Deep explosive focal depths during maar forming magmatic-hydrothermal eruption: Baccano Crater, Central Italy

Abstract

We describe the eruptive activity of the Pleistocene composite Baccano maar crater in the Sabatini Volcanic Complex (Central Italy) combining stratigraphy, grain size/componentry and rare earth element and Yttrium (REY) composition of its eruptive products with the stratigraphy and geothermal data derived from deep wells drilled on the Baccano structural high. The main lithological characteristics of the basal Baccano maar pyroclastic deposit, composed of more than 60% wt of non-thermometamorphosed lithic clasts from the sedimentary basement, show that the first eruption was magmatic-hydrothermal in nature. The lithology of the sedimentary lithic clasts indicates that the fragmentation level was at a depth of −1,000 to −1,200 m, with fragment depth verified by deep well stratigraphy. The 15% wt juvenile non-vesicular glass components suggest that magma played a minor role in powering the eruption. Assuming that the high-salinity hot hydrothermal fluids (365<T<410°C and P∼25 MPa), hosted in the highly permeable and confined aquifer below the Baccano maar are representative of those at the time of the eruption, we propose that hydrofracturing would have triggered the eruption caused by overpressure at the top of the geothermal aquifer. REY analysis performed on pyroclastic fragments and basement rocks suggest that partial dissolution of the deeper limestones (>−1,400 m) by the aggressive hydrothermal fluids enriched in acid components (HF, HCl, and H2SO4) may have contributed to increased CO₂ partial pressure that helped to drive the hydrofracturing. This could have caused rapid vapour separation and pressure drop, allowing the almost simultaneous breaking of the aquifer cover and brecciation of the calcareous units down to −1,000 to −1,200 m depth. The relative abundance of calcareous lithics in the basal part of the first Baccano eruptive unit, representing about the upper 200 m of stratigraphy below the top of the Baccano structural high, reveals the descent of the piezometric surface during the eruption. Combining deep well information and maar product stratigraphy, using also REY data from maar pyroclastic fragments and the basement rocks we draw an interpretative model for the Baccano maar-forming eruption, concluding that a) magmatic-hydrothermal eruptions may originate deeper than previously thought, and b) hydrothermal fluids circulating in limestone aquifers may play an important role in triggering such eruptions.

Appendix: Analytical procedure used for REY analysis

After the preliminary grinding and homogenization, samples were dissolved by means of microwave-assisted acid dissolution (with a FKV—Milestone—MLS 1200 Mega). Each dissolution was executed on approximately 400 mg of sample, accurately weighed (directly) in Teflon vessels, using the following acid mixture: 4 ml HNO₃ + 1 ml H₂O₂ + 2 ml HF + 1 ml HClO₄. All these reagents were of high purity grade (BDH-Aristar®). The microwave cycle was prearranged in four steps: 5 min at 250 W (heating power), 10 min at 400 W, 10 min at 600 W and 5 min at 250 W. After cooling at room temperature, the solutions were transferred into Teflon open recipients and then heated near to dryness on a hotplate, in order to remove HF and the excess of the other acids. This heating process was repeated three times and at the beginning of each time a small amount of acid was added (0.5 ml of HClO₄ the first time, then 1 ml of HNO₃). The final nearly dry residue was transferred to a 50 mL volumetric flask adding 0.5 ml of HNO3 and ultra-pure (U.P.) deionized water (∼18.4 MΩ cm) up to the final volume. Twenty blanks were also prepared, using the same analytical scheme but omitting the sample, in order to establish the practical limits of detection of the method (whole analytical procedure). The dissolution procedure was checked, with reference to REEs recovery, analysing the GSP-2 (CRM purchased by USGS). The determination of the REEs in the solutions coming from the dissolution of the samples was performed by means of an ICP-MS (Perkin-Elmer-ELAN-6100) configured with the following operational parameters: RF power 1,100 W, nebiliser flow 0.92 L min-1, resolution 0.7 amu, dwell time 40 ms, replicates 3, measurement mode peak hopping. REEs are easily determined by ICP-MS since they occur at the middle to the high portion of the mass range and have few interferences from polyatomic species. Rhodium (Rh) was added as internal standard to all the solutions, in order to correct any instrumental drifts. Blanks, calibration standards and sample solutions were prepared and, if necessary, diluted in a 1% HNO₃ matrix (obtained using U.P. deionized water and high purity grade HNO₃). Standard solutions, containing the REEs in the range 0.01–50 μg/L, were prepared daily by dilution of a stock multi-element solution (CAL1-1 supplied by AccuStandard) at 10 mg/L. The solutions obtained at the end of the dissolution procedure were generally diluted with a factor of 1 to 10 before the final determination by ICP-MS.

Limits of detection (LOD) for REEs have been calculated as Xmb+3Smb, where Xmb and Smb are respectively the mean and the standard deviation of the measurements of method blanks and are 27 μg kg-1 for La, 75 μg kg-1 for Ce and in the range 0.5–5 μg kg-1 for the other elements.

The geochemical reference material GSP-2 was used to verify the accuracy and precision of the analytical method. Five distinct tests were carried out on this CRM, therefore five separate solutions were analyzed with the ICP-MS. Table 1 shows the excellent agreement between the results obtained in our laboratory and the certified values.