Rationale: Aerodyne Tunable Infrared Laser Differential Absorption Spectrometer (TILDAS) allows for the measurement of carbonate-clumped (∆ 638) and stable oxygen isotope ratios (δ 628) without the mass interference of 17O, and understanding acid fractionation factors during phosphoric acid digestion is essential for inter-laboratory data comparison. Here, we present δ 628 and ∆ 638 ratios in CO 2 generated during phosphoric acid digestion of a reference carbonate at different temperatures. Methods: Carrara Marble (MAR-J1) calcite is digested with ⁓104% phosphoric acid using a) Break Seal method is a modified version of a McCrea-type reaction vessel where complete equilibration is achieved between product CO 2 and H 2O generated from the digestion and b) Individual Acid Bath (IAB) where the samples are digested with fresh aliquot of acid each time and the product CO 2 and H 2O simultaneously frozen in a U-trap connected to the reaction chamber by liquid N 2. Results: The regression equations for acid fractionation of δ 628 are: a) Break Seal Method: 1000 ln α (dig. temp- 25℃) = (0.532 ± 0.030) × 10 6/T 2 + (-6.185 ± 0.308) ; R 2 = 0.98. b) IAB: 1000 ln α (dig. temp- 25℃) = (0.359 ± 0.027) × 10 6/T 2 + (-4.114 ± 0.247) ; R 2 = 0.97. The regression equations for acid fractionation of ∆ 638 are: a) Break Seal Method: 1000 ln α (dig. temp - 25℃) = (0.0189 ± 0.0017) × 10 6/T 2 + (-0.2151 ± 0.0174) ; R 2 = 0.95. b) IAB: 1000 ln α (dig. temp - 25℃) = (0.0529 ± 0.0041) × 10 6/T 2 + (-0.6072 ± 0.0356) ; R 2=0.97. Conclusion: The shallow slope of the IAB method for δ 628 and the steeper slope for ∆ 638 indicate minimal resetting of the isotope composition of product CO 2 due to re-equilibration or interaction with free H 2O molecules. This observation enables inter-laboratory comparisons of δ 628 and ∆ 638 in calcite.

The potential for carbonate clumped isotope thermometry to independently constrain both the formation temperature (T∆47 ) of carbonate minerals and fluid oxygen isotope composition allows insight into long-standing questions in the Earth sciences, but remaining discrepancies between calibration schemes hamper interpretation of T∆47 measurements. To address discrepancies between calibrations, we designed and analyzed a sample suite (41 total samples) with broad applicability across the geosciences, with an exceptionally wide range of formation temperatures, precipitation methods, and mineralogies. We see no statistically significant offset between sample types, although comparison of calcite and dolomite remains inconclusive. When data are reduced identically, the regression defined by this study is nearly identical to that defined by four previous calibration studies that used carbonate-based standardization; we combine these data to present a composite carbonate-standardized regression equation. Agreement across a wide range of temperature and sample types demonstrates a unified, broadly applicable clumped isotope thermometer calibration.