Flexible Electronics
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers Hongkyw Choi, Jin Sik Choi, Jin-Soo Kim, Jong-Ho Choe, Kwang Hyo Chung, Jin-Wook Shin, Jin Tae Kim, Doo-Hyeb Youn, Ki-Chul Kim, Jeong-Ik Lee, Sung-Yool Choi, Philip Kim, Choon-Gi Choi,* and Young-Jun Yu*
Graphene
leading to high surface-to-volume ratio and outstanding conductivity is applied for gas molecule sensing with fully utilizing its unique transparent and flexible functionalities which cannot be expected from solid-state gas sensors. In order to attain a fast response and rapid recovering time, the flexible sensors also require integrated flexible and transparent heaters. Here, large-scale flexible and transparent gas molecule sensor devices, integrated with a graphene sensing channel and a graphene transparent heater for fast recovering operation, are demonstrated. This combined all-graphene device structure enables an overall device optical transmittance that exceeds 90% and reliable sensing performance with a bending strain of less than 1.4%. In particular, it is possible to classify the fast (≈14 s) and slow (≈95 s) response due to sp2-carbon bonding and disorders on graphene and the self-integrated graphene heater leads to the rapid recovery (≈11 s) of a 2 cm × 2 cm sized sensor with reproducible sensing cycles, including full recovery steps without significant signal degradation under exposure to NO2 gas.
1. Introduction Utilizing the electrical response to adsorbed molecules on sp2-bonded carbon networks, field effect transistors (FETs) based on graphene and reduced graphene oxide (rGO) have
H. Choi, Dr. J. S. Choi, Dr. J.-S. Kim, Dr. J.-H. Choe, Dr. K. H. Chung, Dr. J. T. Kim, Dr. D.-H. Youn, Prof. K.-C. Kim, Dr. C.-G. Choi, Dr. Y.-J. Yu Creative Research Center for Graphene Electronics Electronics and Telecommunications Research Institute (ETRI) 218 Gajeong-ro, Yuseong-gu, Daejeon 305–700, Korea E-mail: [email protected]; [email protected] H. Choi, Dr. C.-G. Choi Department of Advanced Device Technology University of Science and Technology (UST) Daejeon 305–333, Korea J.-W. Shin, Dr. J.-I. Lee OLED Research Center, Electronics and Telecommunications Research Institute (ETRI) 218 Gajeong-ro, Yuseong-gu, Daejeon 305–700, Korea
been used in gas sensor applications.[1–20] Molecular sensors using graphene and rGO have demonstrated a substantial enhancement of their sensitivity, a reduced 1/f noise level, and selectivity of the target gas molecules.[1–20] Recently, similar gas sensor applications have been extended to other
Prof. K.-C. Kim Department of Advanced Chemical Engineering Mokwon University Daejeon 302–729, Korea Prof. S.-Y. Choi Department of Electrical and Engineering and Graphene Research Center Korea Advanced Institute of Science and Technology (KAIST) Daejeon 305–701, Korea Prof. P. Kim Department of Physics Columbia University New York, NY 10027, USA
DOI: 10.1002/smll.201400434 small 2014, DOI: 10.1002/smll.201400434
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
full papers
H. Choi et al.
two-dimensional (2D) materials, such as molybdenum disulfide (MoS2).[21–25] For these two-dimensional systems relying on the adsorption of gas molecules on the field effect transistors channel, recovering the sensors’ responsibility required the complete desorption of the adsorbed molecules, which is usually achieved by heating the device beyond the desorption temperature. For flexible and transparent gas sensing applications using these 2D materials, however, this heating process for the recovery step is often accompanied with a technical challenge due to the limited performance of the microfabricated heaters in the gas sensory applications.[11–14] Moreover, a conventional sensor with this type of micro-sized heater could not be supported for large sized and low-cost, design-free applications employing 2D materials. Although there are previous reports as the graphene application for flexible and transparent heaters which exceeds to temperature of 150 °C,[26–29] the optimization of the both graphene sensor and heater combination has been required for high performance molecules sensor. In this work, we demonstrate the centimeter scale transparent gas sensors with either laterally or vertically built-in graphene heaters on a flexible substrate. This integration of a sensor and heater device
consisting of graphene exhibits high sensitivity, high optical transmittance, high bending strain resistance and a rapid recovery time.
