Search

The design of efficient graphene-silicon (GSi) Schottky junction photodetectors requires detailed understanding of the spatial origin of the photoresponse. Scanning-photocurrent-microscopy (SPM) studies have been carried out in the visible wavelengths regions only, in which the response due to silicon is dominant. Here we present comparative SPM studies in the visible ($\lambda$ = 633nm) and infrared ($\lambda$ = 1550nm) wavelength regions for a number of GSi Schottky junction photodetector architectures, revealing the photoresponse mechanisms for silicon and graphene dominated responses, respectively, and demonstrating the influence of electrostatics on the device performance. Local electric field enhancement at the graphene edges leads to a more than ten-fold increased photoresponse compared to the bulk of the graphene-silicon junction. Intentional design and patterning of such graphene edges is demonstrated as an efficient strategy to increase the overall photoresponse of the devices. Complementary simulations and modeling illuminate observed effects and highlight the importance of considering graphene's shape and pattern and device geometry in the device design.

Graphene-silicon (GS) Schottky junctions have been demonstrated as an efficient architecture for photodetection. However, the response speed of such devices for free space light detection has so far been limited to 10's-100's of kHz for wavelength $\lambda >$ 500nm. Here, we demonstrate graphene-silicon Schottky junction photodetectors fabricated on a silicon-on-insulator substrate (SOI) with response speeds approaching 1GHz, attributed to the reduction of the photo-active silicon layer thickness to 10$\mu$m and with it a suppression of speed-limiting diffusion currents. Graphene-silicon-on-insulator photodetectors (GSOI-PDs) exhibit a negligible influence of wavelength on response speed and only a modest compromise in responsivities compared to GS junctions fabricated on bulk silicon. Noise-equivalent-power (NEP) and specific detectivity (D$^*$) of GSOI photodetectors are 14.5pW and 7.83$\times10^{\rm{10}}$ cm Hz$^{\rm{1/2}}$W$^{\rm{-1}}$, respectively, in ambient conditions. We further demonstrate that combining GSOI-PDs with micro-optical elements formed by modifying the surface topography enables engineering of the spectral and angular response.

Graphene-Silicon Schottky diode photodetectors possess beneficial properties such as high responsivities and detectivities, broad spectral wavelength operation and high operating speeds. Various routes and architectures have been employed in the past to fabricate devices. Devices are commonly based on the removal of the silicon-oxide layer on the surface of silicon by wet-etching before deposition of graphene on top of silicon to form the graphene-silicon Schottky junction. In this work, we systematically investigate the influence of the interfacial oxide layer, the fabrication technique employed and the silicon substrate on the light detection capabilities of graphene-silicon Schottky diode photodetectors. The properties of devices are investigated over a broad wavelength range from near-UV to short-/mid-infrared radiation, radiation intensities covering over five orders of magnitude as well as the suitability of devices for high speed operation. Results show that the interfacial layer, depending on the required application, is in fact beneficial to enhance the photodetection properties of such devices. Further, we demonstrate the influence of the silicon substrate on the spectral response and operating speed. Fabricated devices operate over a broad spectral wavelength range from the near-UV to the short-/mid-infrared (thermal) wavelength regime, exhibit high photovoltage responses approaching 10$^6$ V/W and short rise- and fall-times of tens of nanoseconds.

The combination of plasmonic nanoparticles and graphene enhances the responsivity and spectral selectivity of graphene-based photodetectors. However, the small area of the metal-graphene junction, where the induced electron-hole pairs separate, limits the photoactive region to sub-micron length scales. Here, we couple graphene with a plasmonic grating and exploit the resulting surface plasmon polaritons to deliver the collected photons to the junction region of a metal-graphene-metal photodetector. This results into a 400% enhancement of responsivity and a 1000% increase in photoactive length, combined with tunable spectral selectivity. The interference between surface plasmon polaritons and the incident wave introduces new functionalities, such as light flux attraction or repulsion from the contact edges, enabling the tailored design of the photodetector's spectral response. This architecture can also be used for surface plasmon bio-sensing with direct-electric- readout, eliminating the need of complicated optics.

Graphene's high mobility and Fermi velocity, combined with its constant light absorption in the visible to far-infrared range, make it an ideal material to fabricate high-speed and ultra-broadband photodetectors. However, the precise mechanism of photodetection is still debated. Here, we report wavelength and polarization dependent measurements of metal-graphene-metal photodetectors. This allows us to quantify and control the relative contributions of both photo-thermo- and photoelectric effects, both contributing to the overall photoresponse. This paves the way for a more efficient photodetector design for ultra-fast operating speeds.

