Gibbsite (Al(OH)3) in the Porto Trombetas bauxite horizon (lower Amazon, Brazil) occurs as at least two isotopically distinct populations, which Horbe (2011) interpreted as one generation formed under conditions resembling the modern Amazonian climate and one formed under a warmer, more seasonal, low-rainwater-recycling regime attributed to the Late Oligocene–Early Miocene. Those temperature estimates rest on oxygen and hydrogen isotope data from mineral separates that may contain unquantified kaolinite and amorphous contamination, on a single chosen gibbsite–water fractionation calibration, and on the assumption that formation water had the isotopic composition of modern local rainwater. The open problem is to determine quantitatively — and isotopically — the formation temperatures and parent-water isotopic compositions (and hence climatic-hydrologic regime) of the distinct gibbsite generations in this allitic profile, controlling contamination, calibration choice, and paleo-water composition.

Background

Intense tropical weathering (lateritization) desilicates rocks and enriches them in aluminium, producing allitic profiles in which a bauxite horizon dominated by gibbsite — the aluminium tri-hydroxide Al(OH)3 — is commonly overlain by a thick kaolinite-rich Oxisol (a deeply weathered tropical soil). Because gibbsite and kaolinite precipitate from soil waters, their stable isotope compositions (oxygen-18/oxygen-16, reported as δ18\delta^{18}O, and deuterium/hydrogen, reported as δ\deltaD, on the VSMOW scale in permil) record both the temperature and the isotopic composition of the water from which they crystallized. If the mineral–water isotope fractionation factor is known as a function of temperature, measured mineral values can in principle be inverted into a formation temperature — but only if the water composition is known independently, since one measurement constrains only the combination of the two.

For kaolinite, the widely used fractionation equations of Sheppard and Gilg (1996) make it a standard paleoclimate archive in deeply weathered terrains. For gibbsite the calibration basis is weaker and divided: the empirical synthesis-and-natural-sample study of Bird et al. (1994), the theoretical calculation of Zheng (1998), and the decade-long low-temperature synthesis experiments of Vitali et al. (2000) for oxygen and Vitali et al. (2001) for hydrogen give different temperature–isotope relations, and no study has reconciled them or tested their disagreement on natural data. Bird et al. (1993) had earlier measured very low gibbsite δ18\delta^{18}O (+6.4 to +7.4 permil) in the Paragominas bauxite of the eastern Amazon and attributed it to strongly depleted formation waters (δ18\delta^{18}O near −9 permil) rather than to high temperature — the mirror image of the high-temperature interpretation, which illustrates how sharply conclusions depend on the assumed water composition.

Horbe (2011) measured δ18\delta^{18}O and δ\deltaD of gibbsite separates from the Porto Trombetas bauxite horizon and of kaolinite from the overlying Oxisol, and computed gibbsite formation temperatures using the Bird et al. (1994) equation and the composition of modern local pristine rainwater (Honório et al., 2010). One gibbsite population resembles modern Amazonian conditions, while the more 18^{18}O-enriched population was interpreted as a second generation formed under a seasonal, low-recycling or coastal-proximal regime during the Late Oligocene–Early Miocene, when the Andes were barely exhumed and western Amazonia was flooded by the Pebas mega-wetland; the kaolinite of the overlying Oxisol records cooler, wetter, less seasonal conditions. Several aspects of these estimates remain to be tested: whether the analyzed gibbsite separates retain kaolinite or amorphous material that isotopically biases the measured values; how sensitive the two-population split is to the choice among the published gibbsite–water calibrations (Bird et al., 1994; Zheng, 1998; Vitali et al., 2000, 2001); and whether the formation water can be assumed to have had the isotopic composition of modern local rainwater. Since then, (U–Th)/He dating — a low-temperature thermochronometric method that dates a mineral by the helium accumulated from uranium and thorium decay — of Fe-(oxyhydr)oxides in lateritic duricrusts (indurated weathering crusts) has begun to date Amazonian bauxitization directly (bauxite formation at 25–15 Ma in central-northern Amazonia; dos Santos Albuquerque et al., 2020, 2025), and Ansart et al. (2025) demonstrated at Manaus how phase proportions quantified by XRD (X-ray diffraction) bound contamination effects on clay isotope values and how EPR (electron paramagnetic resonance) dating of kaolinite and (U–Th)/He dating of goethite resolve successive weathering episodes — tools that could optionally support an attack on the Trombetas gibbsite question, but have not been applied to it.

Problem Statement

For the polygenetic allitic weathering profile at Porto Trombetas (lower Amazon, Brazil) — a gibbsite-rich bauxite horizon overlain by a kaolinitic Oxisol — determine quantitatively, using isotope measurements, the formation conditions of the distinct gibbsite generations that coexist within the bauxite horizon. Specifically: (1) establish whether the two (or more) isotopically distinct gibbsite populations recognized by Horbe (2011) — one with signatures compatible with formation under modern Amazonian conditions, one more 18^{18}O-enriched — are genuine successive generations rather than artifacts of kaolinite or amorphous-material contamination in the analyzed mineral separates; and (2) for each genuine generation, determine its formation temperature and the δ18\delta^{18}O and δ\deltaD of the parent weathering water (and hence the implied climatic-hydrologic regime, such as rainfall seasonality, degree of rainwater recycling, or coastal proximity), with uncertainties that account for the disagreement among the published gibbsite–water fractionation calibrations (Bird et al., 1994; Zheng, 1998; Vitali et al., 2000, 2001) and that do not presuppose the modern rainwater isotopic composition. The Porto Trombetas profile is taken as the representative case study of such profiles; extension of this quantitative analysis to comparable polygenetic lateritic bauxite profiles across Amazonia is not required, although the same question of generation-specific formation conditions applies to them.

