SUPPLEMENTARY INFORMATION

Transcription

1 SUPPLEMENTARY INFORMATION Nanowires of Methylammonium Lead Iodide (CH 3 NH 3 PbI 3 ) prepared by low temperature solution-mediated crystallization E. Horváth 1 *, M. Spina 1, Zs. Szekrényes 2, K. Kamarás, 2 R.Gaal, 1 D. Gachet 3, L. Forró 1 1 Laboratory of Physics of Complex Matter (LPMC), Ecole Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland 2 Institute for Solid State Physics and Optics, Wigner Research Centre for Physics, Hungarian Academy of Sciences, 1525 Budapest, Hungary 3 Attolight AG, Innovation Park, 1015 Lausanne, Switzerland 1

2 Synthesis of MAPbI3 nanocrystallites by slip-coating method MAPbI3 single crystals and polycrystalline powder were synthesized using the method described by Poglitch and Weber. 1 Single crystals were prepared by precipitation from a concentrated aqueous solution of hydriodic acid (57 w% in H2O, % Sigma-Aldrich) containing lead (II) acetate trihydrate ( %, Acros Organics) and a respective amount of CH3NH2 solution (40 w% in H2O, Sigma-Aldrich). A constant o C temperature gradient was applied to induce the saturation of the solute at the low temperature part of the solution. Besides the formation of hundreds of submillimeter sized crystallites (referred as polycrystalline powder) several large MAPbI3 crystals with 3x5 mm silver-grey mirror-like facets were grown after 21 days. Leaving the crystals in open air resulted in a silver-grey to green-yellow color change. In order to prevent this unwanted reaction with moisture the as synthesized crystals were immediately transferred and kept in a desiccator prior the measurements. Large single crystals were kept as a reference material for qualitative analysis. The as-prepared polycrystalline powder was dissolved in organic solvents. Filiform crystallites: 10 microliters of saturated solution ( 50 w%) of MAPbI3 in dimethylformamide (DMF, Sigma-Aldrich) was dropped onto a microscope glass slide (Thermoscientifictype, 76x26 mm) and covered with a second microscope slide so that the excess yellow solution squeezed out; the remaining solution formed a homogenous liquid film between the glass plates (Fig 1 a-c). The excess of the MAPbI3 solution was removed from the sides by soaking with a tissue. Next, the bottom substrate was held in place while gradually sliding the upper glass plate, exposing the thin liquid film to air. Note that the sliding was performed manually. In order to prepare thicker wires 20 microliters of the same saturated solution was applied. Solvent evaporation from the uncovered surface caused an instantaneous yellow to brown- 2

3 red color change. The process was performed at room temperature. Unless otherwise specified, the same protocol was applied on SiO2 covered microfabricated chips. Nanoparticle based films: were prepared by the same procedure described above applying saturated solution (40 w%) of MAPbI3 in gamma-butyrolactone (GBL, Sigma-Aldrich) solvent (Fig.S8 a-c). Figure S1: Optical microscope images and SEM micrographs of MAPbI 3 nanowires slip-coated on a film composed of TiO 2 nanoparticles (a, b) and on a flat, compact TiO 2 ( 100 nm) grown by ALD (atomic layer deposition) on a highly p-doped Si substrate (c, d). Note that the MAPbI 3 nanowire growth similarly takes place not only on flat, compact TiO 2 surfaces (c, d) but on substrates coated with porous titania as well (a, b). 3

4 Figure S2: A representative SEM micrograph showing a set of the MAPbI 3 nanowires used for the statistical morphological data analysis. The same area was imaged by atomic force microscopy. More details about the morphological analysis can be found below in the section entitled Device geometry calculations (a). Statistical distribution of nanowire length (L) over width (W) ratio was obtained by analyzing SEM and AFM data of 40 individual nanowires (b). Material characterization Energy Dispersive X-ray Spectroscopy (EDX) Scanning Electron Microscope images were taken with a MERLIN Zeiss electron microscope. The elemental composition of the fibrous crystallites was analyzed by EDX (accelerating voltage of 8 kv, working distance of 8.5 mm). Samples were mounted on Al pucks with carbon tape with electrical contact to the surface also formed by carbon tape. Energy-dispersive X-ray spectroscopy (EDS) measurement was performed with an X-MAX EDS detector mounted at a 35 degree takeoff angle with a SATW window. EDS spectra and map were obtained at a working distance of 8.5 mm with 8 kev accelerating voltage and a current held at 92 pa channels were used for the acquisitions, corresponding to energy of 5eV per channel. The EDS map was taken at a 4

