Simultaneous Thermoelectric and Optoelectronic Characterization of

Nov 3, 2015 - We also show that under bias, both above- and below-bandgap illumination leads to a photoresponse in the channel with signatures of pers...
0 downloads 4 Views 3MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Simultaneous Thermoelectric and Optoelectronic Characterization of Individual Nanowires François Léonard,*,† Erdong Song,‡ Qiming Li,§ Brian Swartzentruber,∥ Julio A. Martinez,*,‡ and George T. Wang*,§ †

Sandia National Laboratories, Livermore, California 94551, United States Department of Chemical & Materials Engineering, New Mexico State University, Las Cruces, New Mexico 88003, United States § Sandia National Laboratories, Albuquerque, New Mexico 87123, United States ∥ Center for Integrated Nanotechnologies, Sandia National Laboratories, Albuquerque, New Mexico 87185, United States ‡

S Supporting Information *

ABSTRACT: Semiconducting nanowires have been explored for a number of applications in optoelectronics such as photodetectors and solar cells. Currently, there is ample interest in identifying the mechanisms that lead to photoresponse in nanowires in order to improve and optimize performance. However, distinguishing among the different mechanisms, including photovoltaic, photothermoelectric, photoemission, bolometric, and photoconductive, is often difficult using purely optoelectronic measurements. In this work, we present an approach for performing combined and simultaneous thermoelectric and optoelectronic measurements on the same individual nanowire. We apply the approach to GaN/AlGaN core/shell and GaN/AlGaN/GaN core/shell/shell nanowires and demonstrate the photothermoelectric nature of the photocurrent observed at the electrical contacts at zero bias, for above- and below-bandgap illumination. Furthermore, the approach allows for the experimental determination of the temperature rise due to laser illumination, which is often obtained indirectly through modeling. We also show that under bias, both above- and below-bandgap illumination leads to a photoresponse in the channel with signatures of persistent photoconductivity due to photogating. Finally, we reveal the concomitant presence of photothermoelectric and photogating phenomena at the contacts in scanning photocurrent microscopy under bias by using their different temporal response. Our approach is applicable to a broad range of nanomaterials to elucidate their fundamental optoelectronic and thermoelectric properties. KEYWORDS: Nanowires, GaN, GaN/AlGaN, photocurrent, thermoelectric, photogating

S

measurements (including spatially resolved) neither the Seebeck coefficient nor the temperature increase during illumination are directly measured. Recent progress to identify the photothermoelectric effect has been made in atomically thin materials by using electrostatic gating in conjunction with the Mott relation or by employing double-gated structures.18,27,28 A combination of optoelectronic and thermoelectric measurements on the same device would provide a new approach to establish the photocurrent mechanism and its characteristic features, in particular for materials or devices where electrostatic gating is not feasible or practical, or when the Mott relation is not applicable. The value of this approach has recently been demonstrated for photodetectors based on macroscopic films of carbon nanotubes29,30 where the macroscopic size of the devices allows their transfer between different measurement platforms. This approach has not been applied to

emiconducting nanowires (NWs) have attracted interest for numerous applications in electronics1−4 and optoelectronics.4−11 Progress toward useful devices requires a detailed understanding of the mechanisms of electronic transport in NWs, charge injection at contacts, and photocurrent generation. For example, in the case of solar cells or photodetectors, the particular photocurrent mechanism determines the quantum efficiency and the bases for optimal device design. However, differentiating between the possible mechanisms such as photothermoelectric, bolometric, photovoltaic, and photoemission in NWs is challenging because of their often similar and overlapping signatures. There is currently extensive discussion in the scientific literature about differentiating these mechanisms in NWs,12−17 carbon nanotubes,18−20 graphene,21,22 and 2D materials.23−25 While these studies have provided key insights, the determination of the photoresponse mechanisms can be difficult if only photocurrent measurements are used. For example, the photothermoelectric effect is argued to dominate the photoresponse in graphene26 and 2D materials.25 Unfortunately, in standard photocurrent © XXXX American Chemical Society