2. Results and Discussion The graphene in this experiment was synthesized by chemical vapor deposition (CVD),[30–32] and a polyethersulfone (PES) substrate was employed as a transparent and flexible supporting layer (see Experimental Section). In our initial design, indeed the graphene sensing channel also served as a heater. However, since the typical recovery process requires a large heating power, the sensory characteristic of graphene channels often altered significantly after the heating process due to the large resistance variation of the channel. We speculate that the contacts degradation is responsible for such alteration owing to the large current flow during the heating and recovery process. Therefore, we separate the heater and sensory parts of the device as in our current design to achieve a more reliable device operation. Figure 1a illustrates the preparation of the lateral gas molecule sensor-heater
Figure 1. a) Schematic diagram of fabrication for laterally positioned graphene sensor and heater on flexible transparent substrate. b,c) Optical microscope image of b) transparent and c) flexible SLG sensor channel-BLG heater on PES substrate. d) Optical microscope images of laser patterned interface between graphene sensor and heater. Here because of high optical transmittance of SLG and BLG (Tr > 90%), sufficient contrast between laser patterned interface and graphene on PES substrate was not guaranteed. Thus this optical microscope image was measured on SiO2 substrate.
2 www.small-journal.com
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
Figure 2. a) The represented Raman spectroscopy of SLG, BLG, and FLG channels. b) The represented statistical analysis of intensity ratio between Raman D and G peak (ID/IG) of SLG from spatial distribution images of ID/IG in inset. The scale bar in inset is 20 µm. Here, because sufficient Raman scattering signal of graphene on PES substrate was not guaranteed due to high optical transmittance of PES substrate, these Raman spectra were measured on SiO2 substrate. c) The relative resistance variation ΔR/R0 of SLG (green dots), BLG (blue dots), and FLG (red dots) channels with 6 mm width and 20 mm length as a function of time under NO2 gas (40 ppm). Here black lines are fitted line for ΔR/R0 and large variation of ΔR/R0 of SLG is −51%.
association. Upon the transfer of graphene onto the PES substrate, single-layer graphene (SLG) for the channel and bilayer graphene (BLG) for the heater were fabricated at the center and sides of the PES substrate, respectively as shown in Figures 1b and c. The graphene channel and heaters were contacted with Cr/Au (≈10 nm/40 nm thickness) electrodes. Here, the optical transmittance (Tr) values of the SLG and BLG are remained to be higher than 90% and they exhibit a similar surface roughness condition (see Supporting Information (SI), Figure S1). In order to rule out undesired connections between the gas sensing channel and the heater, we utilized a laser patterning method (see Experimental section). Figure 1d shows optical images of the disconnected interface between the channel and the heater created by laser lithography. The quality and number of the graphene layers are crucial for gas sensor applications, because the response time and sensitivity of graphene gas sensors have been determined by the binding sites on the channel surface. While the low-energy binding sites consists of the sp2-bonded carbon providing a rapid response, the high-energy binding sites such as vacancies and oxygen functional groups produce a slow sensor response.[2,4] The Raman spectra in Figure 2a exhibit fingerprints reflecting the elevation of the number of graphene layers from SLG to few-layer graphene (FLG) channels with the broadening of the 2D peak and the decrease in the intensity ratio (I2D/IG) between G and 2D.[33] Moreover, the increasing intensity (ID) of Raman D-peak originating from the graphene edge as well as the vacancies and oxygen functional groups on the graphene appears at FLG. In Figure 2b, the spatial distribution between low- and highenergy binding sites on surface of SLG channel reflected in the Raman mapping leads to the statistical analysis results of ID/IG within range of 0.2–0.5. Figure 2c shows the relative resistance variation (ΔR/R0) of SLG, BLG, and FLG sensors as a function of time under an NO2 gas environment (40 ppm), where ΔR and R0 are the resistance variation and initial resistance of the graphene sensor channel, respectively. Although a variation of ΔR/R0 in all graphene sensors emerges due to the adsorbed molecules, different slopes of ΔR/R0 under each graphene condition are observed. Upon small 2014, DOI: 10.1002/smll.201400434