References

  1. Horbe, A.M.C. (2011). Oxygen and hydrogen isotopes in pedogenic minerals — Implications for paleoclimate evolution in Amazonia during the Cenozoic. Geoderma 163, 178–184. DOI 10.1016/j.geoderma.2011.03.017, https://doi.org/10.1016/j.geoderma.2011.03.017
  2. Bird, M.I., Longstaffe, F.J., Fyfe, W.S., Tazaki, K., Chivas, A.R. (1994). Oxygen-isotope fractionation in gibbsite: synthesis experiments versus natural samples. Geochimica et Cosmochimica Acta 58, 5267–5277. DOI 10.1016/0016-7037(94)90310-7, https://doi.org/10.1016/0016-7037(94)90310-7
  3. Bird, M.I., Longstaffe, F.J., Fyfe, W.S., Kronberg, B.I., Kishida, A. (1993). An Oxygen-Isotope Study of Weathering in the Eastern Amazon Basin, Brazil. In: Climate Change in Continental Isotopic Records, Geophysical Monograph 78, pp. 295–307. DOI 10.1029/GM078p0295, https://doi.org/10.1029/GM078p0295
  4. Zheng, Y.-F. (1998). Oxygen isotope fractionation between hydroxide minerals and water. Physics and Chemistry of Minerals 25, 213–221. DOI 10.1007/s002690050105, https://doi.org/10.1007/s002690050105
  5. Vitali, F., Longstaffe, F.J., Bird, M.I., Caldwell, W.G.E. (2000). Oxygen-isotope fractionation between aluminum-hydroxide phases and water at <60 °C: results of decade-long synthesis experiments. Clays and Clay Minerals 48, 230–237. DOI 10.1346/ccmn.2000.0480209, https://doi.org/10.1346/ccmn.2000.0480209
  6. Vitali, F., Longstaffe, F.J., Bird, M.I., Gage, K.L., Caldwell, W.G.E. (2001). Hydrogen-isotope fractionation in aluminum hydroxides: synthesis products versus natural samples from bauxites. Geochimica et Cosmochimica Acta 65, 1391–1398. DOI 10.1016/S0016-7037(00)00604-9, https://doi.org/10.1016/S0016-7037(00)00604-9
  7. Sheppard, S.M.F., Gilg, H.A.M. (1996). Stable isotope geochemistry of clay minerals. Clay Minerals 31, 1–24. DOI 10.1180/claymin.1996.031.1.01, https://doi.org/10.1180/claymin.1996.031.1.01
  8. Giral-Kacmarcík, S., Savin, S.M., Nahon, D.B., Girard, J.-P., Lucas, Y., Abel, L.J. (1998). Oxygen isotope geochemistry of kaolinite in laterite-forming processes, Manaus, Amazonas, Brazil. Geochimica et Cosmochimica Acta 62, 1865–1879. DOI 10.1016/S0016-7037(98)00103-3, https://doi.org/10.1016/S0016-7037(98)00103-3
  9. Honório, B.A.D., Horbe, A.M.C., Seyler, P. (2010). Chemical composition of rainwater in western Amazonia — Brazil. Atmospheric Research 98, 416–425. DOI 10.1016/j.atmosres.2010.08.001, https://doi.org/10.1016/j.atmosres.2010.08.001
  10. Ansart, C., Guinoiseau, D., Quantin, C., Calmels, D., Gautheron, C., Agrinier, P., Bouchez, J., Fekiacova, Z., Pinna-Jamme, R., Demri, D., Balan, E., Heller, B., Bollaert, Q., Ataytür, Ö., Allard, T. (2025). Lateritic Cenozoic paleoenvironmental and paleoclimatic conditions in the Central Amazon basin, Brazil, inferred from mineralogy, geochemistry and geochronology. Palaeogeography, Palaeoclimatology, Palaeoecology 662, 112674. DOI 10.1016/j.palaeo.2024.112674, https://doi.org/10.1016/j.palaeo.2024.112674
  11. dos Santos Albuquerque, M.F., Horbe, A.M.C., Danišík, M. (2020). Episodic weathering in Southwestern Amazonia based on (U–Th)/He dating of Fe and Mn lateritic duricrust. Chemical Geology 553, 119792. DOI 10.1016/j.chemgeo.2020.119792, https://doi.org/10.1016/j.chemgeo.2020.119792
  12. dos Santos Albuquerque, M.F., Horbe, A.M.C., Gautheron, C., Vieira, J.G.C., Val, P., Pinna-Jamme, R., Salazar, C.A. (2025). (U–Th)/He geochronology of Fe and Fe–Al lateritic duricrusts and bauxites in central-northern Amazonia: implications for paleoclimate and landscape evolution. Chemical Geology 689, 122835. DOI 10.1016/j.chemgeo.2025.122835, https://doi.org/10.1016/j.chemgeo.2025.122835
  13. Vho, A., Lanari, P., Rubatto, D. (2019). An internally-consistent database for oxygen isotope fractionation between minerals. Journal of Petrology 60, 2101–2129. DOI 10.1093/petrology/egaa001, https://doi.org/10.1093/petrology/egaa001