5 magnification of 3764x corresponding at a map area of 30.6x22.9 µm 2 with a resolution of 2048x1536 pixels. Spectra were acquired over 1573 seconds of live time with detector dead time averaging of 4% and a dwell time per pixel of 500 µs. Quantitative EDS analysis utilized AZtec software provided by Oxford Instrument Ltd. The Pb:I ratio in the as-synthesized filiform nano crystallites and single crystals is found to be 3, which is consistent with formula of MAPbI3. Element Line Type Apparent Concentration k Ratio Wt% Wt% Sigma Atom% Si KA I LA Pb MA Total Table S1: EDS Analysis Results 5

6 Figure S3: SEM micrograph of an assembly of slip-coated MAPbI 3 nanowires (a). EDS window integral maps of Si (b), I (c), Pb (d). Note that the majority of Pb and I signals were detected from the area covered with MAPbI 3 nanowires; excluding the possibility an underlying 2D film exists below the 1D structures. The lowintensity signals detected from the uncovered area are due to the presence of a small number of aggregates of ~10 nm sized isotropic MAPbI 3 crystallites. These particles can clearly be seen on TEM and SEM micrographs (Fig 2 a, c, e, g) EDS sum spectrum calculated from the data acquired from all the pixels in the secondary electron image (e). 6

7 Powder X-ray Diffraction (XRD) X-ray diffraction patterns were collected with a RIGAKU diffractometer (40 kv and 30 ma) using a source of Cu Kλ ( Å). XRD diffractogram shows high intensity diffraction peaks at 2 Theta 14 º (110) and 28.4 º (220), which were identified as the characteristic peaks of the cubic MAPbI3 phase 2 (Fig S4 a,b). The presence of the low intensity peak at is assigned to a PbI2 phase which presumably formed as a result of the humidity-induced partial decomposition of MAPbI3 during the PXRD measurement (Fig S4b). The presence of two major reflection peaks suggests that highly oriented MAPbI3 crystallites (along the (110) direction) are present on the substrate. In FigS4 the broad and low intensity diffraction peak centered at 2 Theta 24 degrees comes from the microscope glass slide support 2. Figure S4: PXRD diffractogram of the as prepared MAPbI 3 nanowires. (a) Figure 3 a, c, e, (b) Figure 3 b, d, f. 7

8 Raman Spectroscopy The structure of the filiform perovskites was analyzed by Raman spectroscopy (HORIBA LabRAM HR Raman spectrometer). Spectra were taken using a 532nm green excitation laser. The laser power was reduced in order to avoid photodegradation of the sample. The focal spot size was about 10 µm using a 50x long working distance objective. Figure S5: Raman spectrum of MAPbI 3 nanowires slip-coated on SiO 2/Si substrate. Numbers indicate the frequency of the measured lines. The recorded Raman spectrum shows great similarities with the reported Raman modes of MAPbI3 by Quarti and coworkers. According to their measured Raman spectra assisted by DFT 8

9 calculations the bands at 69 and 92 cm 1 are assigned to the characteristic modes of the inorganic cage (bending and stretching modes of the Pb I bonds). They have assigned the modes at 119 and 154 cm 1 to the librations of the organic cations and the broad, unstructured cm 1 features to the torsional mode of the methylammonium cations 3. Fourier Transform Infrared Spectroscopy (FTIR) Measurements were performed with a Bruker Tensor FTIR spectrometer with a DTGS detector and 4 wavenumber resolution. The bulk crystal was measured in transmission mode on Si while the diffuse reflectance (DRIFT) mode was employed for the analysis of the nanowires. For the DRIFT we used an integration sphere and a flat gold surface as a reference. In both samples we found the vibrational peaks at identical frequencies. These correspond exactly to those of Onoda- Yamamuro, Matsuo and Suga 4 on bulk crystals of CH3NH3PbI3 prepared carefully to avoid formation of (MA)4PbI6 2H2O. It is frequently reported in the literature that the temperature must be kept above 40 ºC otherwise these types of crystals will form. Since the fabrication of the filiform crystallites was performed at room temperature the presence of (MA)4PbI6 2H2O phase was expected. The phase pure (MA)4PbI6 2H2O or mixed phase (MA)4PbI6 2H2O@(MA)PbI3 compound would cause water absorption bands in the spectrum. Taking the analogy with PbI2 5 such bands would appear at 3650 and 1600 cm -1. The absence of these vibration modes in the FT- IR spectrum of filiform crystallites and bulk single crystal suggests identical chemical composition of CH3NH3PbI3. 9

