observed, which seemingly resulted in a growth area for biofilms within the precipitate. However, these findings are merely judged on the basis of measuring the optical density of the dispersion (which should be strongly dependent on the GO concentration [47]) and microscopy. In the context of this study, it should also be noted that testing for bacterial growth inhibition zones on agar plates should only produce negatives (no inhibition) for graphene, since the antibacterial effect is not based on the leaching of soluble species like toxic ions.
In accordance with the previously described study by Akhavan and Ghaderi [44], a slight enhancement of bacterial growth on randomly deposited graphene oxide was reported. The enhanced growth of bacteria on rough (but non-perpendicular) GO films is utilized by Wang et al. to enhance the activity of anaerobic ammonium oxidation bacteria [48].
The apparent discrepancy between biocompatibility and differing antibacterial activities has been explained by Hui et al. [47] In their work, the authors showed that the adsorption of proteins and other biomolecules that are commonly found in nutrient broth on basal planes (i.e., a flat surface) of graphene deactivates its antibacterial activity. However, most antibacterial tests are performed in simple saline solutions that do not contain ingredients that are prone to adsorb on graphene. Consequently, results of the antibacterial activity of graphene (and other nanomaterials [49]) have to be interpreted regarding whether nutrient broths or similar media containing biomolecules that are strongly surface active and readily adsorb on (carbon) nanomaterials were used in the biological experiments. Written concurrently to this overview, these findings on the antibacterial activity have also been very recently summarized in several extensive and detailed reviews [50–52].
2.2. Carbon Nanotubes
Carbon nanotubes (CNT) can be described as hollow structures with an extremely high aspect ratio, which are formed by rolled graphene sheets. Depending on the rolling angle, carbon nanotubes can be metallic or semi-conductive. Furthermore, nanotubes are categorized as single-walled nanotubes (SWNTs) and multi-walled nanotubes (MWNTs). The latter consist of several single-walled tubes that are nested inside each other. Carbon nanotubes, especially SWNTs have a significantly higher antibacterial effect than most carbon nanomaterials [12]. The antibacterial activity of purified SWNTs was first demonstrated by Kang et al. [53]. In their detailed follow-up study, Kang et al. show that purified SWNTs and MWNTs seriously impact bacterial membrane integrity upon direct contact. Accordingly, metabolic activity and morphology are compromised as well [20]. They also show that SWNTs exhibit a stronger antibacterial activity than MWNTs, probably caused by their smaller size that facilitates membrane perturbation and provides a larger surface area (Figure 3). Oxidative stress likely plays an additional, albeit minor, role in the antibacterial mechanism [20]. Liu et al. add further detail to the investigation of mechanical effects that govern the antibacterial activity of CNTs [54]. In accordance with the work by Chen et al. [55], they add weight to the hypothesis that SWNT can act as “nano-darts” that pierce bacterial membranes by comparing bacteria with differing membrane robustness and ruling out other mechanisms by performing extensive control experiments. Further studies indicated that MWNTs showed no mutagenic activity in bacteria assays with E. coli and S. typhimurium [56].
A distinct feature of sp2 carbon nanomaterials, including nanotubes, is their special electronic
structure causing semi-conductivity, or, in the case of some CNTs, even (pseudo) metallic conductivity. This aspect was investigated by Vecitis et al. [57] who could clearly demonstrate that metallic nanotubes exhibit a much higher antibacterial activity than semi-conducting CNTs. Consequently, electronic effects can also contribute to the antibacterial activity of nanotubes and this might also apply for other carbon nanomaterials. Similar to the effect described in more detail for fullerenes (see below), CNTs can be activated via photosensitization causing the formation of additional ROS [58].