Letter pubs.acs.org/NanoLett
Switching of Photonic Crystal Lasers by Graphene Min-Soo Hwang,† Ha-Reem Kim,† Kyoung-Ho Kim,† Kwang-Yong Jeong,† Jin-Sung Park,† Jae-Hyuck Choi,† Ju-Hyung Kang,† Jung Min Lee,†,‡ Won Il Park,‡ Jung-Hwan Song,§ Min-Kyo Seo,§ and Hong-Gyu Park*,† †
Department of Physics, Korea University, Seoul 02842, Republic of Korea Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Republic of Korea § Department of Physics and Institute for the NanoCentury, KAIST, Daejeon 34141, Republic of Korea ‡
S Supporting Information *
ABSTRACT: Unique features of graphene have motivated the development of graphene-integrated photonic devices. In particular, the electrical tunability of graphene loss enables high-speed modulation of light and tuning of cavity resonances in grapheneintegrated waveguides and cavities. However, efficient control of light emission such as lasing, using graphene, remains a challenge. In this work, we demonstrate on/off switching of single- and double-cavity photonic crystal lasers by electrical gating of a monolayer graphene sheet on top of photonic crystal cavities. The optical loss of graphene was controlled by varying the gate voltage Vg, with the ion gel atop the graphene sheet. First, the fundamental properties of graphene were investigated through the transmittance measurement and numerical simulations. Next, optically pumped lasing was demonstrated for a graphene-integrated single photonic crystal cavity at Vg below −0.6 V, exhibiting a low lasing threshold of ∼480 μW, whereas lasing was not observed at Vg above −0.6 V owing to the intrinsic optical loss of graphene. Changing quality factor of the graphene-integrated photonic crystal cavity enables or disables the lasing operation. Moreover, in the double-cavity photonic crystal lasers with graphene, switching of individual cavities with separate graphene sheets was achieved, and these two lasing actions were controlled independently despite the close distance of ∼2.2 μm between adjacent cavities. We believe that our simple and practical approach for switching in graphene-integrated active photonic devices will pave the way toward designing high-contrast and ultracompact photonic integrated circuits. KEYWORDS: Graphene, photonic crystals, nanolasers, switching
B
graphene-integrated metamaterials were shown to exhibit gateinduced switching and linear modulation of terahertz waves.23 Although the possibility of efficiently controlling desired optical properties in these graphene-based passive devices has been demonstrated, further studies are needed for developing novel active photonic devices incorporating graphene. In particular, it remains a challenge to control the optical loss and lasing operation in graphene-integrated active nanocavities. In this work, we demonstrate on/off switching of single- and double-cavity PhC lasers with monolayer graphene sheets. The optical loss of graphene atop the PhC cavities was efficiently controlled by electrolyte gating with the ion gel. Optically pumped room-temperature lasing is demonstrated in single PhC cavities with graphene, exhibiting a low threshold of ∼480 μW at a gate voltage below −0.6 V, although lasing was not achieved initially owing to the intrinsic optical loss of graphene.
ecause of its unique features, atomic-scale-thickness graphene has been used in a variety of photonic applications.1 For example, high carrier mobility (∼200,000 cm2 V−1 s−1) and high saturation velocity (5.5 × 107 cm/s) of graphene2,3 are advantageous for high-speed optoelectronic devices such as photodetectors4−7 and saturable absorbers.8,9 In addition, graphene exhibits a high optical transmittance, 97.7% in free-standing monolayer graphene,10 and thus has been used in transparent electrodes in nanowire light-emitting diodes11 and electrically driven microdisk lasers.12 However, such a small optical absorption of graphene is not negligible in nanoscale photonic structures. For example, quality (Q) factor of a microdisk cavity decreased ∼12-fold when a five-layer graphene with a thickness of only 1.5 nm was introduced atop the cavity.12 The electrical tunability of graphene loss can be employed to adequately address this issue. Since the demonstration of substantial tuning of optical transitions in monolayer graphene by electrical gating,13−15 high-speed waveguide-integrated electro-absorption modulators16−18 and graphene-photonic crystal (PhC) cavities with tunable cavity resonances19−22 have been successfully developed. In addition, © XXXX American Chemical Society
Received: December 15, 2016 Revised: January 19, 2017 Published: February 6, 2017 A
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. Tuning of graphene optical loss. (a) Photograph of a device used for measuring the graphene transmittance. The monolayer graphene sheet (red rectangle) was placed between the Ti/Au contacts and covered by the ion gel (blue rectangle). A transparent glass substrate was used. Scale bar, 1 cm. (b) Measured transmittance through the glass/graphene/ion gel structure in panel a for a 1550 nm wavelength laser diode, as a function of the gate voltage of graphene Vg (black dots, left axis). The incident laser power was 135 μW. The error bars denote one standard deviation from the average over 30 measurements. The change in the transmittance, ΔT, is plotted as a function of Vg (blue line, right axis); these data were obtained using FEM simulations. The inset shows the calculated Fermi level |EF| (blue line) and the real part of optical conductivity (green line) of graphene. The red dotted line at Vg = 0.9 V indicates the Dirac point or the charge neutrality point (EF = 0).