Received: September 4, 2015 Revised: October 13, 2015

A

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. (a) Cross-sectional STEM image of a GaN/AlGaN nanowire. (b) SEM image of the platform for thermoelectric and optoelectronic measurements. (c) Illustration of the combined thermoelectric and optoelectronic platform. (d) Typical I−V curves at 300 K for GaN/AlGaN and GaN/AlGaN/GaN nanowires.

process in situ with MOCVD deposition of an AlGaN shell layer at ∼1050 °C and 75 Torr.34,40,41 For GaN/AlGaN/GaN core/shell/shell NWs, the AlGaN shell layer growth was followed by a GaN shell layer grown at the same pressure and temperature. Cross-section scanning transmission electron microscopy (STEM) images show AlGaN shell thicknesses ranging from approximately 10−35 nm thick (Figure 1a) and Al contents from ∼20−30% (bandgap ∼4 eV) as determined by energy dispersive X-ray spectroscopy. The synthesized NWs, with lengths of ∼15 μm, are tapered with side lengths increasing from ∼150 nm at the tip of the NW to ∼250 nm at the base. The characterization platform was produced by standard photolithography methods on a 1 μm-thick LPCVD silicon nitride deposited on a lightly doped silicon substrate. Electronbeam (E-beam) metal evaporation was performed to sequentially deposit 5 nm Ti (first metal layer), 50 nm Au, 30 nm Ti, 100 nm of Al, and 20 nm Ti followed by metal lift-off in acetone for ∼4 h. This step was used to produce the heater, thermometers, and main contact leads. E-beam lithography patterning between contacts 2 and 3 followed by fluorine plasma etching was carried out to produce a 300 nm deep trench between contacts 2 and 3 in order to increase the temperature gradient during Seebeck coefficient measurements. Single nanowires were cleaved from the growth substrate and placed on top of the contact metal leads of the characterization platform using a nanomanipulator tool developed by our group (a field-emission SEM, JEOL 6701 F modified to include two probe tips).42 E-beam lithography was carried out to pattern contact lines on top of the nanowires in order to achieve good electrical and thermal contacts. This was achieved by using ebeam metal evaporation of 80 nm Ti (first layer), 80 nm of Al, 20 nm of Ti, and 30 nm of Au after PMMA development. Liftoff of the E-beam resist was carried out in acetone, followed by rapid thermal annealing at 750 °C for 30 s in forming gas (5% H2 in N2) to improve contact quality.43 This procedure produced nanowires suspended between contacts (Figure 1b).

individual nanostructures due to the need for an integrated measurement platform. In the more specific case of GaN and GaN/AlGaN nanowires, previous global illumination studies31−34 have shown photogating due to surface band-bending. However, spatially resolved photocurrent measurements have not been presented in this material system, nor the zero-bias versus finite-bias properties revealed. Understanding photocurrent mechanisms in these materials is relevant to their potential application as UV photodetectors.27,35 Furthermore, thermoelectric measurements of individual III−N NWs are scarce36 and have only addressed the case of GaN NWs, instead of the core/shell GaN/AlGaN NWs studied here. Here we address the challenges of identifying photocurrent mechanisms and photophysics of III−N core/shell NWs by presenting an approach for the simultaneous thermoelectric and optoelectronic characterization of individual semiconducting NWs. We apply this approach to GaN, GaN/AlGaN core/ shell, and GaN/AlGaN/GaN core/shell/shell NWs and study both above- and below-bandgap illumination at the contacts and in the channel. A thermoelectric measurement on the same platform and on the same nanowire shows the photothermoelectric origin of the response at the contacts, which originates from a large Seebeck coefficient in these III−N NWs. Furthermore, scanning photocurrent microscopy gives unusual profiles as a function of source-drain bias due to the concomitant presence of photothermoelectric and photogating mechanisms at and near the contacts, which can be revealed due to their different temporal signatures. Unintentionally n-type doped GaN NWs (bandgap ≈ 3.4 eV) were grown by metal−organic chemical vapor deposition (MOCVD) at ∼900 °C and 450 Torr on r-plane sapphire using thin Ni films deposited by e-beam evaporation as catalyst for vapor−liquid−solid (VLS) growth.37−39 The NWs grow along the [112̅0] direction with triangular cross sections comprising a (0001̅) facet and two (11̅01) facets. GaN/AlGaN core/shell NWs were grown separately by following the GaN NW growth B