Amongst the wide spectrum of potential applications of graphene, ranging from transistors and chemical-sensors to nanoelectromechanical devices and composites, the field of photonics and optoelectronics is believed to be one of the most promising. Indeed, graphene's suitability for high-speed photodetection was demonstrated in an optical communication link operating at 10 Gbit/s\cite. However, the low responsivity of graphene-based photodetectors compared to traditional III-V based ones is a potential drawback. Here we show that, by combining graphene with plasmonic nanostructures, the efficiency of graphene-based photodectors can be increased by up to 20 times, due to field concentration in the area of a p-n junction. Additionally, wavelength and polarization selectivity can be achieved employing nanostructures of different geometries.

The recently found material graphene is a truly two-dimensional crystal and exhibits, in addition, an extreme mechanical strength. This in combination with the high electron mobility favours graphene for electromechanical investigations down to the quantum limit. Here, we show that a monolayer of graphene on SiO2 provides natural, ultra-small membranes of diameters down to 3 nm, which are caused by the intrinsic rippling of the material. Some of these nano-membranes can be switched hysteretically between two vertical positions using the electric field of the tip of a scanning tunnelling microscope (STM). They can also be forced to oscillatory motion by a low frequency ac-field. Using the mechanical constants determined previously, we estimate a high resonance frequency up to 0.4 THz. This might be favorable for quantum-electromechanics and is prospective for single atom mass spectrometers., Comment: 9 pages, 4 figures

In this work, double-gated field effect transistors manufactured from monolayer graphene are investigated. Conventional top-down CMOS-compatible processes are applied except for graphene deposition by manual exfoliation. Carrier mobilities in single- and double gated graphene field effect transistors are compared. Even in double-gated graphene FETs, the carrier mobility exceeds the universal mobility of silicon over nearly the entire measured range. At comparable dimensions, reported mobilities for ultra thin body silicon-on-insulator MOSFETs can not compete with graphene FET values., Comment: 7 pages, 9 figures

Using scanning tunneling microscopy (STM) in ultra high vacuum and atomic force microscopy, we investigate the corrugation of graphene flakes deposited by exfoliation on a Si/SiO2 (300 nm) surface. While the corrugation on SiO2 is long-range with a correlation length of about 25 nm, some of the graphene monolayers exhibit an additional corrugation with a preferential wave length of about 15 nm. A detailed analysis shows that the long range corrugation of the substrate is also visible on graphene, but with a reduced amplitude, leading to the conclusion that the graphene is partly freely suspended between hills of the substrate. Thus, the intrinsic rippling observed previously on artificially suspended graphene can exist as well, if graphene is deposited on SiO2., Comment: 10 pages, 11 figures, including supplementary material

The absence of a band gap in graphene restricts its straight forward application as a channel material in field effect transistors. In this letter, we report on a new approach to engineer a band gap in graphene field effect devices (FED) by controlled structural modification of the graphene channel itself. The conductance in the FEDs is switched between a conductive "on-state" to an insulating "off-state" with more than six orders of magnitude difference in conductance. Above a critical value of an electric field applied to the FED gate under certain environmental conditions, a chemical modification takes place to form insulating graphene derivatives. The effect can be reversed by electrical fields of opposite polarity or short current pulses to recover the initial state. These reversible switches could potentially be applied to non-volatile memories and novel neuromorphic processing concepts., Comment: 14 pages, 4 figures, submitted to IEEE EDL

Conventional field effect transistor operation in graphene is limited by its zero gap and minimum quantum conductivity. In this work, we report on controlled electrochemical modification of graphene such that its conductance changes by more than six orders of magnitude, which enables reversible bipolar switching devices. The effect is explained by a chemical reaction of graphene with hydrogen (H+) and hydroxyl (OH-), which are catalytically generated from water molecules in the sub-stochiometric silicon oxide gate dielectric. The reactive species attach to graphene making it nonconductive but the process can subsequently be reversed by short current pulses that cause rapid local annealing. We believe that the demonstrated electrochemical field effect devices are viable candidates for future logic circuits, non-volatile memories and novel neuromorphic processing concepts.

In this letter, a top-gated field effect device (FED) manufactured from monolayer graphene is investigated. Except for graphene deposition, a conventional top-down CMOS-compatible process flow is applied. Carrier mobilities in graphene pseudo-MOS structures are compared to those obtained from top-gated Graphene-FEDs. The extracted values exceed the universal mobility of silicon and silicon-on-insulator MOSFETs., Comment: 12 pages, 3 figures