10 Figure S6: FTIR spectra obtained for MAPbI 3 nanowires and a bulk crystal of CH 3NH 3PbI 3. The Si-O vibration of the substrate is marked with an asterisk. 10

11 Cathodoluminescence (CL) As shown in figure S7(a), quantitative cathodoluminescence (CL) of CH3NH3PbI3 nanowires and micron-sized pieces of grinded CH3NH3PbI3 single crystals was carried out at room temperature using an Attolight Rosa 4634 CL microscope, which tightly integrates a high numerical aperture (NA 0.72) achromatic reflective lens within the objective lens of a field emission gun scanning electron microscope (FEG-SEM). The focal plane of the light lens matches the FEG-SEM optimum working distance. The CL collection efficiency is constant over a 300 μm field of view so that CL emission can be compared quantitatively between two separate points. CL was spectrally resolved with a Czerny-Turner spectrometer (320 mm focal length, 150 grooves/mm grating) and measured with an UV-Vis. CCD camera. The electron beam energy was set at 2 kev and the value of the probe current was kept moderate (around 2 na). Figure S7: Cathodoluminescence (CL) measurements performed on CH 3NH 3PbI 3 nanowires and micron sized pieces of grinded CH 3NH 3PbI 3 single crystals. SEM image (a) and the corresponding cathodoluminescence map of CH 3NH 3PbI 3 nanowires (b). The white cross depicts the position where the characteristic CL spectrum was acquired (c). SEM image (d) and the corresponding cathodoluminescence map of a micron sized particle prepared by grinding a CH 3NH 3PbI 3 single crystal (e). The white cross depicts the position where the characteristic CL spectrum was acquired (f). 11

12 Both, the micron sized particles prepared by grinding a CH3NH3PbI3 single crystal and the filiform CH3NH3PbI3 nanoparticles show similar CL spectra with a single peak centered at ~767 nm (~1.61eV) and ~770 nm (~1.61eV), respectively. These peak positions correspond well with the band gaps of reported MAPbI3 perovskites ( ev). Note that the luminescence intensity significantly increases at the edges of the crystallite (FigS7e). Repeated measurements on the same area resulted in a decreasing luminescence signal pointing out the material sensitivity to the applied electron beam (electron-beam damage). Slip-coating with GBL Figure S8: Optical (a) and corresponding SEM (b, c) images of MAPbI 3 crystals slip-coated from GBL solution. All images show the absence of filiform crystallites and the formation of agglomerates composed of isotrophic particles (c) 12

13 Device fabrication The devices were fabricated by slip-coating of MAPbI3 solution (GBL-nanoparticles, DMFfiliform crystallites) onto a highly p-doped silicon substrate with 300 nm thermally grown SiO2 on top. Source and drain contacts were patterned by e-beam evaporation (Leybold Optics LAB 600 H) of Pt (100 nm) in high vacuum (<10-6 mbar, at room temperature) through a microfabricated hard mask. The characteristic width of the fabricated contact pads was 100 µm, their length was varying between 5 µm and 50 µm. The sketch and an optical microscopy image of a representative device are shown in Fig 4a and 4b. Device characterization Electrical measurements If it is not otherwise stated, all measurements were performed on freshly prepared samples under ambient conditions. Two-point electrical measurements (d.c.) were carried out using a National Instruments GPIB-USB-HS controller and a Keithley 2400 source meter. In order to minimize sources of external noise, the measurements were performed in a home-built Faraday cage. A microscope objective and a micromechanical stage were used to locate the device. Photocurrent measurements We probed the devices and their time-dependent responsivity to laser excitation using a laser beam (COHERENT laser module, model , λ=633nm) with an illumination power from 0.1 mw to 10 mw. The spot size of 4 mm 2 resulted in an estimated maximum illumination power of 0.25 Wcm -2. The time response of the photocurrent was acquired by modulating the laser beam with a mechanical chopper (217 Hz) and detecting the photocurrent with a current preamplifier. 13