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 2. (a) UV-SPCM at zero bias for a GaN/AlGaN/GaN nanowire. Dotted lines indicate position of NW and source and drain contacts. (b) UV-SPCM for the same nanowire as in panel (a) at a bias of −0.1 V. (c) Photocurrent at zero bias as a function of incident UV laser power when the laser is positioned over the source contact. (d) Time dependence of the photocurrent at zero bias when the UV laser is focused on the source contact. (e) Photocurrent as a function of incident laser power when the UV laser is focused in the channel for a drain-source bias of −0.1 V. (f) Time dependence of the photocurrent for a drain-source bias of 0.1 V when the UV laser is focused in the channel and turned ON and OFF. The inset shows the turn-on behavior in more detail.

coefficient35 of bulk GaN is on the order of 10−3 cm−1 for above-bandgap illumination we estimate that ∼0.1% of the incident light is absorbed by the NW (assuming a diffractionlimited spot size). The sub-bandgap absorption is much lower and can vary by orders of magnitude depending on the density of gap states.35 Electronic and Thermoelectric Measurements. Electrical and thermoelectric transport properties as a function of temperature for our samples were studied prior to optoelectronic studies. Current−voltage (I−V) characteristics were measured to corroborate the presence of Ohmic contacts (Figure 1d). In all cases, we observe linear I−V curves with NW conductivities from ∼53 S/cm to ∼72 S/cm for GaN/AlGaN, and ∼63 S/cm for GaN/AlGaN/GaN at room temperature. The linear I−V curves in conjunction with the high conductivity implies good Ohmic contacts and highly doped NWs. The linear I−V curves observed for all NWs also indicate that the shells do not create a significant barrier for charge injection. We further characterized the electronic transport properties by measuring the temperature dependence of the four-probe electrical conductivity from 25 to 310 K. In all cases, we find that the I−V curves remain linear, implying that Schottky barriers are insignificant in these devices, regardless of the number of shells. We measured the thermoelectric properties of a number of GaN/AlGaN and GaN/AlGaN/GaN nanowires as detailed in the Supporting Information. The Seebeck coefficient at 300 K is found to be −78 μV/K for GaN/AlGaN/GaN, and between −120 and −150 μV/K for GaN/AlGaN nanowires. The negative Seebeck coefficients imply n-type behavior in all these NWs, and the Seebeck values are typical of highly doped samples:45 the values for GaN/AlGaN/GaN correspond to a GaN doping of ∼1019 cm−3, while the value for the GaN/ AlGaN samples corresponds to a doping of ∼2 × 1018 cm−3 for

Samples were electrically tested in a probe station and later mounted in a helium closed-loop cryostat to perform thermoelectric measurements from 25 to 310 K. The Seebeck coefficient of our nanowires were obtained by measuring the thermal voltage (ΔVth) and the temperature difference (ΔT) between leads 2 and 3 when a DC current was applied to the heater (lead 5, Figure 1b) and applying the relation S = −ΔVth/ ΔT36,44 (see Supporting Information for details). Measurements were taken in the dark at a pressure of around 10−6 Torr or lower. Four-probe electrical conductivity was measured with a lock-in amplifier set at 13 Hz frequency with source-drain bias applied between leads 1 and 4 and voltage drop measured between leads 2 and 3 (Figure 1b). We used scanning photocurrent microscopy (SPCM) at room temperature to characterize the optoelectronic properties of the devices and to perform combined thermoelectric and optoelectronic measurements (Figure 1c). For sub-bandgap studies, the beam from a red HeNe laser (633 nm, 1.95 eV) was fed into a microscope and focused on the devices using a 50×, NA = 0.55 objective, and the current was measured using a DL 1211 current preamplifier and a probe station. Different beam powers (measured by replacing the sample with a Newport 818 photodetector) were realized with a variable density filter placed between the laser source and the microscope. A computer-controlled stage was used to move the devices in 50 nm steps with respect to the laser beam, while the current was recorded. The intensity of the reflected light was collected simultaneously to create spatially resolved reflection images that can be analyzed and overlapped with SPCM images to unambiguously identify the positions of the electrodes and the nanowire. Additional details of the SPCM measurements can be found in the Supporting Information and in a previous publication.24 For the above-bandgap measurements, we focused the beam of a HeCd laser (325 nm, 3.8 eV) using a 15×, NA = 0.28 reflective objective. Since the absorption C

