In the context of increasing atmospheric carbon dioxide (CO2) caused by anthropogenic activity, knowledge of the impacts of CO2 on the climate system, ecosystems, and sea-level is of primary importance. Palaeoclimate archives offer the unique possibility to study periods where climate was different and to investigate causal links between natural processes, atmospheric CO2 and the climate response. One major driver of natural variations in atmospheric CO2 is the biological pump in the oceans which strongly influences atmospheric CO2 by sequestering carbon out of the atmosphere into the deep ocean. Hence for a comprehensive understanding of natural climate, reconstructing both CO2 and palaeoproductivity is a powerful combination. Boron isotopes in planktonic foraminifera are a promising proxy to reconstruct pH and therefore atmospheric CO2 beyond the ice core record (0-800 thousand years before present). Yet it has only been tested in few locations and at relatively low resolution, obscuring our ability to robustly quantify its accuracy, and spatial variability. Furthermore, the standard approach requires complex and time-consuming analytical procedures, limiting the generation of a large number of data sets due to the specialist training required. Here I present three aims: (1) Testing the boron-isotope-CO2 proxy against the ice-core-CO2 record at several locations and at high temporal resolution in order to build the confidence in the application of the proxy deeper geological time and to investigate the relationship between ice core CO2 forcing and δ11B-derived relative pH change. I find that the broad pattern of δ11B-CO2 reconstructions closely follow the ancient atmosphere record in the Pacific, and sub-Atlantic Oceans, although there are details within each record which both expand and caution the utility of the proxy. (2) Present the method development of automating the boron purification using the prepFAST-MC from carbonate matrix in order to remove the analytical bottleneck and increase the throughput of samples. I show that this method is capable of accurate and complete matrix separation of B from carbonate matrices in a variety of materials. (3) Explore the potential of barium isotopes in planktonic foraminifera over the last deglaciation to track biological productivity in the palaeorecord as a new proxy of productivity. I show that interspecies differences are important hindering the use of δ138Ba in bulk foraminifera. I also show that non-spinose foraminifera may record marine particle microenvironment making spinose foraminifera better suited to explore ocean barium cycling in the paleorecord.