neutrality point, and increased as Vg increased or decreased from 0.9 V (inset of Figure 1b). The Fermi level is useful for explaining the gate-dependent transmittance change in terms of the electron occupancy of the valence and conduction bands. In particular, the transmittance change is almost zero near Vg = 0.9 V because the graphene sheet is absorptive because of the interband transition from the valence to conduction band.10 On the other hand, the transmittance change increases for other Vg because of the reduction of valence band electrons or the accumulation of conduction band electrons for the interband transition.15 We took advantage of the efficient loss control of graphene to demonstrate the on/off switching of a PhC laser. Figure 2a schematically shows our graphene−PhC laser system as well as its switching operation. The optical loss can be reduced by applying a gate voltage to the graphene sheet, thus achieving lasing from the PhC cavity and switching of the laser. A sideview schematic of the laser structure elucidates the working principle in more detail (Figure 2b). A monolayer graphene sheet was placed on a free-standing InGaAsP slab with a singlecell PhC cavity,27 and a 100 nm-thick SiO2 spacer was inserted between the slab and the graphene sheet. The SiO2 spacer optically separated the graphene sheet from the PhC cavity and reduced the cavity loss initially originating from the graphene sheet. In addition, the ion gel covered the entire graphene− PhC cavity structure and the two metal contacts. A dc electric field was applied to the graphene sheet for electrolyte gating with the ion gel. To examine lasing behaviors obtained by tuning the graphene optical loss, we fabricated graphene−PhC laser structures with ion gels using the following procedure. First, free-standing PhC slab structures, consisting of InGaAsP quantum wells,27,28 were fabricated using electron-beam lithography and dry etching (see Methods), and a 100 nmthick layer of SiO2 was deposited on the structures by plasmaenhanced chemical vapor deposition (PECVD). A large-area graphene sheet, grown by chemical vapor deposition, was transferred atop the SiO2/PhC structures.29 Next, the entire sample including the obtained graphene−PhC structures was moved and attached to the glass substrate, after Ti/Au metal contacts were fabricated on the graphene−PhC structures and the glass substrate. The sample of graphene−PhC structure was
The reduction in the graphene loss as a function of the gate voltage was examined by the transmittance measurement and the corresponding numerical simulation. In addition, individual graphene switching systems were formed in two adjacent PhC lasers, which are separated by ∼2.2 μm. First, to examine the loss control of graphene by electrical gating we fabricated a simple glass/graphene/ion gel structure (Figure 1a). The monolayer graphene sheet was placed between the two contacts and was electrically connected to one of them. The ion gel was cut with a razor blade and transferred onto the graphene sheet and contacts using tweezers.24 The gap between the two contacts was sufficiently narrow; thus, a strong direct current (dc) electric field was applied to the sample with the ion gel.24 We then measured the transmittance of this glass/graphene/ion gel structure as a function of the gate voltage of graphene Vg (black dots, Figure 1b). A continuous wave (cw) laser diode with a wavelength of 1550 nm was used as an incident light source, because the PhC lasers in Figures 2−4 had similar wavelengths. The transmittance was ∼90.8% in the −0.2 ≤ Vg ≤ 1.8 V range, as a result of optical loss and reflection in the graphene sheet, ion gel, and glass substrate. However, increased transmittance was measured approximately for Vg < −0.2 V and Vg > 1.8 V. In particular, the transmittance increased up to ∼91.8% at Vg = −1.1 V, owing to the reduced graphene loss. We note that the optical loss of graphene could be efficiently controlled by electrical gating with the ion gel. By comparing the measured transmittance with simulation results using a theoretical model of the optical conductivity of graphene, one can