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 3. (a) Sub-bandgap SPCM at zero bias for a GaN/AlGaN/GaN nanowire. (b) Photocurrent at zero bias as a function of incident laser power when the red laser is focused on the source electrode. (c) Time dependence of the photocurrent at zero bias when the red laser is focused on the source contact. The inset shows the decay in more detail. (d) Voltage across a GaN/AlGaN nanowire as a function of the voltage applied to the heater in the dark and when either of the two contacts is illuminated with the focused red laser. The two sets of dark/light data were acquired as the laser was turned on and off.

While the above-bandgap photoresponse under bias can be understood from photogating, the zero-bias photocurrents at the contacts must have a different origin because of the different power and time dependence. To study it in more detail, we performed sub-bandgap SPCM with a red laser as shown in Figure 3a. Similar to the zero bias case for UV-SPCM, at zero bias the photocurrent is only observed at the metal contacts, with opposite sign at the two contacts. The photocurrent depends linearly on the laser power (Figure 3b) and the time dependence is limited by the preamp (Figure 3c), consistent with the UV results. Photocurrent due to bandbending in the nanowire near the contacts is not consistent with our observations since the maximum photocurrent in our case is always situated directly on the electrode, for both aboveand below-bandgap illumination. In addition, the bolometric and photoconductive mechanism do not apply due to the zero bias. Therefore, we need to differentiate between the photothermal and photoemission mechanisms, which can be done using our thermoelectric and optoelectronic platform, as we now discuss. Photoresponse Mechanism at Contacts. Having performed optoelectronic and thermoelectric characterization on the same NW, we are in a position to differentiate between the photothermoelectric and photoemission mechanisms for the photoresponse observed at the contacts at zero bias. In the case of the photothermoelectric effect, local illumination causes a local temperature increase, generating a voltage across the two contacts given by

AlGaN. These values are consistent with the measured linear I−V characteristics and high conductivities mentioned above. Optoelectronic Measurements. We first measured the above bandgap photoresponse of the devices using UV-SPCM. Figure 2a shows the UV-SPCM map for a GaN/AlGaN/GaN NW at zero bias; two photocurrent spots of opposite sign are observed when the laser is over the two electrodes with a photocurrent magnitude of about 1 nA, with no measurable photocurrent in the channel. In contrast, when a source-drain bias is applied (Figure 2b), the photocurrent is observed when the laser is over the channel with a magnitude that increases by a factor of ∼200, and that changes sign with the sign of the applied voltage (not shown). The different origins of the above-bandgap photocurrent at the contacts and in the channel can be revealed by considering their power and time dependence. In the case of contact illumination at zero external bias the photocurrent shows a linear dependence with laser power, and time dynamics that are limited by the preamp time constant of 30 ms (Figure 2c,d). In contrast, channel illumination at finite bias gives strongly sublinear power dependence and persistent photoconductivity (Figure 2e,f). This behavior has previously been observed for global illumination and shown to arise due to photogating of the channel, whereby excited electrons and holes are spatially separated by radial surface band-bending; this leads to trapping of holes at the nanowire surface, sharpening of the radial bandbending, and an increase of the effective conducting crosssection.31−33 This effect leads to a current increase when a finite source-drain bias is applied. Importantly, the power and time dependence rule out bolometric (linear power dependence, exponential time decay) and photoconductive phenomena (linear or square root power dependence, exponential time decay) as the sources of the photocurrent at finite bias.

ΔV = VD − VS = (SM − S NW )(TD − TS)

(1)

where SNW is the nanowire Seebeck coefficient, SM is the Seebeck coefficient of the metal contact, TD is the drain temperature, and TS is the source temperature. Because SM ≪ D

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 4. (a) Sub-bandgap SPCM at a bias of −0.1 V for a GaN/AlGaN nanowire. The dark current was subtracted from the data to show only the photocurrent. (b) Time-dependence of the photocurrent when the red laser is focused in the channel for a drain-source bias of −0.1 V. (c) Time dependence of the photocurrent when the red laser is focused on the source contact for different bias voltages. (d) Time dependence of the photocurrent when the UV laser is focused on the source contact for a bias of −0.02 V.

