host montebrasite and the boundaries corresponded to well-developed {110} planes. The observed microtexture was newly discovered, and it is an important evidence of the exsolution process. In XRD experiments conducted at high temperature, the unit-cell parameters of amblygonite were closer to that of monoclinic structures such as lacroixite with increasing temperature. Results suggested that scattered patches or lamellae of lacroixite were exsolution textures from a high-temperature phase. Montebrasite and amblygonite specimens from other localities involved varying textures corresponding to their occurrence. The amblygonite–montebrasite series from petalite-bearing pegmatite included low to high lacroixite contents and that from lower-temperature pegmatite with spodumene either did not possess or involved low lacroixite contents. Gem-quality montebrasite from drusy vugs formed at low temperature did not include any exsolution texture or lacroixite. The variety of texture of the amblygonite–montebrasite series indicated in this study generated new possibilities as the indicator of pegmatite-forming process.

The sound velocity of hcp Fe 0.89 Si 0.11 (Fe-6wt% Si) alloy was measured at pressures from 45 to 84 GPa and temperatures of 300 and 1800 K using inelastic X-ray scattering (IXS) from laser-heated samples in diamond-anvil cells (DACs). The compressional velocity ( ν P ) and density (ρ) of the Fe-Si alloy are observed to follow a linear relationship at a given temperature. For hcp Fe 0.89 Si 0.11 alloy we found ν P = 1.030 (±0.008) × ρ – 1.45 (±0.08) + [3.8 × 10 −5 ( T – 300)×(ρ – 15.37)], including non-negligible temperature dependence. The present results of sound velocity and density of hcp Fe 0.89 Si 0.11 alloy indicates that 3~6 wt% of silicon in the inner core with additional amount of Ni can explain the compressional velocity ( ν P ) and density (ρ) of the “preliminary Earth reference model” (PREM), assuming a temperature of 5500 K and that silicon is the only light element in the inner core

The newly discovered Nayongzhi Zn-Pb deposit (>20 Mt ores at 1.11–15.65 wt% Zn and 0.59–0.97 wt% Pb) in NW Guizhou province, South China, is hosted by late Ediacaran and early Cambrian carbonate rocks. The ore body is structurally controlled by a kilometer-scale reverse fault-anticline system and occurs as stratiform, lentiform, or steeply dipping vein structures. Its geological feature is comparable to that of the Mississippi Valley-type (MVT) Zn-Pb deposits. δ 34 S values (+11.8 to +33.0‰) of sulfide minerals determined by NanoSIMS have a larger range than those determined by conventional bulk analysis (δ 34 S = +18.12 to +24.79‰). This suggests that S isotopes determined by in situ analysis can reflect the nature of fractionation involved in mineralization. Furthermore, cores of sulfide crystals have higher δ 34 S values (+26.1 to +33.0‰) than their rims (+11.8 to +24.5‰). This implies a mixture of multiple S reservoirs or a Rayleigh fractionation of S isotopes occurred during ore formation process. Additionally, both S isotopic compositions determined by in situ and bulk analyses reflect the enrichment of 34 S in hydrothermal fluid (δ 34 S fluid > +11.8‰), a typical characteristic of marine sulfate-derived S. Such S isotopic signatures also show that thermochemical sulfate reduction (TSR) is the dominant mechanism for the incorporation of S 2− from SO42− $\rm SO_{4}^{2-} $. Pb isotopic ratios of galena obtained by femtosecond LA-MC-ICPMS plot in the field that overlaps with the Pb evolution curve of upper crust contributed to the orogeny and the field of modern lower crust, and can be compared to the Proterozoic metamorphic rocks. The means that the majority of Pb metal is sourced from the basement rocks. Although δ 13 C values (–4.1 to +0.5‰) of calcite separates and corresponding fluids are similar to both fresh limestone (–1.7 to +1.3‰) and typical marine carbonate rocks, the δ 18 O values (+12.4 to +14.1‰) are significantly lower than both limestone (+24.1 to +25.5‰) and marine carbonate rocks. Such C-O isotopic characteristics suggest that the source of C is ore-hosting carbonate rocks, whereas O has a mixed source of metamorphic fluids and carbonate rocks resulting from water/rock (W/R) interaction. This study demonstrates that (1) fluid mixing caused rapid sulfide precipitation, resulting in significant fractionation of S isotopes; and (2) both the W/R interaction and CO 2 degassing controlled local carbonate cyclic process of dissolution → re-crystallization, which provided metastable physical and chemical conditions for giant sulfide mineralization. These two processes are crucial in forming MVT deposits.

