Individual single-walled carbon nanotubes with covalent sidewall defects have emerged as a class of photon sources whose photoluminescence spectra can be tailored by the carbon nanotube chirality and the attached functional group/molecule. Here we present electroluminescence spectroscopy data from single-tube devices based on (7, 5) carbon nanotubes, functionalized with dichlorobenzene molecules, and wired to graphene electrodes. We observe electrically generated, defect-induced emissions that are controllable by electrostatic gating and strongly red-shifted compared to emissions from pristine nanotubes. The defect-induced emissions are assigned to excitonic and trionic recombination processes by correlating electroluminescence excitation maps with electrical transport and photoluminescence data. At cryogenic conditions, additional gate-dependent emission lines appear, which are assigned to phonon-assisted hot-exciton electroluminescence from quasi-levels. Similar results were obtained with functionalized (6, 5) nanotubes. We also compare functionalized (7, 5) electroluminescence data with photoluminescence of pristine and functionalized (7, 5) nanotubes redox-doped using gold­(III) chloride solution. This work shows that electroluminescence excitation is selective toward neutral defect-state configurations with the lowest transition energy, which in combination with gate-control over neutral versus charged defect-state emission leads to high spectral purity.

Graphene, a zero-gap semiconductor, absorbs 2.3% of incident photons in a wide wavelength range as a free-standing monolayer, whereas 50% is expected for ∼90 layers. Adjusting the layer number allows the tailoring of the photoresponse; however, controlling the thickness of multilayer graphene remains challenging on the wafer scale. Nanocrystalline graphene or graphite (NCG) can instead be grown with controlled thickness. We have fabricated photodetectors from NCG that are spectrally flat in the near-infrared to short-wavelength infrared region by tailoring the layer thicknesses. Transfer matrix simulations were used to determine the NCG thickness for maximum light absorption in the NCG layer on a silicon substrate. The extrinsic and intrinsic photoresponse was determined from 1100 to 2100 nm using chromatic aberration-corrected photocurrent spectroscopy. Diffraction-limited hyperspectral photocurrent imaging shows that the biased photoresponse is unipolar and homogeneous across the device area, whereas the short-circuit photoresponse gives rise to positive and negative photocurrents at the electrodes. The intrinsic photoresponses are wavelength-independent, indicative of bolometric and electrothermal photodetection.