the fitting of ΔR/R0 as a function of time during exposure to NO2 gas by exponential decay, and the response time constant (τ) can be determined from the time evolution of ΔR/R0. Here, we extracted τ from the exponential equation as ΔR/R0 (t) = ΔR/R0 (t0) exp[–(t – t0)/τ], where t0 is the time of changing the curvature of ΔR/R0, the negative sign indicates the exponential decay for NO2 gas, respectively and depending on the molecules binding site, the time constant can be distinguished as either a fast (τf) or a slow (τs) response ascribing to the molecules adsorbed at low-energy or high-energy bonding sites, respectively. Since the ΔR/R0 leads to either multi- (τf and τs) or single-variation (τs) as a function of time, we employed multi- or single-exponential equations, respectively. For example, we could observe that τs for BLG and FLG is more dominant than it is in SLG. Although there are the adsorbing molecules at the low-energy binding sites on surface channel, the encapsulated inner layers in BLG and FLG can lead to a sustaining channel conductance. This degrades sensor response, resulting slower sensing responses for BLG and FLG are mainly exhibited with τs values of 60 s at a ΔR/R0 ratio of −32% and 104 s with −31%, respectively. On the other hand, SLG shows a rapid sensing response with a τf value of 14 s up to an ΔR/R0 ratio of −20% at the initiatory stage and with the ΔR/R0 ratio exceeding −51% with a slow response with a τs of 95 s. Furthermore, the low-ratio of structural disorders (ID/IG < 0.1), that is, the high-ratio of the low-energy binding sites on graphene channel allows us to improve sensor sensitivity and response speed. (see SI, Figure S2) Consequently, the SLG channel with minor structural disorders led to an enhancement of the sensor sensitivity and response. At room temperature, the full recovery time of molecules adsorbed onto 2 cm × 2 cm sized graphene in air is observed to be longer than around several thousand seconds, as shown in Figure 3a. In order to reduce such a long recovery time of the sensors, we employed a graphene heating layer for fast and reproducible sensing performance. Figure 3b shows the recovery time constant (τr) as derived by exponential fitting (≈1 – exp(–t/τr)) of the recovery curves after applying different heating temperatures (inset of Figure 3b). When the graphene heater is activated to increase the channel
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
3
full papers
H. Choi et al.
Figure 3. a) The relative resistance variation ΔR/R0 of SLG channel as a function of time under NO2 gas (40 ppm) including recovery step without heating. Here yellow area in 1800–5000 s indicates recovery steps with turning NO2 gas off. b) Recovering time constant τr as a function of heater temperature. Inset : the recovering curves of the ΔR/R0 as a function of time under different temperature from room temperature to 250 °C. c) Temperature distribution along transverse (x-axis) and longitudinal (y-axis) direction of sensor-heater device structured as laterally intercalated SLG sensor channel (6 mm width) between BLG heaters (7 mm width) with applied 1.7 W of electric power. Here the red dot and blue dot are temperature profiles of thermal image in inset along x-axis and y-axis with origin at center on channel, respectively. Inset : Spatial temperature distribution of graphene heaters (7 mm width) which intercalate 6 mm width graphene sensor with applied 1.7 W. Here three broken squares indicate center channel and side heaters area, respectively. The scale bar is 7 mm. d) The relative resistance variation ΔR/R0 leading to saturation on −39% of SLG channels as a function of time under NO2 gas (40 ppm) including recovery step with Ton ≈ 165 °C heating. Inset : the recovering curves of ΔR/R0 of SLG (black dots), BLG (blue dots) and FLG (red dots) channels under NO2 gas (40 ppm) with Ton as a function of time.