investigate the gate-dependent variation in the Fermi level EF and determine the charge neutrality point. First, the optical conductivity of graphene was calculated using the gate-dependent formalism, for voltages in the −1.5 < Vg < 2.3 V range (inset of Figure 1b and Methods).21,25,26 Next, a finite-element method (FEM) simulation was performed in which the obtained optical conductivity of graphene was employed, and the transmittance change was calculated as a function of Vg (ΔT; right axis in Figure 1b). We note that the simulated transmittance was in an excellent agreement with the measured one. By comparing the measured and the simulated transmittance changes, the gate-dependent Fermi level |EF| was obtained: |EF| was zero at Vg = 0.9 V, which is the charge B
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. Switching of the graphene-integrated single-cell PhC laser. (a) Schematic of the graphene−PhC laser structure. On/off laser switching is achieved by applying a gate voltage to the graphene sheet. (b) Side-view schematic of the laser structure. The graphene sheet is placed on the freestanding InGaAsP PhC cavity. The 100 nm thick SiO2 layer is between the PhC slab and the graphene sheet. The optical loss of the graphene−PhC laser structure is controlled by electrolyte gating with the ion gel. Inset, a SEM image of the fabricated PhC cavity with graphene. Scale bar, 1 μm. (c) Measured PL spectra of the single-cell PhC laser in the −1.1 ≤ Vg ≤ 0 V range. The peak pump power was fixed at 523 μW for all the spectra. The wavelengths of the resonant peaks remained unchanged. All spectra were normalized by the peak intensity with a wavelength of 1491.3 nm at Vg = −1.1 V. The output power at Vg = −1.1 V was ∼105 times lower than the pump power. (d) Measured peak intensity and fwhm line width (inset) of the resonant peak with a wavelength of 1491.3 nm in (c) versus Vg. (e) Measured mode images at Vg = −0.5, −0.6, and −0.7 V (left to right). Scale bar, 10 μm. (f) Calculated Q factors of the hexapole mode in the single-cell PhC cavity versus Vg. (Inset) calculated electric field intensity profile of the hexapole mode at a wavelength of 1485.4 nm. Scale bar, 1 μm.
placed in the location of the graphene sheet in Figure 1a. A fabricated PhC cavity with graphene is shown in the inset of Figure 2b. Finally, ion gel was transferred to the sample (see Methods). We then characterized the optical properties of the resulting graphene−PhC lasers lasers by optically pumping the fabricated structures using a 980 nm pulse laser diode (10 ns pulse width and a 1% duty cycle) and electrically gating with the ion gel (dc voltage) at room temperature (see Methods). At
each gate voltage of graphene, the light emitted from the structures was collected by a 40× objective lens and analyzed using either a spectrometer or an infrared (IR) camera. We measured the photoluminescence (PL) spectra of the single-cell PhC laser structures at different Vg, for a fixed peak pump power of 523 μW (Figure 2c). First, we varied Vg from 0 to −1.1 V, guaranteeing the stable operation of the ion gel.24 Two resonance peaks were observed at 1463.4 and 1491.3 nm C
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters at Vg = 0 V. Among them, the peak with the wavelength of 1491.3 nm rapidly increased with decreasing Vg, while the wavelength was the same. In particular, the peak intensity (1491.3 nm) at Vg = −1.1 V was ∼240 times larger than the intensity at Vg = 0 V. To clearly see the Vg−dependence of the PL intensities, the peak intensities in Figure 2c were plotted as a function of Vg (Figure 2d). The peak intensity was quite small in the −0.6 V < Vg ≤ 0 V range but significantly increased in the −1.1 V ≤ Vg ≤ −0.6 V range, as shown in Figure 2d. In addition, we measured the full width at half-maximum (fwhm) spectral line width of the peak at 1491.3 nm, as a function of Vg (inset, Figure 2d). The line width rapidly decreased from 1.5 to 0.7 nm by varying Vg from 0 to −0.6 V and increased slightly up to 0.8 nm at Vg = −1.1 V. The measured mode image reveals a central intensity antinode in the single-cell PhC cavity, which becomes brighter