CMMMech
CMMMech > Mineral > Stevensite

Stevensite mineral

Stevensite is a Mg-rich smectite name listed with questionable status Q, original-description year 1873, in the September 2026 IMA list; it is not an approved or grandfathered entry. The listed formula is (Ca,Na)xMg3-ySi4O10(OH)2. The experimental material below is a synthetic, poorly crystalline Mg-rich trioctahedral smectite assigned by the authors as likely stevensite, not independently established phase-pure mineral. This owner is distinct from saponite, hectorite, amorphous Mg-silicate, and its carbonate transformation products. No constituent criticality designation is inherited.

Microbial & Biochemical Mechanisms (2)

Community-associated low-abundance carbonate coating on source-assigned stevensite

biomineralization Microbial

A mixed microbial-mat incubation produced a low-abundance dypingite coating on synthetic stevensite in Last Chance Lake water. Dissolution followed by carbonate precipitation is the proposed transformation. Spatial association and comparison arms support a community-associated effect, without isolating a causal taxon, proving biological necessity, or tracing the coating magnesium exclusively to the clay.

Experimental Context & Analytical Limits:
Methods 2.2-2.5 used filtered (0.22 micrometre) lake water amended with BG11 components and mixed mat inoculum from Milk Lake, British Columbia. The mat was collected in June 2018 and waters in August 2020; approximately 0.15 g wet mat was inoculated into 200 mL amended water. The uncharacterized community contained cyanobacteria, heterotrophs and diatoms; no isolate-level taxon is assigned. BIO1, BIO2 and BIO3 used Milk, Goodenough and Last Chance waters respectively. Incubations were exposed to atmosphere with monthly deionized-water replacement. Dypingite was observed at six months in BIO3. The experiment extended approximately one year; some Milk Lake observations extend to 13 months. No-biofilm clay controls began five months later, were foil-covered and reached seven months, so light and duration were not matched. Biofilm-only controls lacked added clay. BG11 and lake water also supplied magnesium. This was a laboratory experiment, not an industrial carbon-capture or mineral-recovery process.
Curated Evidence Points:
Sections 3.3, 4.1 and 4.3, Figure 5 and Table 3 report dypingite associated with BIO3 clay; BIO1 and BIO2 showed no mineralogical change, and BIO3null no additional precipitate. BIO3control produced gaylussite and trona, not the same clay coating. BIO1control formed an unresolved Mg-silicate, which is not assigned to stevensite here. The different control timing and illumination prevent a strict living-cell-necessity inference.
Figures S5 and S6 show the weak dypingite diffraction feature and Mg-C-bearing EDS interaction volumes, respectively; the platy coating morphology is shown in main Figure 5. The small sample and very low carbonate abundance prevented quantitative phase analysis. Silicon and sodium in EDS may include underlying clay. The EDS concentration inset in Figure S6 reports lower, not higher, Mg weight percentage in the coating volume than the starting clay; the main text’s Mg-enrichment wording is therefore not adopted.
Section 3.1 assigns likely stevensite using broad diffraction, a 1.53 angstrom 060 reflection, Mg/Si 0.73 and treated basal-spacing changes: 13.95 to 14.53 angstrom with ethylene glycol, and 12.32 after heating and glycol treatment. These support the authors’ poorly crystalline smectite interpretation, not a refined pure-phase identification. Section 4.1 proposes photosynthesis, heterotrophy and EPS effects but does not separately measure those causal contributions; Section 4.3 notes that borosilicate vessels may also supply silicon.
Figures S1-S4 document treated natural/synthetic clay diffraction, biotic bottle photographs and abiotic solid diffraction; Table S1 characterizes digested synthetic clay. Tables S2-S3 contain the abiotic and biotic solution measurements. Quartz is attributed to preparation contamination. Evaporation and sampling complicate concentration comparisons. Main Tables 3 and 4 reverse the final BIO3/BIO3null Mg values (0.70 and 0.95 mmol/L), so no quantitative Mg-removal efficiency or clay-specific carbonate yield is adopted. Theoretical carbon-storage capacity and modeled saturation are not measured capture.

Abiotic dissolution and unsuccessful carbonate precipitation during short pH swings

other

Acidification dissolved some Mg and Si from synthetic source-assigned stevensite, but subsequent pH increase did not produce a detected carbonate phase in these short abiotic experiments. Reprecipitation or sorption can explain some solution changes without proving a particular new silicate phase.

Experimental Context & Analytical Limits:
Methods 2.3 and Tables 1-2 used approximately 1 g synthetic clay per 100 mL solution, with 10% CO2 in nitrogen or hydrochloric acid and subsequent sodium-hydroxide pH adjustment. Goodenough Lake water and deionized-water arms were examined over approximately 28 hours. These abiotic comparisons were not duration- and medium-matched controls for the year-long mat experiment.
Curated Evidence Points:
Sections 3.2 and 4.2 and Figures 2-4 report dissolution and no carbonate detected by XRD, SEM or FTIR. Saturation modeling did not constitute observed carbonate formation. The null result is specific to the supplied recipes and duration.
Table S2 preserves the abiotic solution measurements, Table S4 the modeled carbonate saturation indices, and Figures S2 and S4 the synthetic-clay and abiotic solid diffraction. The ABIOswing 207-minute pH is 7.27 in Table 1/Table S2 but 8.21 in Results 3.2.2; initial Goodenough alkalinity is 609 mEq/L in methods but 465 in Table S2/main Table 4. No numerical dissolution efficiency is inferred across those inconsistent entries.

Sources & References

  1. ima_2026_09: IMA-CNMNC. The New IMA List of Minerals, September 2026, Stevensite row, PDF p205; Q means questionable. Retained original re-inspected on the access date.
    https://cnmnc.units.it/files/editor/IMA_Master_List_(2026-09).pdf (accessed 2026-10-14)
  2. handbook_stevensite: Handbook of Mineralogy. Stevensite, version 1 (2001). Complete one-page identity and smectite-group assignment.
    https://handbookofmineralogy.org/pdfs/stevensite.pdf (accessed 2026-10-14)
  3. arizaleta_2026: Arizaleta et al. (2026). Microbial and abiotic approaches to carbonating stevensite: what works and what doesn’t. Geochimica et Cosmochimica Acta 414:202-216. Complete 15-page version of record from Umeå University repository, diva2:2045678; Figures 1-5 and Tables 1-4. Online publication 24 December 2025.
    https://doi.org/10.1016/j.gca.2025.12.052 (accessed 2026-10-14)
  4. arizaleta_2026_si: Official 31-page supplementary material to Arizaleta et al. (2026), article S0016703725007446, mmc1.pdf; Figures S1-S7, Tables S1-S7, calculations and PHREEQC input.
    https://ars.els-cdn.com/content/image/1-s2.0-S0016703725007446-mmc1.pdf (accessed 2026-10-14)

Curation History

Timestamp Curator Action Summary
2026-10-14T18:55:06.653612Z codex (GPT-6; codex) CREATE Created evidence-backed mineral identity with complete primary characterization, explicit experimental phase and control limits, and no inherited criticality or deployment claim.
2026-10-14T19:05:11.377487Z codex (GPT-6; codex) EDIT Corrected primary section and supplementary figure/table locators for review issue #263, preserving phase assignments, observations, controls and uncertainty. Initial F1 and expanded F1/F2 observations remain immutable.

Record source: data/records/stevensite.yaml