temperature T, τr starts to decrease rapidly. For T < 100 °C, τr > 100 s, but for T < 250 °C, τr ≈ 11 s, two orders of magnitude shorter than that of room temperature. Here, in order to investigate the direct thermal influence from the heater to the channel, we utilized a vertically stacked SLG channel and a FLG heater junction (2 cm × 2 cm size) as isolated by a PES supporting layer (see SI, Figure S3). With the lateral sensor channel and heater combination on the PES substrate, we employed two BLG heaters positioned beside the SLG sensor channel to reduce the full recovery time, as shown at Figure 1. Here, because the heating temperature (T = 100 °C) and the electric power (0.6 W) of SLG are an insufficient, we employ BLG for stiffening heater performance (see SI, Figure S4). Although the BLG in this work leads to a sheet resistance range of ≈500 Ω sq–1, larger than the stacked FLG heater with a doping treatment as reported previously,[26–29] the highest temperature by the BLG heater reached about 250 °C under an applied bias voltage VDS = 30 V corresponding to the power consumption 1.7 W. With this maximum high temperature of the sensor recovery operation, we achieve 27 s of the rising time and 55 s of the decay times under turn-on and turn-off conditions, respectively (see SI, Figure S4). Because thermal energy is provided along a lateral path from the heater to the sensor, the effective temperature
4 www.small-journal.com
at the sensor is lower than that at the heater. In this experiment, we set the limit, T = 250 °C empirically considering the thermal limitation for sustaining the PES substrate without melting. Therefore, in order to increase the temperature on the channel by laterally positioned two heaters with an upper-bound temperature of about 250 °C, the width of the channel should be decreased to narrow the gap between the heaters. In a simulation of the thermal propagation between the center channel and the side heaters by the heat diffusion equation of κ∇2T = 0 to describe the steady-state temperature profile, we found that a channel width of 6 mm leads to a sufficient temperature of 155 °C on a channel as supplied by heaters at 250 °C, where κ and T are the thermal conductivity and the temperature, respectively (see SI, Figure S5). Indeed, decreasing the distance between the heaters as well as channel width increased the temperature, but this modification increase the resistance of the graphene channel. Considering these two competing factors, we optimized the device design to produce the channel resistance ≤ 2 kΩ with the channel width ≈6 mm. Furthermore, the ΔR/R0 of the channels which are wider than 4 mm could be converged on sufficient sensitivity higher than −20%. (see SI, Figure S6) Figure 3c shows a thermal image and temperature line profiles along the transverse (x-axis) and longitudinal (y-axis) directions
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
Figure 4. a) The relative resistance variation ΔR/R0 of SLG channels as a function of time including recovery step with 100–165 °C heating under different NO2 gas concentration from 40 to 0.5 ppm. b) The ΔR/R0 of SLG channels as a function of time including recovery step by 100–165 °C heating under NO2 (1 ppm) with 0 to 1.4% bending strain. Inset : Schematic diagram of strained graphene sensor in NO2 gas condition. Here arrows exhibit uniaxial tensile strain direction (y-axis). c) The ΔR/R0 of sensor at initial, saturated and fully recovered conditions denoted as 1) off, 2) on, and 3) reset at (b), respectively, under 1 ppm NO2 gas as a function of strain from 0 to 1.4%. Inset: Optical image of bended graphene gas sensor-heater junction under 1.2% bending strain.
