Calibration of the Oxygen and Clumped Isotope Thermometers for (Proto-)Dolomite Based on Synthetic and Natural Carbonates

Dolomite is a very common carbonate mineral in ancient sediments, but is rarely found in modern environments. Because of the difficulties in precipitating dolomite in the laboratory at low temperatures, the controls on its formation are still debated after more than two centuries of research. Two important parameters to constrain the environment of dolomitization are the temperature of formation and the oxygen isotope composition of the fluid from which it precipitated. Carbonate clumped isotopes (expressed with the parameter Delta(47)) are increasingly becoming the method of choice to obtain this information. However, whereas many clumped isotope studies treated dolomites the same way as calcite, some recent studies observed a different phosphoric acid fractionation for Delta(47) during acid digestion of dolomite compared to calcite. This causes additional uncertainties in the Delta(47) temperature estimates for dolomites analyzed in different laboratories using different acid digestion temperatures. To tackle this problem we present here a (proto-)dolomite-specific Delta(47)-temperature calibration from 25 to 1100 degrees C for an acid reaction temperature of 70 degrees C and anchored to widely available calcite standards. For the temperature range 25 to 220 degrees C we obtain a linear Delta(47)-T relationship based on 289 individual measurements with R-2 of 0.864: Delta(47) (CDES 70 degrees C) = 0.0428 +/- 0.0020 x (10(6)/T-2) + 0.1481 +/- 0.0160 (T in Kelvin) When including two isotopically scrambled dolomites at 1100 degrees C, the best fit is obtained with a third order polynomial temperature relationship (R-2 = 0.924): Delta(47) (CDES 70 degrees C) (parts per thousand) = -0.0002 x (10(6)/T-2)(3) +0.0041 x (10(6)/T-2)(2) + 0.0115 x (10(6)/T-2) + 0.2218. Applying a calcite Delta(47)-T relationship produced under identical laboratory conditions results in 3 to 16 degrees C colder calculated formation temperatures for dolomites (with formation temperature from 0 to 100 degrees C) than using the (proto-)dolomite specific calibration presented here. For the synthetic samples formed between 70 and 220 degrees C we also determined the temperature dependence of the oxygen isotope fractionation relative to the water. Based on the similarity between our results and two other recent studies (Vasconcelos et al., 2005 and Horita, 2014) we propose that a combination of the three datasets represents the most robust calibration for (proto-)dolomite formed in a wide temperature range from 25 to 350 degrees C. 10(3)alpha(CaMg-carbonates-Water) = 2.9923 +/- 0.0557 (10(6)/T-2) - 2.3592 +/- 0.4116 Because of the uncertainties in the phosphoric acid oxygen and clumped isotope fractionation for (proto-)dolomite, we promote the use of three samples that are available in large amounts as possible inter-laboratory reference material for oxygen and clumped isotope measurements. A sample of the middle Triassic San Salvatore dolomite from southern Switzerland, the NIST SRM 88b dolomite standard already reported in other Delta(47) studies and a lacustrine Pliocene dolomite from La Roda (Spain). This study demonstrates the necessity to apply (proto-)dolomite specific Delta(47)-T relationships for accurate temperature estimates of dolomite formation, ideally done at identical acid digestion temperatures to avoid additional uncertainties introduced by acid digestion temperature corrections. In addition, the simultaneous analyses of dolomite reference material will enable a much better comparison of published dolomite clumped and oxygen isotope data amongst different laboratories.