14 Device geometry calculations The geometrical factors used for the performance calculations of the photodetectors were extracted from the AFM and SEM analysis of the fabricated devices (Fig. S9 a.-f). The height of the MAPbI3 nanowires and nanoparticles was measured with a Bruker Dimension FastScan atomic force microscope (in tapping mode, FigS9 a, b). The width and length of the nanostructured MAPbI3 was determined from SEM (Scanning Electron Microscope) images (MERLIN Zeiss SEM operating at 1kV and a working distance of ~4mm, Fig S9 c-f). The active area of the devices made of filiform perovskites contained several crystallites (Fig S9 f). In this case the calculations were done by summing up the dimensions of individual nanowires lying in the active area (Fig S9 b, d). For the nanoparticles, a mean value of their height was used and a uniform coverage of the contacts was assumed (Fig S9 a, e). 14

15 Figure S9: AFM (a, b) and SEM (c, d, e, f) of several MAPbI 3 nanoparticles (a, c, e) and nanowires (b, d, f) used in the fabrication of photodetectors. 15

16 The average thickness of the nanoparticle (390nm) -and nanowire-based films (365nm) depicted in Figure 5 were determined by AFM in tapping mode. The device channel length and width was also estimated from the SEM micrographs. The nanowire-based device had a channel width and length of 106 µm and 9.5 µm, respectively. The active area of the device built from nanoparticles was 15.5 µm long and 100 µm wide. Device Height (H) Width (W) Length (L) Nanowire 365nm 106µm 9.5µm Nanoparticle 390nm 100µm 15.5µm Table S2: Active area; channel dimensions of photodetectors depicted in Fig 5 in the manuscript. Because of the slight difference in device dimensions, we report their output characteristics as a function of the electric field applied between the source and drain contacts. For similar reasons, (slight difference in height between the nanowire and the nanoparticle based film) the current density was reported instead of the source-to-drain current. Note that the corresponding raw data (dark and photocurrent I-V charateristics) are depicted in FigS10. The longitudinal electric field was determined as: E d = V d L Where E d is the longitudinal electric field, V d is the applied source-to-drain voltage and L is the device channel length. The Current density was calculated by: J d = I d A Where J d is the longitudinal current density, I d is the source-to-drain current and A is the crosssectional area of the perovskite coatings. 16

17 The area of the nanoparticle film was evaluated as the average height measured by AFM (H) times the width (W) estimated from the SEM micrograph: A = H W The area of the nanowire film was determined by summing up the cross-sectional areas of individual nanowires. They were calculated as the height of individual nanowires measured by AFM (H i ) times its width (W i ) estimated from the SEM micrograph. 17 H i W i i=0 The area coverage was estimated from the SEM top view images. The nanowires covered 27% while the nanoparticles 97% of the active area between the two Pt contacts. Consequently, exposing to a laser light (0.25 Wcm -2, λ=633nm), the calculated incident power was 6.8µW (nanowires) and 38.3µW (nanoparticles). Device Width (W) Length (L) Device Area (WxL) Area Coverage Perovskite Area Incident Power Nanowire 106µm 9.5µm 1007µm 2 27% 271µm 2 6.8µW Nanoparticle 100µm 15.5µm 1550µm 2 97% 1532µm µW Table S3: Parameters used for calculating the light power incident on the two devices In order to compare the incident photon to current conversion efficiency, the external quantum efficiencies (EQE) as a function of the applied source to drain voltage were calculated as follows: EQE(%) = hci ph eλp in

18 where h = Js is the Planck constant; c = ms 1 is the speed of light in vacuum; e = C is the elementary charge; λ is the laser wavelength; I ph is the photocurrent; P in is the incident optical power. Calculated photocurrent (I ph ), responsivity and external quantum efficiency (EQE) values of devices (Fig5) are shown in Table S4. Nanowire Responsivity Nanoparticle Responsivity Vd EQE EQE I ph (na) I ph (na) (V) (ma/w) (%) (ma/w) (%) Table S4: Calculated photocurrent (I ph ), responsivity and external quantum efficiency (EQE) values of devices (Fig5). 18

19 Figure S10: (a) Combined SEM-optical micrographs showing the surface of the thin film composed of nearly isotropic MAPbI 3 particles (bottom) and nanowires (top) with the Pt source-drain contacts deposited by e-beam evaporation (a). The grain boundaries in the nanoparticle based film are clearly perceivable with a green-blue color (a, bottom). Note the absence of such a contrast in the case of filiform perovskites (a, top). Comparison of the dark current density of the nanoparticle and nanowire based devices (b). Comparison of the photocurrent-dark current ratio of the nanoparticle and nanowire based devices (c). (d). Comparison of the external quantum efficiency of the nanoparticle and nanowire based devices. 19