SNW at room temperature we have ΔV ≈ − SNW(TD − TS). In the configuration used to measure the SPCM, the current measured by the preamp is given by I = −ΔV/RNW where RNW is the nanowire resistance (see Supporting Information). Therefore, the photothermoelectric (PTE) current due to illumination should be equal to IPTE = SNW(TD − TS)/RNW. When the drain electrode is illuminated, we have TD > TS; because our direct measurement of SNW gives negative values, IPTE should be negative, and this is precisely what we observe in our optoelectronic measurements. To further show the photothermoelectric origin of the signal, we performed a simultaneous thermoelectric and optoelectronic measurement (Figure 1c and Supporting Information). In this experiment, we use the heater to generate a thermal gradient across the source and drain as in the normal thermoelectric measurement, and at the same time we focus the red laser on a particular location on the nanowire device. Figure 3d shows the results of such measurements for a GaN/ AlGaN NW when the laser is focused on either electrode. In the absence of light, the voltage across the NW depends on the heater power (quadratically on the heater voltage); for illumination of the already hot (cold) electrode, the voltage increases (decreases) indicating that the light serves the same role as heat. The combined SPCM-TE platform also allows us to extract the temperature increase due to laser illumination since we know the actual value of the Seebeck coefficient for the specific nanowire under measurement. From Figure 3b and eq 1, we extract a temperature increase of 16.5 K/mW; at the largest power in Figure 3b this corresponds to a temperature increase of 2.3 K. From this measured temperature increase we can extract the thermal properties of the contact/NW system. For example, by modeling the metal contact as a thin film with heat

transfer at the bottom (Supporting Information) we extract a heat transfer coefficient of ∼2 × 108 W/m2 K, a value typical of metal/semiconductor interfaces.46 Furthermore, the same model gives a time scale for heat dissipation of ∼5 ns, consistent with our observation of the time response limited by our measurement electronics. The photoemission mechanism, whereby electrons are photoexcited over a Schottky barrier at the contact,14 can be ruled-out based on the Ohmic nature of the contacts in our NWs and because of the negligible optical transmission through the thick metal stack that forms the contacts. Indeed, we estimate that the attenuation through the contact should be on the order of 10−11, implying that very little of the incoming light makes it to the nanowire/metal interface (Supporting Information). However, a thermal model based on optical heating at the contact surface predicts temperatures on the order of a few K/mW at the metal/nanowire interface (Supporting Information). Having established the photothermoelectric nature of the signal observed at the metal contacts for above- and belowbandgap illumination at zero external bias, we now turn to subbandgap SPCM at finite bias. Figure 4a shows the SPCM map taken with a red laser at a bias voltage of −0.1 V. In the channel we observe a uniform photocurrent along the length of the nanowire; this suggests the presence of defect and surface states34,47 in the nanowire bandgap, which give rise to a similar photogating mechanism as for UV illumination. This is supported by the long decay time when the laser is positioned in the channel (Figure 4b). The SPCM map of Figure 4a shows a remarkable feature: the two photocurrent peaks at the metal contacts have the same sign. This is qualitatively different from the maps of Figure 2a or Figure 3a, and also from a number of reports in the literature E