as |Vg| increases (Figure 2e). In particular, the mode image is more clearly seen for Vg ≤ −0.6 V, as expected from Figure 2d. Moreover, we measured the PL spectra by varying Vg from 1.0 to 2.0 V, at the peak pump power of 523 μW (Supporting Information Figure S1). The same resonant mode as for the negative Vg was obtained for Vg ≥ 1.7 V, and its peak intensity increased with increasing Vg although it was not lased. The measured peak intensities for the positive Vg are plotted together in Figure 2d. The dependence of the resonant peak intensity on Vg (Figure 2d) was almost the same as the Vg dependence of the transmittance change (ΔT) in the graphene/ion gel structure (Figure 1b). Both the peak intensity and ΔT were negligibly small at low |Vg| but increased at higher |Vg|. For example, as Vg changed from −0.6 to −1.1 V, the peak intensity in the graphene-integrated PhC cavity increased 40-fold while the transmittance increased by ∼1.0% in the monolayer graphene. In addition, a significant change of line width in the PhC cavity was observed in the −0.6 < Vg ≤ 0 V range, although the line width slightly increased again for Vg < −0.6 V because of the thermal effect on the leakage current at higher |Vg|.30 These observations strongly indicate that the optical loss of graphene can be modulated by Vg and directly affect the measured peak intensity and spectral line width of a resonant mode in the graphene-integrated PhC cavity. The resonant mode, which was initially suppressed by the optical loss of graphene, exhibited an increased peak intensity and reduced line width after applying Vg. Next, by utilizing the dependence of the graphene optical conductivity on electrical gating, which was obtained in the inset of Figure 1b, we calculated the Q factor of a resonant mode excited in the graphene−PhC cavity as a function of Vg (Figure 2f and Methods).31 The structural parameters were obtained from the SEM image of a fabricated single-cell PhC cavity (inset, Figure 2b). Additional structures, including the SiO2 spacer and ion gel, were also considered for more accurate calculation. In our FEM simulation, the hexapole mode was excited at a wavelength of 1485.4 nm (inset, Figure 2f), which is similar to the measured lasing wavelength in Figure 2c. We then calculated the Q factor of the hexapole mode (Figure 2f).27,28 The Q factor, related to the total optical loss including the radiative loss into the air and substrate and the graphene loss, increased after electrical gating of graphene, exhibiting a trend similar to that of the transmittance in Figure 2d.20,21 The calculated Q factor at Vg = −1.1 V was 2580, nearly 1.6 times larger than the Q factor of 1610 at Vg = 0 V. This increase in the Q factor at lower Vg enables the measurement of the lasing operation in the PhC cavity. In fact, the calculated Q factor
without graphene was 3750, which is higher than those of the graphene-integrated PhC cavity. The evanescent field overlap of the resonant mode with the top graphene sheet is the main origin of the optical loss. Thus, using the hexapole mode with an originally high Q factor27 sufficiently reduces the optical loss of the graphene-integrated cavity in response to small changes in Vg. To better assess the potential lasing of the observed resonance peak at negative Vg in Figure 2, we measured the peak intensity for different pump power and negative Vg values (Figure 3a). In the −0.6 < Vg ≤ 0.0 V range, the peak intensity
Figure 3. Dependence of PL intensity on the pump power and gate voltage. (a) Measured peak intensity of the graphene-integrated singlecell PhC laser versus the peak pump power and gate voltage Vg. The voltage Vg was varied from 0 V to −1.1 V while the peak pump power was varied from 424 to 570 μW. (b) Peak intensity versus the peak pump power, for the −1.1 ≤ Vg ≤ 0 V range. The threshold pump powers for lasing were ∼480 and ∼520 μW for Vg = −1.1 V and Vg = −0.6 V, respectively. All of the peak intensities were normalized by the maximal intensity at Vg = −1.1 V. Inset, the measured threshold pump powers versus Vg, for the −1.1 ≤ Vg ≤ −0.6 V range. Lasing was not observed for Vg > −0.6 V.