of the laterally aligned sensor-heater structure with a 6-mm width for the SLG channel and a 7-mm width for the BLG heaters with an applied electric power of 1.7 W. Because the supplied thermal energy from the heaters is mainly focused on the center of the channel and given that there is some degradation of the temperature toward the edges, the temperature profiles along the x- and y-axes with the origin at the center of the channel exhibits a temperature deviation between 100 °C to 165 °C under the supplied heater temperature of 250 °C. Thus, this heating system results in a temperature sufficiently higher than 100 °C for the rapid recovery step. Consequently, as shown in Figure 3d, this sensor-heater device can complete the full recovery step from ΔR/R0 of −39% within 20 s for a 2 cm × 0.6 cm sensor channel in both NO2 gas environment (40 ppm) and the recovery curves of BLG and FLG also exhibit similar scale as SLG (see inset of Figure 3d). Furthermore, the initially hole-doped CVD graphene channels by undesired residues on surface lead to higher sensing sensitivity for the electron acceptor molecules than donor molecules (see SI, Figure S7).[11,12] Utilizing this lateral graphene sensor-heater combined device, the sensitivity variation under different NO2 gas concentrations (40–0.5 ppm) is measured, as shown in Figure 4a. A feasible reduction of ΔR/R0 from −40 to −12% is observed while decreasing the NO2 gas concentration from 40 to 1 ppm, and the high sensitivity of our sensors is guaranteed even at the low concentration of a ΔR/R0 ratio of −10% for 0.5 ppm. Moreover, each sensing cycle under a different NO2 concentration include a full recovery step, and we are able to measure the gas sensing result rapidly without signal degradation. Turning our attention to the verification of the flexibility of our graphene sensor, we also investigate the sensing performance under different bending conditions. With an applied uniaxial tensile strain along the y-axis, as shown in the inset of Figure 4b, gas sensing and recovery cycles are measured in Figure 4b. Although there is slight variation in the decay curves as the bending strain increased from 0 to 1.4%, the sensing curvature is not distorted. Note that the bending strain (ε) was extracted from the bending geometry equation as ε = (δp + δg)/2Rc × (1 + 2η + χη2)/(1 + χη + χη + χη2), where η = δg/δp, χ = Yg/Yp, and Rc is bending radius, and small 2014, DOI: 10.1002/smll.201400434
δp, δg, and Yp, Yg are thickness and Young’s modulus for PES substrate and graphene, respectively.[26,34,35] In Figure 4c, we display ΔR/R0 ratio as a function of strain at three different conditions corresponding the (1) off, (2) on, and (3) reset in Figure 4b. We find that ΔR/R0 exhibit a deviation within ≈5% under a strain level of 1.4%. Repeated straining between flat and 1.0% bending, produces a device performance change less than 10%, presumably due to the residual static charging on channel surface (see SI, Figure S8). Furthermore, we also observe reproducible results from a vertically stacked SLG channel and a FLG heater junction device (see SI, Figure S3); hence, this flexible and transparent graphene sensor-heater device can be extended to variously scaled applications with unlimited sensor designs.
3. Conclusion We demonstrate a flexible and transparent graphene gas molecule sensor integrated with a graphene heater. Laterally positioned sensor and heaters on a PES substrate with transparency (Tr > 90%) are fabricated to exhibit high-performance gas molecules sensing with a ΔR/R0 ratio of 10% with 0.5 ppm NO2 gas and a rapid response and recovery time smaller than 20 s for a centimeter-scale sensor channel. The realization of large scale transparent gas sensor on a flexible substrate is a novel combination that has never been realized as an integrated system. Upon comparing with conventional solid-state oxide gas sensor (SnO2) which operates with response time of 50–90 s and recovery time of 26 s to 8 min at 250 °C,[36,37] SLG channel-BLG heater junction sensor exhibits significantly fast-response and recovery performance. In addition, the flexibility of the graphene sensor-heater combined devices are guaranteed with a gas sensing sensitivity ratio ΔR/R0 of 12% for 1 ppm NO2 gas under a bending strain rate of 1.4%. This flexible and transparent all graphene sensor and heater scheme presents the possibility for an unconstrained size and design of sensor applications such as smart window and wearable nose-sensors application with transparent flexible 2D materials.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
