Atmospheric CO2 level (pCO2), < 280 ppm in pre-industrial times, is predicted to exceed 1000
ppm by the year 2100 under high growth and low mitigation scenarios. Per doubling of pCO2 the
global surface temperature is expected to increase by 1.5° to 4.5°C (IPCC, 2013). Global war-
ming by 2°C (relative to pre-industrial times) is already predicted to cause dramatic environmental
changes, amongst which are a significant loss in terrestrial biodiversity and a major perturbation
of the marine ecosystem (IPCC, 2018). An important method by which climate models are as-
sessed is by comparing simulations of past climate states to past reconstructions of temporal
variations in temperature and pCO2. The evolution of Earth’s surface temperature has been re-
constructed from common elemental (e.g., Mg/Ca ratio of marine carbonates; Nürnberg et al.,
1996), molecular (e.g., TEX86 of sedimentary organic matter; Schouten et al., 2002) and bulk
isotopic proxy data (e.g., oxygen isotopic composition of marine carbonates; Epstein et al., 1953)
of well-dated sedimentary archives. However, in addition to temperature, the chemical and iso-
topic composition of sedimentary archives can be controlled by reaction kinetics involved in their
formation and by the chemical and/or isotopic composition of the waters these archives originated
from. Moreover, after their formation and deposition, they can be exposed to secondary, diage-
netic processes which may alter their pristine chemical and isotopic composition. These additional
processes can introduce large scatter and biases in reconstructed Earth’s surface temperatures
of up to tens of degrees C (e.g., Veizer & Prokoph, 2015). They may provide an explanation why
state-of-the-art climate models cannot reproduce the relatively shallow paleo-latitudinal tempera-
ture gradients obtained for the high-pCO2 intervals of the Eocene (Evans et al., 2018) and early
Cretaceous (Price et al., submitted) and why sometimes even contrasting results have been ob-
tained concerning the past coupling between atmospheric pCO2 and Earth’s surface temperature
(Veizer et al., 2000; Came et al., 2007), especially for the Archean (Catling & Zahnle, 2020).

Current disequilibrium ∆47 and ∆48 trajectories rely on modeled reaction rates and kinetic
isotope fractionations (Guo & Zhou, 2019; Guo, 2020). These need to be verified for the different
reactions involved in carbonate (bio)mineralization to be able to eliminate the kinetic bias in car-
bonate formation temperature correctly. Kinetic isotope fractionations and exchange rates occur-
ring in the solution (associated with diffusion, (de)hydration and (de)hydroxylation reactions) and
at the solution/crystal interface will be determined experimentally under controlled conditions. In
cooperation with M. Dietzel and M. Böttcher, calcite and aragonite will be precipitated in sets of
systematic experimental approaches at variable temperatures (characteristic of Earth’s surface),
variable CO2 diffusion times, variable pH and insufficient, but variable time for O-isotope ex-
change between DIC, H2O and OH– (e.g., Dietzel et al., 2009). Starting from solutions equilibrated
at distinct pH (Beck et al., 2005), the precipitation rate and carbonate saturation state (Devriendt
et al., 2017) will be varied. Experimentally determined kinetic fractionations will then be used to
improve model simulations (collaborator: Weifu Guo).