was negligibly small even as the peak pump power increased up to 570 μW. However, in the −1.1 ≤ Vg ≤ −0.6 V range, the peak intensity clearly increased with increasing pump power. Throughout this region, we observed more rapidly increased peak intensities for higher |Vg|. To clarify these properties at each Vg, the peak intensity was plotted versus the peak pump power (Figure 3b). A superlinear increase in the peak intensity was observed for Vg ≤ −0.6 V, indicating a clear lasing operation in our graphene−PhC cavity structure. The lasing D
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 4. Graphene-integrated double-cavity PhC laser. (a) Schematic of a double-cavity PhC laser with top graphene layers. Graphene was disconnected between the two PhC cavities to apply the gate voltages separately, left Vg and right Vg, to the left and right graphene layers, respectively. The direction of Vg was opposite to that in Figures 1−3; the common contact on the glass substrate was grounded in this structure. (b) SEM image of the fabricated double-cavity PhC structure with graphene. A 420 nm wide gap was formed between the left and right graphene layers. Scale bar, 1 μm. (c) Measured PL spectra at (left Vg, right Vg) = (0 V, 0 V) (black line; top of panel), (1 V, 0 V) (red line; middle of panel), and (0 V, 1 V) (blue line; bottom of panel). The incident peak pump power was 3.23 mW in each case. For comparison, all of the peak intensities were normalized by the maximal intensity at (left Vg, right Vg) = (1 V, 0 V). (d) Measured output intensities of the lasing peaks with wavelengths of 1525.0 nm (top of panel) and 1513.0 nm (bottom of panel) versus the peak pump power. In both graphs, the black, red, and blue lines were measured at (left Vg, right Vg) = (0 V, 0 V), (1 V, 0 V), and (0 V, 1 V), respectively.
threshold was ∼480 μW for Vg = −1.1 V. Larger thresholds were measured for smaller |Vg|. The threshold was increased to ∼520 μW for Vg = −0.6 V, and lasing did not occur for Vg > −0.6 V. These features are discernible in the plot that shows the dependence of the threshold on Vg in the region of lasing (inset, Figure 3b). Our systematic measurements showed that on/off switching of the graphene-integrated single-cell PhC laser was for the first time successfully achieved by electrolyte gating of the graphene sheet with the ion gel. The optical loss of graphene was sufficiently reduced for Vg ≤ −0.6 V to excite a lasing mode. Moreover, the optical loss was further reduced at higher |Vg|, thus reducing the lasing threshold (inset, Figure 3b). In fact, the thickness of the SiO2 layer was critical for determining the threshold and the minimal value of |Vg| for the lasing action in the graphene−PhC structure; these values for lasing will further decrease with increasing the SiO2 layer’s thickness or the Q factor of the graphene−PhC cavity. With a faster tuning of the graphene optical loss, our approach to efficient switching of
nanolasers will also enable various photonic applications, such as a high-speed modulation of lasing.16,32 To fully utilize our graphene-based switching system, we extended our approach to laser arrays enabling independent control of lasing operations. In fact, it is not straightforward to achieve independent control of each laser in laser arrays, particularly when the laser cavities are closely located. However, graphene-based switching of individual nanolasers is likely to be feasible, because submicron-sized graphene structures can be easily fabricated using conventional top-down fabrication procedures. Figure 4a shows the schematic of our grapheneintegrated double-cavity PhC laser structure. The center-tocenter distance between two similar PhC cavities was only ∼2.2 μm. Two separate monolayer graphene sheets were placed on top of the two PhC cavities, respectively, with the optical absorption of each PhC cavity controlled by independent electrical gating of graphene. Left and right gate voltages were applied separately to the left and right graphene layers, respectively, while the metal pad on the glass substrate functioned as their common contact. We fabricated a doubleE