20 Figure S11: Colored SEM micrographs of nanowire-based (a) and nanoparticle-based photodetectors (b) fabricated to test the white light illumination responses. Comparison of photocurrent responses to a white LED light of lux (c). External quantum efficiency calculated from the measured photocurrents (d). The devices shown in Fig S11 were built following the previously used procedure (Fig5). Their performances were assessed both under a white light illumination (LED with light intensity of lux) and red laser (0.25 Wcm -2, λ=633nm). The surface coverage in the active area is 34% (nanowires) and of 83% (nanoparticles). 20

21 Device Width (W) Length (L) Height (H) Area Coverage Incident Power Nanowire 113.9µm 32.3µm 136nm 34% 31.3µW Nanoparticle 108.7µm 34.7µm 138nm 83% 78.5µW Table S5: Parameters used for calculation of the light power incident on the two devices (Fig S11) Effect of the gate Voltage We tested the effect of a transversal electric field applied on the semiconducting MAPbI3 nanowire channel. The device configuration is shown in Figure S12a. As a gate dielectric, 300 nm thick SiO2 was used. We have observed that during the slip-coating process the erosion/friction forces could damage the thinner gate dielectric resulting in a significant leakage current during the measurement. The 300 nm thick dielectric showed higher success rate in the fabrication of working devices with leakage currents under 5 na. A constant gate voltage (up to ±40V) was applied while sweeping the source-to-drain electric field between -5kV/cm and +5kV/cm. Even for relatively high gate voltages (±40V) no modulation of the charge carrier concentration was observed (Fig. S12b). Figure S12: Schematic representation of the photodetector made of filiform perovskites (a). Output characteristics at different gate voltages (V g) (b). As we can clearly see from the inset, gate voltage has no influence on the measured current density. 21

22 Breakdown Voltage determination To study the performance limit and the maximum electric field applicable to the MAPbI3 nanowires under illumination we tested several devices under increasing source-to-drain electric field (an example is shown in Fig S13). Figure S8a shows that when applying a longitudinal field larger than 20kV/cm the repeated I-V curves start to diverge. Decreasing photocurrent and an increasing hysteresis loop appears in the output characteristic (Figure S8a, curve 3). Electric field to values larger than 30kV/cm causes the total breakdown of the device (Figure S8b). Figure S13: Output characteristics of MAPbI 3 nanowire photodetector with increasing longitudinal electric field (a). An electric field exceeding 30±10kV/cm causes irreversible rupture of the active area (b). 22

23 Photodetector stability We tested the reliability of the fabricated photodetectors by measuring the photocurrent of the devices for ~100 cycles (from -200mV to + 200mV, measurement time ~ 1h). In order to prevent humidity induced degradation of the material a drop-casted top PMMA layer (A4) was applied prior the device cyclability test. The device showed identical I-V characteristics up to 300s. After 1h of continuous measurement, the photocurrent increased by ~5% (this improvement is presumably due to some contacts adjustments). Figure S14: Photocurrent of the device after several consecutive cycles. The current is stable until (300s) and it increases with time by about 5% (probably because of contacts adjustment). 23

24 External Quantum Efficiency calculation The external quantum efficiency (E.Q.E.) is the ratio of the number of carriers generated and collected by the photodetector to the number of photons of a given energy incident on the device. For a given incident optical power Pin and a generated photocurrent Iph, E.Q.E. can be calculated as: I ph e P in hν (Equation 1) Where e is elementary charge, h is the Planck constant and ν is the speed of light. The performances of the devices based on nano-perovskites were calculated by assuming that all the incident light was absorbed by the devices and converted into electron-hole pairs, thus neglecting the effect of optical losses due to transmission and reflection. It is important to mention, that the transmission and reflection losses have not been determined in this work. Since the presence of these optical phenomena could highly affect the calculated E.Q.E., in our case the reported values can be seen as characteristic lower bound values for this material. 24

25 Chart 1. Responsivity of state-of-the-art photodetectors Chart S1: Photoresponsivity of several recent state-of-the-art photodetectors based on nanomaterials. 25