Mineral weathering processes in soils are important controls on soil characteristics and on bio- and geochemical cycling. Elucidation of these processes and their mechanisms is crucial for understanding soil environments and their influence globally. An Umbric Podzol from the Falkland Islands was studied while investigating possible ways to counteract soil degradation and loss. The soil had lost the O, E, and Bs horizons through erosion, thus revealing the transitional B/C horizon, which grades into the underlying parent material. Samples were taken from the B/C surface and 5 cm below the surface, then analyzed with X-ray diffraction, scanning electron microscopy with energy-dispersive X-ray spectroscopy, organic C and N analysis, analysis of extractable Fe and Al with the dithionite-citrate-bicarbonate and ammonium oxalate methods, and Fourier-transform infrared analysis. The soil fabric and mineralogy were compatible with derivation from sandstone rock. Clasts of heterogeneous mineral composition as well as loose material from disaggregated clasts were present. The soil had large proportions of quartz and albite, and minor amounts of muscovite, chlorite, plagioclase, feldspar, kaolinite, and non-diffracting Fe oxide (goethite and/or ferrihydrite). The most peculiar characteristic was a large component (~7 wt% of the bulk soil) of an amorphous (non-X-ray diffracting) silicate phase of small particle size (<1 μm), non-extractable, with heterogeneous composition. The average composition of this phase is similar to that of the bulk soil and approaches that of Al-Fe-rich smectite. The amorphous phase is not allophane or imogolite by any of the analyses carried out. The amorphous silicate phase is formed partly by the translocation of metals from O, E, and Bs horizons and partly by dissolution of the primary minerals of the B/C horizon, both of which precipitated in combination with low water mobility causing rapid saturation of the interstitial water. There are no reports of amorphous silicate phases with these characteristics or abundances from soils or other weathering environments. Thus, our observations indicate the existence of complex, successive weathering steps not yet identified that could be investigated in materials subjected to slow weathering such as the soil described here.

The rapidly growing use of rare earth elements and yttrium (REE) in modern-day technologies, not least within the fields of green and carbon-free energy applications, requires exploitation of new REE deposits and deposit types. In this perspective, it is vital to develop a fundamental understanding of the behavior of REE in natural hydrothermal systems and the formation of hydrothermal REE deposits. In this study, we establish a mineralogical, textural, and mineral-chemical framework for a new type of deposit, the hydrothermal Olserum-Djupedal REE-phosphate mineralization in SE Sweden. An early, high-temperature REE stage is characterized by abundant monazite-(Ce) and xenotime-(Y) coexisting with fluorapatite and subordinate amounts of (Y,REE,U,Fe)-(Nb,Ta) oxides. During a subsequent stage, allanite-(Ce) and ferriallanite-(Ce) formed locally, partly resulting from the breakdown of primary monazite-(Ce). Alteration of allanite-(Ce) or ferriallanite-(Ce) to bastnäsite-(Ce) and minor synchysite-(Ce) at lower temperatures represents the latest stage of REE mineral formation. The paragenetic sequence and mineral chemistry of the allanites record an increase in Ca content in the fluid. We suggest that this local increase in Ca, in conjunction with changes in oxidation state, were the key factors controlling the stability of monazite-(Ce) in the assemblages of the Olserum-Djupedal deposit. We interpret the alteration and replacement of primary monazite-(Ce), xenotime-(Y), fluorapatite, and minor (Y,REE,U,Fe)-(Nb,Ta) oxide phase(s), to be the consequence of coupled dissolution-reprecipitation processes. These processes mobilized REE, Th, U, and Nb-Ta, which caused the formation of secondary monazite-(Ce), xenotime-(Y), fluorapatite, and minor amounts of allanite-(Ce) and ferriallanite-(Ce). In addition, these alteration processes produced uraninite, thorite, columbite-(Fe), and uncharacterized (Th,U,Y,Ca)-silicates. Textural relations show that the dissolution-reprecipitation processes affecting fluorapatite preceded those affecting monazite-(Ce), xenotime-(Y), and the (Y,REE,U,Fe)-(Nb,Ta) oxide phase(s). The mineralogy of the primary ore mineralization and the subsequently formed alteration assemblages demonstrate the combined mobility of REE and HFSE in a natural F-bearing high-temperature hydrothermal system. The observed coprecipitation of monazite-(Ce), xenotime-(Y), and fluorapatite during the primary REE mineralization stage highlights the need for further research on the potentially important role of the phosphate ligand in hydrothermal REE transporting systems.

Leesite (IMA2016-064), K(H 2 O) 2 [(UO 2 ) 4 O 2 (OH) 5 ]·3H 2 O, is a new uranyl-oxide hydroxyl-hydrate found underground in the Jomac mine, Brown’s Rim, White Canyon mining district, San Juan County, Utah. Laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) analyses provided the empirical formula K 0.67 Na 0.004 Ca 0.012 U 4 O 20 H 15.31 , based on 4 U and 20 O apfu. Sheets in the crystal structure of leesite adopt the fourmarierite anion topology, and so belong to the schoepite family of related structures that differ in the interlayer composition and arrangement, and charge of the sheet. Leesite may form as one of the principal components of “gummite” mixtures formed during the alteration of uraninite, and the unit cell of leesite resembles the previously described, but poorly understood mineral, paraschoepite. Uptake of dangerous radionuclides ( 90 Sr, 135 Cs, 137 Cs, 237 Np, 238 Pu) into the structure of leesite and other members of the family has important implications for the safe disposal of nuclear waste.