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

■

on a broad range of nanomaterials.12−15,18−26 To understand the origin of this behavior, we focused the red laser in the center of one of the contacts and measured the time dependence of the photocurrent at different source-drain voltages. Figure 4c shows that the sign of the photocurrent depends on the bias voltage: at small voltages, the photocurrent is positive, but it becomes negative at larger biases. Furthermore, at intermediate voltages, the photocurrent is initially positive but becomes negative when the light is turned off (Figure 4d). The SPCM map and the photocurrent dynamics can be understood by considering the coexistence of photothermoelectric and photogating phenomena. At zero bias, it is clear that the photocurrent is due to the photothermoelectric effect with a short time scale for decay, while channel illumination at finite bias gives a longer rise time, and a much longer decay time, with a photocurrent of opposite sign. Thus, at short times after the light is turned on, the photocurrent is photothermoelectric and positive; as time increases the photogating contribution increases, causing the total photocurrent to decrease, and even become negative at large enough bias. An additional signature of this effect is the photocurrent overshoot at early times in Figure 4d. When the light is turned off, the photothermoelectric current decreases rapidly, leaving the slowly decaying photogating response. The presence of concomitant photothermoelectric and photogating phenomena at and near the contacts explains the unusual SPCM map in Figure 4a: at the bias of −0.1 V, the positive zero-bias photocurrent at the source electrode becomes negative due to the dominance of the photogating effect. The observation of spatial and dynamic photocurrent reversal at contacts requires the coexistence of two photoresponses with opposite sign and different dynamics. Clearly, this requires that both phenomena be active during illumination. In SPCM experiments, the finite size of the laser spot means that contact and channel phenomena will invariably be captured when the contacts are illuminated by the laser unless special designs are made (e.g., very wide electrodes) to prevent one or the other. However, even in this case joint phenomena should be expected when the laser is within a spot size from the electrode edge. In conclusion, we present a combined thermoelectric and optoelectronic measurement approach for the characterization of individual nanostructures. The method allows for the direct measurement of the Seebeck coefficient of individual nanostructures simultaneously with their optoelectronic response, thus eliminating some of the uncertainties when identifying the active photoresponse mechanisms. When applied to III−N core−shell nanowires, it allows identification of the photoresponse mechanisms at electrical contacts and in the channel; the temperature increases during illumination and reveals the presence of intertwined phenomena at or in the near-contact region. More generally, this approach should be applicable to a broad range of nanomaterials and nanosystems, allowing for an improved understanding and optimization of optoelectronic nanodevice performance. In addition, modifications of the thermoelectric platform could lead to characterization of systems with spatially varying Seebeck coefficients, for example, by placing the heating element in the channel or by taking advantage of scanning thermal probes to provide local heating.

Letter

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b03572. Details of the optoelectronic and thermoelectric measurements. Estimates of temperature increase during contact illumination. Estimates of optical transmission through metal contact. Thermal model for contact heating. Additional data (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: fl[email protected]. *E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division. E.S. and J.A.M. acknowledge partial support from the New Mexico State University. This work was performed, in part, at the Center for Integrated Nanotechnologies, an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science by Los Alamos National Laboratory (Contract DE-AC5206NA25396) and Sandia National Laboratories (Contract DE-AC04-94AL85000). Sandia National Laboratories is a multiprogram laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC0494AL85000.

■

REFERENCES

(1) Li, Y.; Xiang, J.; Qian, F.; Gradecak, S.; Wu, Y.; Yan, H.; Blom, D. A.; Lieber, C. M. Nano Lett. 2006, 6, 1468−1473. (2) Chuang, S.; Gao, Q.; Kapadia, R.; Ford, A. C.; Guo, J.; Javey, A. Nano Lett. 2013, 13, 555−558. (3) Razavieh, A.; Mehrotra, S.; Singh, N.; Klimeck, G.; Janes, D.; Appenzeller, J. Nano Lett. 2013, 13, 1549−1554. (4) Howell, S. L.; Padalkar, S.; Yoon, K.; Li, Q.; Koleske, D. D.; Wierer, J. J.; Wang, G. T.; Lauhon, L. J. Nano Lett. 2013, 13, 5123− 5128. (5) Mariani, G.; Wong, P.-S.; Katzenmeyer, A. M.; Léonard, F.; Shapiro, J.; Huffaker, D. L. Nano Lett. 2011, 11, 2490−2494. (6) Saxena, D.; Mokkapati, S.; Parkinson, P.; Jiang, N.; Gao, Q.; Tan, H. H.; Jagadish, C. Nat. Photonics 2013, 7, 963−968. (7) Wang, R.; Nguyen, H. P. T.; Connie, A. T.; Lee, J.; Shih, I.; Mi, Z. Opt. Express 2014, 22, A1768−A1775. (8) Wallentin, J.; Anttu, N.; Asoli, D.; Huffman, M.; Åberg, I.; Magnusson, M. H.; Siefer, G.; Fuss-Kailuweit, P.; Dimroth, F.; Witzigmann, B.; Xu, H. Q.; Samuelson, L.; Deppert, K.; Borgström, M. T. Science 2013, 339, 1057−1060. (9) Deshpande, S.; Heo, J.; Das, A.; Bhattacharya, P. Nat. Commun. 2013, 4, 1675. (10) Wong, A. B.; Brittman, S.; Yu, Y.; Dasgupta, N. P.; Yang, P. Nano Lett. 2015, 15, 4096−4101. (11) Zhang, P.; Liu, P.; Siontas, S.; Zaslavsky, A.; Pacifici, D.; Ha, J.Y.; Krylyuk, S.; Davydov, A. V. J. Appl. Phys. 2015, 117, 125104. (12) Persano, A.; Taurino, A.; Prete, P.; Lovergine, N.; Nabet, B.; Cola, A. Nanotechnology 2012, 23, 465701.