5
full papers
H. Choi et al.
4. Experimental Section Graphene Growth by Chemical Vapor Deposition and Transferring on PES Substrate: Single layer graphene (SLG) film was grown by chemical vapor deposition (CVD) on 25 µm thick Cu foil (99.999% Cu foil from Alfa Aesar). After pre-annealing of Cu foil for 20 min, graphene was grown at temperature of 1000 °C with a gas mixture of CH4 (30 sccm) and H2 (10 sccm) under ambient pressure (3.9 × 10−1 Torr) for 20 min. After graphene growth, rapid cooling step (cooling rate: ≈32 °C min–1) is followed under Ar (100 sccm) gas environment. In order to transfer graphene on 200 µm thick polyethersulfone (PES) substrate, graphene on Cu foil was spincoated with Poly(methyl methacrylate) (PMMA) supporting layer. Then the backside Cu foil was etched by 0.1 M ammonium persulfate solution. Finally the rinsed “PMMA/graphene” films were transferred on selected area of PES substrate and then PMMA was removed by acetone. Employing this transfer process, we prepared the BLG on a PES substrate by layer-by-layer stacking of SLGs. The few-layer graphene (FLG) employed for channel and heater layers was grown on Ni catalyst by CVD, similar to process of SLG on Cu foil, and thickness of these FLG (5–10 layers) were confirmed by tunneling electron microscopy. Graphene Sensor Device Fabrication and Characterizations: The graphene channel and heaters were contacted with Cr/Au (≈10 nm/40 nm thickness) electrodes. The spatial Raman scattering distribution excited by 532 nm laser was employed to inspect layer number and quality of graphene. In order to disconnect between graphene sensor and heater area, confocal focusing and scanning system with infrared laser (1064 nm of wavelength, 2 ns of pulse width, 50 kHz of pulse repetition rate and 0.3 W of power) leading to 30 µm resolution were employed. Chamber which is capable of enclosing graphene gas sensor with controllable NO2 (NH3) gas concentration from 40 to 0.5 ppm by mass flow controller (MFC) was utilized for measurement of gas sensing sensitivity. Gas molecules sensing properties reflecting the resistance variation of graphene channel were measured with 2-probe geometry by standard lock-in techniques. The thermal images and temperature of graphene heaters operated by electrical power supply were measured with infrared camera system (Fluke Thermal imager, Ti100) and the k-type thermocouple was employed for the cross-checked heater temperature.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements H.C. and J.S.C. contributed equally to this work. This work was supported by the Creative Research Program of the ETRI (14ZE1110), Korea and a grant (2011–0031660) from the Center for Advanced Soft Electronics under the Global Frontier Research Program of the Ministry of Science, ICT & Future Planning, Korea. S.Y.C. acknowledges the research grants from Nano-Material Technology
6 www.small-journal.com
Development Program (2012M3A7B4049807) and Global Frontier Research Center for Advanced Soft Electronics (2011-0031640). P.K. acknowledges Nano-Material Technology Development Program (2012M3A7B4049966).