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
other ungated graphene−PhC cavity. Furthermore, the coupling between the two PhC cavities was negligibly small because of a relatively large distance between them, as shown by the numerical simulations (Supporting Information Figure S3). Thus, the optical properties of only one cavity were tuned by gating. We also note that the lasing thresholds measured in the double-cavity PhC lasers were ∼4 times larger than those in the single-cavity laser (Figure 3). Such an increase in the thresholds is mainly attributed to a less efficient optical pumping of a single PhC cavity: the pumping laser in the double-cavity PhC structure illuminated the central region between the two cavities. Furthermore, we expect that an effective high-speed modulation of individual lasing actions can be achieved by controlling the optical loss of graphene in a gating structure composed of thin dielectric layers such as Al2O3 or h-BN between the graphene and PhC cavities.32,33 In summary, we demonstrated on/off switching of grapheneintegrated single- and double-cavity PhC lasers by the electrolyte gating of graphene with ion gel. The measured transmittance of a monolayer graphene sheet as a function of the gate voltage and its comparison with numerical simulations suggest that the Q factor of the graphene−PhC cavity mode can increase 1.6 times at the gate voltage of −1.1 V, as compared with the ungated case. Indeed, optically pumped lasing operation was demonstrated for a graphene-integrated singlecavity PhC cavity with a low lasing threshold of ∼480 μW, as the gate voltage of −1.1 V was applied, although lasing was not achieved initially in the graphene−PhC cavity because of the intrinsic graphene loss. In addition, in the graphene-integrated double-cavity PhC laser structure we formed individual graphene switching systems for each cavity. Despite the close distance of only ∼2.2 μm between adjacent PhC cavities, lasing actions were separately controlled as the left and right gate voltages were applied independently. Therefore, we believe that our efficient approach for switching in these graphene-based active photonic devices is promising for designing high-contrast and ultracompact photonic integrated circuits.
cavity PhC structure using procedures similar to those shown in Figure 2 (see Methods). Elaborate aligned electron-beam lithography and oxygen plasma etching were additionally performed for forming a small ∼420 nm wide air gap in the monolayer graphene sheet. The SEM image in Figure 4b shows that this gap was well-aligned on top of the region between the two PhC cavities. Subsequently, ion gel was transferred to cover the entire sample, including the graphene−PhC structure and the metal contacts.24 Then, PL characteristics were measured for the fabricated graphene−PhC structures. The center of the two PhC cavities was optically pumped using a 980 nm wavelength pulse laser diode with a spot size of ∼3 μm, which was larger than the distance between the cavities. Left and right Vg were also applied from 0 to 1 V for separately controlling the optical loss of each graphene. Figure 4c shows the measured PL spectra at (left Vg, right Vg) = (0 V, 0 V) (top of panel), (1 V, 0 V) (middle of panel) and (0 V, 1 