F

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (13) Triplett, M.; Yang, Y.; Léonard, F.; Talin, A. A.; Islam, M. S.; Yu, D. Nano Lett. 2015, 15, 523−529. (14) Yoon, K.; Hyun, J. K.; Connell, J. G.; Amit, I.; Rosenwaks, Y.; Lauhon, L. J. Nano Lett. 2013, 13, 6183−6188. (15) Thunich, S.; Prechtel, L.; Spirkoska, D.; Abstreiter, G.; Fontcuberta i Morral, A.; Holleitner, A. W. Appl. Phys. Lett. 2009, 95, 083111. (16) Chen, G.; Sun, G.; Ding, Y. J.; Prete, P.; Miccoli, I.; Lovergine, N.; Shtrikman, H.; Kung, P.; Livneh, T.; Spanier, J. E. Nano Lett. 2013, 13, 4152−4157. (17) Fu, D.; Zou, J.; Wang, K.; Zhang, R.; Yu, D.; Wu, J. Nano Lett. 2011, 11, 3809−3815. (18) DeBorde, T.; Aspitarte, L.; Sharf, T.; Kevek, J. W.; Minot, E. D. ACS Nano 2014, 8, 216−221. (19) Amer, M. R.; Chang, S.-W.; Cronin, S. B. Small 2015, 11, 3119− 3123. (20) Nanot, S.; Cummings, A. W.; Pint, C. L.; Ikeuchi, A.; Akiho, T.; Sueoka, K.; Hauge, R. H.; Léonard, F.; Kono, J. Sci. Rep. 2013, 3, 1335. (21) Freitag, M.; Low, T.; Xia, F.; Avouris, P. Nat. Photonics 2013, 7, 53−59. (22) Park, J.; Ahn, Y. H.; Ruiz-Vargas, C. Nano Lett. 2009, 9, 1742− 1746. (23) Wu, C.-C.; Jariwala, D.; Sangwan, V. K.; Marks, T. J.; Hersam, M. C.; Lauhon, L. J. J. Phys. Chem. Lett. 2013, 4, 2508−2513. (24) Klee, V.; Preciado, E.; Barroso, D.; Nguyen, A. E.; Lee, C.; Erickson, K. J.; Triplett, M.; Davis, B.; Lu, I. H.; Bobek, S.; McKinley, J.; Martinez, J. P.; Mann, J.; Talin, A. A.; Bartels, L.; Léonard, F. Nano Lett. 2015, 15, 2612−2619. (25) Buscema, M.; Barkelid, M.; Zwiller, V.; van der Zant, H. S. J.; Steele, G. A.; Castellanos-Gomez, A. Nano Lett. 2013, 13, 358−363. (26) Xu, X.; Gabor, N. M.; Alden, J. S.; van der Zande, A. M.; McEuen, P. L. Nano Lett. 2010, 10, 562−566. (27) Cyril, P.; Akira, H.; Motoaki, I.; Theeradetch, D.; Hiroshi, A.; Isamu, A. Jpn. J. Appl. Phys. 2000, 39, L387. (28) Gabor, N. M.; Song, J. C. W.; Ma, Q.; Nair, N. L.; Taychatanapat, T.; Watanabe, K.; Taniguchi, T.; Levitov, L. S.; Jarillo-Herrero, P. Science 2011, 334, 648−652. (29) He, X.; Wang, X.; Nanot, S.; Cong, K.; Jiang, Q.; Kane, A. A.; Goldsmith, J. E. M.; Hauge, R. H.; Léonard, F.; Kono, J. ACS Nano 2013, 7, 7271−7277. (30) He, X.; Fujimura, N.; Lloyd, J. M.; Erickson, K. J.; Talin, A. A.; Zhang, Q.; Gao, W.; Jiang, Q.; Kawano, Y.; Hauge, R. H.; Léonard, F.; Kono, J. Nano Lett. 2014, 14, 3953−3958. (31) Calarco, R.; Marso, M.; Richter, T.; Aykanat, A. I.; Meijers, R.; v. d. Hart, A.; Stoica, T.; Lüth, H. Nano Lett. 2005, 5, 981−984. (32) Chen, H.-Y.; Chen, R.-S.; Rajan, N. K.; Chang, F.-C.; Chen, L.C.; Chen, K.-H.; Yang, Y.-J.; Reed, M. A. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 205443. (33) Sanford, N. A.; Blanchard, P. T.; Bertness, K. A.; Mansfield, L.; Schlager, J. B.; Sanders, A. W.; Roshko, A.; Burton, B. B.; George, S. M. J. Appl. Phys. 2010, 107, 034318. (34) Armstrong, A.; Li, Q.; Lin, Y.; Talin, A. A.; Wang, G. T. Appl. Phys. Lett. 2010, 96, 163106. (35) Ambacher, O.; Rieger, W.; Ansmann, P.; Angerer, H.; Moustakas, T. D.; Stutzmann, M. Solid State Commun. 1996, 97, 365−370. (36) Lee, C.-H.; Yi, G.-C.; Zuev, Y. M.; Kim, P. Appl. Phys. Lett. 2009, 94, 022106. (37) Li, Q.; Creighton, J. R.; Wang, G. T. J. Cryst. Growth 2008, 310, 3706−3709. (38) Wang, G. T.; Talin, A. A.; Werder, D. J.; Creighton, J. R.; Lai, E.; Anderson, R. J.; Arslan, I. Nanotechnology 2006, 17, 5773. (39) Li, Q.; Wang, G. T. Appl. Phys. Lett. 2008, 93, 043119. (40) Kar, A.; Li, Q.; Upadhya, P. C.; Ah Seo, M.; Wright, J.; Luk, T. S.; Wang, G. T.; Prasankumar, R. P. Appl. Phys. Lett. 2012, 101, 143104. (41) Baird, L.; Ong, C. P.; Cole, R. A.; Haegel, N. M.; Talin, A. A.; Li, Q.; Wang, G. T. Appl. Phys. Lett. 2011, 98, 132104.