26 Reference Material Configuration R (A/W) EQE (%) Response time (s) Light source (nm) V SD (V) V G (V) This Work CH 3NH 3PbI 3 Nanowires < Liu et al. Nature Nano Graphene, single layer (SLG) Stacked SLG-Ta 2O 5-SLG 10 3 ~233x10 3 < Zhang et al. Graphene/ Stacked 1.2x x Nature Sci. Rep MoS 2 SLG/SL-MoS Jin et al. Graphene/ Stacked 1.06x x J.of Mat.Chem HfO 2/ CdSe SLG - HfO 2 - CdSe nanowire Britnell et al. TMDCs/ Stacked ±20 Science Graphene SLG-WS 2-SLG Lopez-Sanchez et al. MoS 2 SL-MoS ~195x Nature Nano Zhang et al. Graphene SLG Nature Comm Quantum dot-like array Konstantatos et al. Nature Nano Graphene/ Quantum dots (QDs) Stacked SLG/BLG- PbS QDs colloidal 4x x Yin et al. MoS 2 SL-MoS ACS NANO

27 Weng et al. ZnO ZnO nanowire 4.7x x ACS AMI Mueller et al. Graphene SLG Nature Phot Xia et al. Graphene SLG < Nature Nano Table S6: Output characteristics of several recent nanoscale photodetectors. R is the responsivity of the device, E.Q.E. is the External Quantum Efficency, V SD is the source-to-drain voltage and V G is the gate voltage. References 1. Poglitsch, A.; Weber, D. The Journal of Chemical Physics 1987, 87, (11), Baikie, T.; Fang, Y.; Kadro, J. M.; Schreyer, M.; Wei, F.; Mhaisalkar, S. G.; Graetzel, M.; White, T. J. J Mater Chem A 2013, 1, (18), Quarti, C.; Grancini, G.; Mosconi, E.; Bruno, P.; Ball, J. M.; Lee, M. M.; Snaith, H. J.; Petrozza, A.; Angelis, F. D. The Journal of Physical Chemistry Letters 2013, 5, (2), Onoda-Yamamuro, N.; Matsuo, T.; Suga, H. Journal of Physics and Chemistry of Solids 1990, 51, (12), Zhengwu, J. I. N. Journal of Materials Science Letters 1997, 16, (20), Liu, C.-H.; Chang, Y.-C.; Norris, T. B.; Zhong, Z. Nat Nano 2014, advance online publication. 7. Zhang, W. J.; Chuu, C. P.; Huang, J. K.; Chen, C. H.; Tsai, M. L.; Chang, Y. H.; Liang, C. T.; Chen, Y. Z.; Chueh, Y. L.; He, J. H.; Chou, M. Y.; Li, L. J. Sci Rep-Uk 2014, Jin, W.; Gao, Z.; Zhou, Y.; Yu, B.; Zhang, H.; Peng, H.; Liu, Z.; Dai, L. Journal of Materials Chemistry C 2014, 2, (9), Britnell, L.; Ribeiro, R. M.; Eckmann, A.; Jalil, R.; Belle, B. D.; Mishchenko, A.; Kim, Y.- J.; Gorbachev, R. V.; Georgiou, T.; Morozov, S. V.; Grigorenko, A. N.; Geim, A. K.; Casiraghi, C.; Neto, A. H. C.; Novoselov, K. S. Science 2013, 340, (6138), Lopez-Sanchez, O.; Lembke, D.; Kayci, M.; Radenovic, A.; Kis, A. Nat Nano 2013, 8, (7),

28 11. Zhang, B. Y.; Liu, T.; Meng, B.; Li, X.; Liang, G.; Hu, X.; Wang, Q. J. Nat Commun 2013, 4, Konstantatos, G.; Badioli, M.; Gaudreau, L.; Osmond, J.; Bernechea, M.; de Arquer, F. P. G.; Gatti, F.; Koppens, F. H. L. Nat Nano 2012, 7, (6), Yin, Z.; Li, H.; Li, H.; Jiang, L.; Shi, Y.; Sun, Y.; Lu, G.; Zhang, Q.; Chen, X.; Zhang, H. Acs Nano 2011, 6, (1), Weng, W. Y.; Chang, S. J.; Hsu, C. L.; Hsueh, T. J. Acs Appl Mater Inter 2011, 3, (2), Mueller, T.; Xia, F.; Avouris, P. Nat Photon 2010, 4, (5), Xia, F.; Mueller, T.; Lin, Y.-m.; Valdes-Garcia, A.; Avouris, P. Nat Nano 2009, 4, (12),