V) (bottom of panel) at a fixed incident peak pump power of 3.23 mW. We applied Vg in the direction opposite to that in Figures 1−3 to gate independent voltages to the left and right graphene. In this case, only positive Vg efficiently tuned the output peak intensity. First, at (left Vg, right Vg) = (0 V, 0 V) two sharp resonance peaks were observed at 1513.0 and 1525.0 nm, although their output intensities were not strong. Second, at (left Vg, right Vg) = (1 V, 0 V) a distinctive behavior was observed in the measured PL spectrum: the peak intensity at 1525.0 nm solely increased significantly, whereas the other peak at 1513.0 nm did not change. Third, at (left Vg, right Vg) = (0 V, 1 V) the opposite behavior was observed: the peak at 1513.0 nm increased significantly, but the peak at 1525.0 nm remained unchanged. The small peak at 1520.9 nm did not change (intensity- or wavelength-wise) in any of the cases. The continuous change of the left (right) Vg from 0 to 1 V clearly shows the Vg-dependent behavior of the peak at 1525.0 nm (1513.0 nm) (Supporting Information Figure S2). Next, we measured output intensities of the resonance peaks at 1525.0 nm (top of panel, Figure 4d) and 1513.0 nm (bottom of panel, Figure 4d) versus the peak pump power to assess their lasing behaviors. Again, three sets of gate voltages were examined: (left Vg, right Vg) = (0 V, 0 V) (black line), (1 V, 0 V) (red line), and (0 V, 1 V) (blue line). Initially, at (left Vg, right Vg) = (0 V, 0 V) both resonance peaks at 1525.0 and 1513.0 nm exhibited lasing characteristics with the thresholds of ∼2.6 and ∼2.7 mW, respectively. At (left Vg, right Vg) = (1 V, 0 V) the threshold for the lasing peak at 1525.0 nm decreased down to ∼2.0 mW while the output intensity increased, whereas the light in-light out (L−L) curve for the resonance peak at 1513.0 nm was almost the same as that at (0 V, 0 V). In addition, at (left Vg, right Vg) = (0 V, 1 V) the threshold for the lasing peak at 1513.0 nm decreased down to ∼2.1 mW and the output intensity increased, whereas the L−L curve for the peak at 1525.0 nm was the same as that at (0 V, 0 V). Consequently, a noticeable reduction in the lasing thresholds at (1 V, 0 V) and (0 V, 1 V) occurred, compared with that at (0 V, 0 V); thus, only one lasing mode could be preferentially excited by rational control of Vg. The independent controls of the lasing peaks at 1525.0 and 1513.0 nm by gating the left and right graphene, respectively, suggest that the peak at 1525.0 nm (1513.0 nm) originated from the resonant mode excited in the left (right) PhC cavity. The reduced optical loss of graphene increased the peak intensity and reduced the lasing threshold of the gated graphene−PhC cavity, while no change was observed in the
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b05207. Detailed description of experimental methods and additional figures (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Hong-Gyu Park: 0000-0002-6375-0314 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS H.-G.P. acknowledges support from the National Research Foundation of Korea (NRF) Grant funded by the Korean government (MSIP) (2009-0081565 and 2014M3A6B3063710) and Korea University Future Research Grant. W.I.P. acknowledges support from NRF Grant (NRFF
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
(30) Cho, J. H.; Lee, J.; He, Y.; Kim, B.; Lodge, T. P.; Frisbie, C. D. Adv. Mater. 2008, 20, 686−690. (31) Andonegui, I.; Garcia-Adeva, A. J. Opt. Express 2013, 21, 4072− 4092. (32) Gao, Y.; Shiue, R. J.; Gan, X.; Li, L.; Peng, C.; Meric, I.; Wang, L.; Szep, A.; Walker, D.; Hone, J.; Englund, D. Nano Lett. 2015, 15, 2001−2005. (33) Yu, R.; Pruneri, V.; García De Abajo, F. J. ACS Photonics 2015, 2, 550−558.