(42) Martinez, J. A.; Cho, J.-H.; Liu, X.; Luk, T. S.; Huang, J.; Picraux, S. T.; Sullivan, J. P.; Swartzentruber, B. S. Appl. Phys. Lett. 2013, 102, 103101. (43) Van Daele, B.; Van Tendeloo, G.; Ruythooren, W.; Derluyn, J.; Leys, M. R.; Germain, M. Appl. Phys. Lett. 2005, 87, 061905. (44) Martinez, J. A.; Provencio, P. P.; Picraux, S. T.; Sullivan, J. P.; Swartzentruber, B. S. J. Appl. Phys. 2011, 110, 074317. (45) Sztein, A.; Haberstroh, J.; Bowers, J. E.; DenBaars, S. P.; Nakamura, S. J. Appl. Phys. 2013, 113, 183707. (46) Stevens, R. J.; Smith, A. N.; Norris, P. M. J. Heat Transfer 2005, 127, 315−322. (47) Polenta, L.; Rossi, M.; Cavallini, A.; Calarco, R.; Marso, M.; Meijers, R.; Richter, T.; Stoica, T.; Lüth, H. ACS Nano 2008, 2, 287− 292.

G

DOI: 10.1021/acs.nanolett.5b03572 Nano Lett. XXXX, XXX, XXX−XXX