Phase relations in the system Mg 2 SiO 4 –MgCr 2 O 4 were studied at 10–24 GPa and 1600 °C using a high-pressure Kawai-type multi-anvil apparatus. We investigated the full range of starting compositions for the forsterite-magnesiochromite system to derive a P – X phase diagram and synthesize chromium-bearing phases, such as garnet, wadsleyite, ringwoodite, and bridgmanite of a wide compositional range. Samples synthesized at 10 GPa contain olivine with small chromium content and magnesiochromite. Mg 2 SiO 4 wadsleyite is characterized by the pressure-dependent higher chromium solubility (up to 7.4 wt% Cr 2 O 3 ). The maximal solubility of chromium in ringwoodite in the studied system (~ 18.5 wt% Cr 2 O 3 ) was detected at P = 23 GPa, which is close to the upper boundary of the ringwoodite stability. Addition of chromium to the system moves the boundaries of olivine/wadsleyite and wadsleyite/ringwoodite phase transformations to lower pressures. Our experiments simulate Cr-rich phase assemblages found as inclusions in diamonds, mantle xenoliths, and UHP podiform chromitites.

Phase relations in the Mg 2 SiO 4 -MgCr 2 O 4 system were investigated in the pressure range of 9.5–27 GPa at 1600 °C to examine the possible deep mantle origin of ultrahigh-pressure (UHP) chromitites in ophiolites. The experimental results indicate that MgCr 2 O 4 -rich chromite (Ch) coexists with Mg 2 SiO 4 - rich olivine (Ol) below ~13.5 GPa in the equimolar Mg 2 SiO 4 ·MgCr 2 O 4 composition. Above ~13.5 GPa, they react to form a three-phase assemblage: garnet (Gt) solid solution in the Mg 4 Si 4 O 12 -Mg 3 Cr 2 Si 3 O 12 system, modified ludwigite (mLd)-type Mg 2 Cr 2 O 5 phase and Mg 14 Si 5 O 24 -rich anhydrous phase B (Anh-B). At ~19.5 GPa, Anh-B is replaced by Mg 2 SiO 4 -rich wadsleyite (Wd). At 22 GPa, MgCr 2 O 4 -rich calcium titanate (CT) phase coexists with Mg 2 SiO 4 -rich ringwoodite (Rw). The assemblage of CT+Rw changes to CT + MgSiO 3 -rich bridgmanite (Brg) + MgO periclase at 23 GPa. These sequential phase changes indicate that Ch+Ol do not directly transform to CT+Rw but to the three-phase assemblage, Gt+mLd+Anh-B (or Wd), that becomes stable at pressures corresponding to the upper and middle parts of the mantle transition zone. Our results suggest that the UHP chromitites that have been studied so far did not reach transition zone depths during mantle recycling processes of the chromitites, because there is no evidence of the presence of the reaction products of Ol and Ch. If the reaction products, in particular mLd and Anh-B, are found in the UHP chromitites, they are good indicators to estimate the subduction depth of the chromitites.

The main source of carbon entering the deep Earth is through subduction of carbonates, including CaMg(CO 3 ) 2 -dolomite. We examine the high-pressure structure and stability of dolomite to understand the means through which carbon can be sequestered as it enters the deep Earth carbon cycle. Dolomite is investigated to 86 GPa using Raman spectroscopy at room temperature: this includes spectroscopic characterization of dolomite-III, a phase stable at deep mantle pressures and temperatures. Between 63–86 GPa, within the dolomite-III structure, we observe spectroscopic evidence for the evolution of a subpopulation of carbonate ions characterized by weaker C-O bonds, with anomalous pressure shifts: this abnormal bonding change is explained by the onset of a 3+1 coordination of the carbon in some of the carbonate ions in the dolomite-III structure, confirming an earlier prediction of Merlini et al. (2012). The wide suite of carbonate ions (both normal threefold and 3+1 coordinate) within this phase at the highest pressures should give rise to a large variety of cation sites; as such, dolomite-III could represent a major host for incompatible elements in the deep mantle, implying that incompatible element distribution may be closely linked to carbon cycling within the deep Earth.

In situ high-pressure single-crystal X-ray diffraction studies of mesolite, an aluminosilicate composed of stacks of Na + -containing natrolite and Ca 2+ -containing scolecite layers in the ratio of 1:2, showed two discrete steps of pressure-induced hydration (PIH): first H 2 O molecules are inserted into the natrolite layers between ∼0.5 and ∼1.5 GPa and subsequently into the scolecite layers. During the PIH in the natrolite layers, the coordination environment of Na + changes from six to seven, the same as that of Ca 2+ in the scolecite layers. While the natrolite layers behave as in the mineral natrolite, the scolecite layers show a different behavior from the mineral scolecite by adopting the super-hydrated natrolite-type structure at higher pressure, as a larger distortion is not favorable in the 1:2 layered framework. This spatial separation of inserted H 2 O during PIH and the growing structural similarity of the two layers result in a weakening of k ≠ 3 n reflections maintaining the 1:2 layer configuration. Our study of this unique behavior of mesolite provides a simple model of structuration under pressure, and the implications of our experimental findings are discussed.