2015R1A2A11001426) and M.-K.S. acknowledges support from NRF Grant (2014M3A6B3063709).
■
REFERENCES
(1) Bonaccorso, F.; Sun, Z.; Hasan, T.; Ferrari, A. C. Nat. Photonics 2010, 4, 611−622. (2) Du, X.; Skachko, I.; Barker, A.; Andrei, E. Y. Nat. Nanotechnol. 2008, 3, 491−495. (3) Meric, I.; Han, M. Y.; Young, A. F.; Ozyilmaz, B.; Kim, P.; Shepard, K. L. Nat. Nanotechnol. 2008, 3, 654−659. (4) Konstantatos, G.; Badioli, M.; Gaudreau, L.; Osmond, J.; Bernechea, M.; de Arquer, F. P. G.; Gatti, F.; Koppens, F. H. L. Nat. Nanotechnol. 2012, 7, 363−368. (5) Mueller, T.; Xia, F.; Avouris, P. Nat. Photonics 2010, 4, 297−301. (6) Xia, F.; Mueller, T.; Lin, Y.; Valdes-Garcia, A.; Avouris, P. Nat. Nanotechnol. 2009, 4, 839−843. (7) Liu, C.-H.; Chang, Y.-C.; Norris, T. B.; Zhong, Z. Nat. Nanotechnol. 2014, 9 (4), 273−278. (8) Bao, Q.; Zhang, H.; Wang, Y.; Ni, Z.; Yan, Y.; Shen, Z. X.; Loh, K. P.; Tang, D. Y. Adv. Funct. Mater. 2009, 19, 3077−3083. (9) Sun, Z.; Hasan, T.; Torrisi, F.; Popa, D.; Privitera, G.; Wang, F.; Bonaccorso, F.; Basko, D. M.; Ferrari, A. C. ACS Nano 2010, 4, 803− 810. (10) Nair, R. R.; Grigorenko, A. N.; Blake, P.; Novoselov, K. S.; Booth, T. J.; Peres, N. M. R.; Stauber, T.; Geim, A. Science 2008, 320, 1308. (11) Lee, J. M.; Choung, J. W.; Yi, J.; Lee, D. H.; Samal, M.; Yi, D. K.; Lee, C. H.; Yi, G. C.; Paik, U.; Rogers, J. A.; Park, W. I. Nano Lett. 2010, 10, 2783−2788. (12) Kim, Y.-H.; Kwon, S.-H.; Lee, J. M.; Hwang, M.-S.; Kang, J.-H.; Park, W.; Park, H.-G. Nat. Commun. 2012, 3, 1123. (13) Wang, F.; Zhang, Y.; Tian, C.; Girit, C.; Zettl, A.; Crommie, M.; Shen, Y. R. Science 2008, 320, 206−209. (14) Li, Z. Q.; Henriksen, E. A.; Jiang, Z.; Hao, Z.; Martin, M. C.; Kim, P.; Stormer, H. L.; Basov, D. N. Nat. Phys. 2008, 4, 532−535. (15) Thareja, V.; Kang, J.-H.; Yuan, H.; Milaninia, K. M.; Hwang, H. Y.; Cui, Y.; Kik, P. G.; Brongersma, M. L. Nano Lett. 2015, 15, 1570− 1576. (16) Liu, M.; Yin, X.; Ulin-Avila, E.; Geng, B.; Zentgraf, T.; Ju, L.; Wang, F.; Zhang, X. Nature 2011, 474, 64−67. (17) Kim, J.; Son, H.; Cho, D. J.; Geng, B.; Regan, W.; Shi, S.; Kim, K.; Zettl, A.; Shen, Y.-R.; Wang, F. Nano Lett. 2012, 12, 5598−5602. (18) Ding, Y.; Zhu, X.; Xiao, S.; Hu, H.; Frandsen, L. H.; Mortensen, N. A.; Yvind, K. Nano Lett. 2015, 15, 4393−4400. (19) Gan, X.; Mak, K. F.; Gao, Y.; You, Y.; Hatami, F.; Hone, J.; Heinz, T. F.; Englund, D. Nano Lett. 2012, 12, 5626−5631. (20) Gan, X.; Shiue, R.-J.; Gao, Y.; Mak, K. F.; Yao, X.; Li, L.; Szep, A.; Walker, D.; Hone, J.; Heinz, T. F.; Englund, D. Nano Lett. 2013, 13, 691−696. (21) Majumdar, A.; Kim, J.; Vuckovic, J.; Wang, F. Nano Lett. 2013, 13, 515−518. (22) Kim, K.-H.; Hwang, M.-S.; Kim, H.-R.; Choi, J.-H.; No, Y.-S.; Park, H.-G. Nat. Commun. 2016, 7, 13893. (23) Lee, S. H. S.; Choi, M.; Choi, C.; Choi, S.-Y.; Zhang, X.; Choi, M.; Kim, T.-T.; Liu, M.; Yin, X.; Min, B. Nat. Mater. 2012, 11, 936− 941. (24) Lee, K. H.; Kang, M. S.; Zhang, S.; Gu, Y.; Lodge, T. P.; Frisbie, C. D. Adv. Mater. 2012, 24, 4457−4462. (25) Falkovsky, L. A. J. Phys. Conf. Ser. 2008, 129, 012004. (26) Emani, N. K.; Chung, T.-F.; Kildishev, A. V.; Shalaev, V. M.; Chen, Y. P.; Boltasseva, A. Nano Lett. 2014, 14, 78−82. (27) Kang, J.-H.; Kim, S.-K.; Jeong, K.-Y.; Lee, Y.-H.; Seo, M.-K.; Park, H.-G. Appl. Phys. Lett. 2011, 98, 211116. (28) No, Y.-S.; Ee, H.-S.; Kwon, S.-H.; Kim, S.-K.; Seo, M.-K.; Kang, J.-H.; Lee, Y.-H.; Park, H.-G. Opt. Express 2009, 17, 1679−1690. (29) Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D.; Piner, R.; Velamakanni, A.; Jung, I.; Tutuc, E.; Banerjee, S. K.; Colombo, L.; Ruoff, R. S. Science 2009, 324, 1312−1314. G
DOI: 10.1021/acs.nanolett.6b05207 Nano Lett. XXXX, XXX, XXX−XXX
