The brain is a dynamic structure with the extracellular space (ECS) taking up almost a quarter of its volume. Signalling molecules, neurotransmitters and nutrients transit via the ECS, which constitutes a key microenvironment for cellular communication and the clearance of toxic metabolites. The spatial organization of the ECS varies during sleep, development and aging and is probably altered in neuropsychiatric and degenerative diseases, as inferred from electron microscopy and macroscopic biophysical investigations. Here we show an approach to directly observe the local ECS structures and rheology in brain tissue using super-resolution imaging. We inject single-walled carbon nanotubes into rat cerebroventricles and follow the near-infrared emission of individual nanotubes as they diffuse inside the ECS for tens of minutes in acute slices. Because of the interplay between the nanotube geometry and the ECS local environment, we can extract information about the dimensions and local viscosity of the ECS. We find a striking diversity of ECS dimensions down to 40 nm, and as well as of local viscosity values. Moreover, by chemically altering the extracellular matrix of the brains of live animals before nanotube injection, we reveal that the rheological properties of the ECS are affected, but these alterations are local and inhomogeneous at the nanoscale.

The intrinsic near-infrared photoluminescence observed in long single-walled carbon nanotubes is known to be quenched in ultrashort nanotubes due to their tiny size as compared to the exciton diffusion length in these materials (>100 nm). Here, we show that intense photoluminescence can be created in ultrashort nanotubes (∼40 nm length) upon incorporation of exciton-trapping sp defect sites. Using super-resolution photoluminescence imaging at <25 nm resolution, we directly show the preferential localization of excitons at the nanotube ends, which separate by less than 40 nm and behave as independent emitters. This unexpected observation opens the possibility to synthesize fluorescent ultrashort nanotubes-a goal that has been long thought impossible-for bioimaging applications, where bright near-infrared photoluminescence and small size are highly desirable, and for quantum information science, where high quality and well-controlled near-infrared single photon emitters are needed.