Graphene and other 2d materials have shown great promise for their ability to sense chemicals. This ability is attributed to the chemical nature of graphene, but also the fact that 2d materials have a large surface-to-area ratio, thus sensitivities towards sensing the single molecule have been achieved, however there is little to no selectivity between analytes. At present, researchers are primarily looking at pollutants or other small molecules using pristine graphene. Alternative sensing motifs are to functionalize graphene; covalent attachment to the graphene lattice, which destroy some properties and lead to irreproducibility across devices. We use a supramolecular approach to non-covalently functionalize the graphene using metal-organic cages. The metal-organic cages can encapsulate a small range of guests, however their binding inside the cage has not been investigated in the presence of graphene. Therefore, we present a novel investigation into the use of pristine graphene as a transducer for the host-guest interactions, in hopes of improving the selectivity of graphene whilst retaining the single molecule sensitivity. We tackle this through Raman spectroscopy and other surface techniques, then move onto fundamental electronic characterisation of graphene with a pyrene-edged metal-organic cage and show that the presence of C60 within the cage influences the electronic characteristics of chemical vapour deposited graphene. After the model system has been established, we extend the prospect into a commercial and scalable setting using liquid phase exfoliated graphene with a sulfonate-edged metal-organic cage with the intent to create an inkjet-printable selective sensor. We encountered expected successes and unexpected failures. Our aim is to establish research towards the field of using 2d materials decorated with 3d architectures that interact non-covalently to sense analytes that are specific to the 3d architecture.