Supporting information for the manuscript
“Preservation of NOM-metal complexes in a modern hyperalkaline stalagmite: implications for speleothem trace element geochemistry”
by A. Hartland et al.
S1. Study site background
This study targeted the hyperalkaline dripwaters of Poole’s Cavern, nr Buxton (53°12’N 1°56’W), UK.
Figure S1 Location of Poole’s Cavern, Buxton, UK.
Poole’s Cavern is a shallow telogenetic cave developed in Lower Carboniferous (Asbian) Limestone, Bee Low Limestone Formation, Peak Limestone Group. Poole’s Cavern is approximately 240 m long and in the largest (main) chamber it is around 15 m wide and 20 m high.
Soils above the cave are 30-60 cm thick consisting of leaf litter and organic-rich top soil (Brown Ranker (Baker & Genty, 1999)) overlying lime waste and fragments of poorly-sorted limestone (see below). The soil profile at Poole’s Cavern is typically characterized by deep, organic-rich A horizons but exhibit a high degree of variability across the site. Lime waste has a pervasive influence over the geochemistry of Poole’s Cavern dripwaters, causing hyperalkalinity and distinct elemental compositions (notably elevated K+ and SO42-). For more information on Poole’s Caverns hyperalkaline environments see Hartland et al., (2010a). Readers are also directed to the study of Baker et al. (1999) which describes the structure of annual fluorescent banding in Poole’s Cavern’s hyperalkaline stalagmites.
Figure S2 Schematic of Poole’s Cavern with drip point locations (drip point PE1 is highlighted with a red circle) (Source: Hartland et al., 2010a).
S2. Stalagmite chronology
Figure S3 Variations in Sr and V concentration in stalagmite PC-08-1 (top) determined by LA-ICPMS. Higher and lower concentrations of V and Sr, are shown to respectively coincide with dark laminae (marked with white rectangles and black dashed lines) in the sample. Because lamina colour was not a reliable measure of the boundary between dark and pale calcite Sr and V concentrations were used to delineate summer (pale) and winter (dark) layers.
S3. Modelling of CaCO3 precipitation and pH change in the PC-08-1 thin film
Atmospheric CO2 seeding of calcite growth under hyperalkaline (pH 12) conditions was studied by Clark et al. (1992). Above pH 9.5, calcite deposition occurs via hydroxylation of CO2(g). The rate at which CaCO3 precipitation proceeds is determined by the rate of CO2(g) diffusion to the water surface and the rate of hydroxylation. The rate of dehydroxylation is function of the temperature and the concentration of CO2(aq) and OH-(aq)
-d[CO2(aq)]/dt = kOH- x [CO2(aq)][OH-(aq)] (S1)
The calculated flux of CO2(g) to the PC-08-1 surface was calculated using the linear velocity of CO2(g) (ideal gas behaviour) and its molar concentration (Cunningham and Williams, 1980) within a representative range for Poole’s Cavern from values of 500-1500 ppm (Hartland et al., 2010a).
JCO2 = ¼nvz (mol cm-2 s-1) (S2)
where n is the molar concentration of CO2 in cave air (mol cm-3), vz is the molecular velocity of CO2 (cm s-1)
Vz = kT2πm (S3)
where k is the Boltzmann constant (gas constant per molecule) 1.380658 x 10-23 J K-1, and m is the molecular mass for 12C16O2 of 7.3665 x 10-26 kg.
Vz = 9.16 x 101 cm s-1
At 500 ppm CO2 JCO2 = 5.09 x 10-7 mol cm-2 s-1
Although the flux of CO2 in the experimental system employed by Clark et al. was higher (higher experimental temperature and pCO2) the relative supply of CO2 into solution was limited by a low SA/V imposed by the experimental vessel. Contrastingly, the thin film at the stalagmite surface covers a relatively large area and contains a comparatively small volume. The surface area and thickness of the PC-08-1 thin film were estimated by assuming a cylindrical geometry (r=1.5 cm; SA=7.07 cm2) and a film thickness (natural range is 100-400 mm) of 200 mm. This gives a predicted drip volume of 0.141 ml which is consistent with a SA/V of 50.
Multiplying JCO2 by the SA/V gives an estimated flux of CO2 into solution of
At 500 ppm CO2 JCO2 = 2.55 x 10-5 mol cm-2 s-1
The reaction constant for the reaction between CO2(g) and OH-(aq) has been evaluated by Pinsent et al. (1956). Using the reaction constant (kOH- = 2550 ± 87 mol s-1) for low salinity solutions and a temperature of 10 °C (cave air temp ca. 8 °C) we calculate the maximum rate of CO2 hydroxylation using equation S4 at 500, 1000 and 1500 ppm CO2(g) as
-d[CO2(aq)]/dt = [kOH-][CO2(aq)][OH-(aq)] (S4)
At 500 ppm CO2(g) -d[CO2(aq)]/dt = [2500][2.55 x 10-5][5.01 x 10-3] = 3.25 x 10-4 mol s-1
Because the flux of CO2(g) to solution is so high in a thin film, and the OH- concentration is comparatively low, the hydroxylation reaction can proceed very quickly. Thus, hydroxylation is not rate limiting and CaCO3 deposition at PC-08-1 is predicted to be driven by the concentration of CO2 in cave air.
Based on the rate of growth of PC-08-1 the efficiency of Ca2+ removal from solution through CaCO3 accrual was calculated to be approximately 20%. Based on our calculations, for ca. 20% removal of Ca2+ for water with an average residence time of 20-25s (average drip rate of PE1 (Baker et al., 1999)) the average pCO2 of cave air is predicted to fall within a range of 400-450 ppm, suggesting that pCO2 values were close to atmospheric levels in the Poached Egg chamber through much the depositional history of PC-08-1.
Cave air pCO2 in the Poached Egg chamber was slightly elevated above atmospheric values in monthly spot measurements (average = 558 ± 182 ppm) for the eight month period from October 2008 to May 2009, but were significantly elevated in the summer months (average = 2303 ± 338) from June – August 2009. Higher extension rates in pale summer calcite in PC-08-1 are consistent with enhanced CO2 seeding in the summer compared with winter. Given the uncertainties involved (e.g. inter-annual variations in drip rate and water chemistry) this modelling exercise successfully demonstrates the influence of cave air pCO2 on hyperalkaline extension rates as evinced by the good agreement between modelled and observed pH change in Poole’s Caverns hyperalkaline dripwaters (see main text).
S4 Additional trace element time series
Figure S4 Co-variation in Co, V and Pb in stalagmite PC-08-1 demonstrates mutual association of these elements with NOM-M complexes. Pb was below solution ICPMS detection limits in PE1 cave waters (<1 ppb). Both Co and V were shown to be predominantly complexed and transported by nano-scale and nominally dissolved NOM in PE1 dripwaters in Hartland et al. (2011).
Figure S5 Co-variation in Co and Al in stalagmite PC-08-1. This may be the result of complexation of Al by NOM, or preferential incorporation of Al at non-lattice sites during high NOM loading. Further work is needed to fully understand the speciation of Al in cave waters. However, complexation of Al by NOM in surface systems is well known (Milne et al., 2003).
Figure S6 Co-variation in Co (bold line) and Br (not quantified) in stalagmite PC-08-1. This may be the result of electrophilic aromatic substitution by Br in NOM. Further work is needed to fully understand the speciation of Br in cave waters.