[1] F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson, K. S. Novoselov, Nat. Mater. 2007, 6, 652–655. [2] Y. Dan, Y. Lu, N. J. Kybert, Z. Luo, A. T. C. Johnson, Nano Lett. 2009, 9, 1472–1475. [3] S. Rumyantsev, G. Liu, M. S. Shur, R. A. Potyrailo, A. A. Balandin, Nano Lett. 2012, 12, 2294–2298. [4] B. Kumar, K. Min, M. Bashirzadeh, A. B. Farimani, M.-H. Bae, D. Estrada, Y. D. Kim, P. Yasaei, Y. D. Park, E. Pop, N. R. Aluru, A. Salehi-Khojin, Nano Lett. 2013, 13, 1962–1968. [5] J. T. Robinson, F. K. Perkins, E. C. Snow, Z. Wei, P. E. Sheehan, Nano Lett. 2008, 8, 3137–3140. [6] J. Kong, N. R. Franklin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, H. Dai, Science 2000, 278, 622–625. [7] Y. Guo, B. Wu, H. Liu, Y. Ma, Y. Yang, J. Zheng, G. Yu, Y. Liu, Adv. Mater. 2011, 23, 4626–4630. [8] W. Li, X. Geng, Y. Guo, J. Rong, Y. Gong, L. Wu, X. Zhang, P. Li, J. Xu, G. Cheng, M. Sun, L. Liu, ACS Nano 2011, 5, 6955–6961. [9] W. Yuan, A. Liu, L. Huang, C. Li, G. Shi, Adv. Mater. 2013, 25, 766–771. [10] M. G. Chung, D.-H. Kim, D. K. Seo, T. Kim, H. U. Im, H. M. Lee, J.-B. Yoo, S.-H. Hong, T. J. Kang, Y. H. Kim, Sens. Actuators B: Chem. 2012, 169, 387–392. [11] F. Yavari, Z. Chen, A. V. Thomas, W. Ren, H.-M. Cheng, N. Koratkar, Sci. Rep. 2011, 1, 166. [12] J. D. Fowler, M. J. Allen, V. C. Tung, Y. Yang, R. B. Kaner, B. H. Weiller, ACS Nano 2009, 3, 301. [13] H. Y. Jeong, D.-S. Lee, H. K. Choi, D. H. Lee, J.-E. Kim, J. Y. Lee, W. J. Lee, S. O. Kim, S.-Y. Choi, Appl. Phys. Lett. 2010, 96, 213105. [14] S. Some, Y. Xu, Y. Kim, Y. Yoon, H. Qin, A. Kulkarni, T. Kim, H. Lee, Sci. Rep. 2013, 3, 1868. [15] A. A. Balandin, Nat. Nanotechnol. 2013, 10, 549–555. [16] X. Huang, X. Qi, F. Boey, H. Zhang, Chem. Soc. Rev. 2012, 41, 666–686. [17] X. Huang, Z. Yin, S. Wu, X. Qi, Q. He, Q. Zhang, Q. Yan, F. Boey, H. Zhang, Small 2011, 7, 1876–1902. [18] Q. He, S. Wu, Z. Yin, H. Zhang, Chem. Sci. 2012, 3, 1764– 1772. [19] H. Li, J. Wu, X. Qi, Q. He, C. Liusman, G. Lu, X. Zhou, H. Zhang, Small 2013, 9, 382–386. [20] J. Wu, H. Li, X. Qi, Q. He, B. Xu, H. Zhang, Small 2014, DOI: 10.1002/smll.201303637. [21] H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu, D. W. H. Fam, A. L. Y. Tok, Q. Zhang, H. Zhang, Small 2012, 8, 63–67. [22] Q. He, Z. Zeng, Z. Yin, H. Li, S. Wu, X. Huang, H. Zhang, Small 2012, 8, 2994–2999. [23] J. D. Late, Y.-K. Huang, B. Liu, J. Acharya, S. N. Shirodkar, J. Luo, A. Yan, D. Charles, U. V. Waghmare, V. P. Dravid, C. N. R. Rao, ACS Nano 2013, 7, 4879–4891. [24] F. K. Perkins, A. L. Friedman, E. Cobas, P. M. Campbell, G. G. Jernigan, B. T. Jonker, Nano Lett. 2013, 13, 668–673. [25] X. Huang, Z. Zeng, H. Zhang, Chem. Soc. Rev. 2013, 42, 1934–1946. [26] J. Kang, H. Kim, K. S. Kim, S.-K. Lee, S. Bae, J.-H Ahn, Y.-J. Kim, J.-B. Choi, B. H. Hong, Nano Lett. 2011, 11, 5154–5158. [27] D. Sui, Y. Huang, L. Huang, J. Liang, Y. Ma, Y. Chen, Small 2011, 7, 3186–3192. [28] J. J. Bae, S. C. Lim, G. H. Han, Y. W. Jo, D. L. Doung, E. S. Kim, S. J. Chae, T. Q. Huy, N. V. Luan, Y. H. Lee, Adv. Funct. Mater. 2012, 22, 4819–4826.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
[29] H. Kim, Y. Kim, K. S. Kim, H. Y. Jeong, A. R. Jang, S. H. Han, D. H. Yoon, K. S. Suh, H. S. Shin, T. Y. Kim, W. S. Yang, ACS Nano 2013, 7, 5769–5776. [30] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H Ahn, P. Kim, J.-Y. Choi, B. H. Hong, Nature 2009, 457, 706–710. [31] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Velamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo, R. S. Ruoff, Science 2009, 324, 1312–1314. [32] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H. R. Kim, Y. I. Song, Y.-J. Kim, K. S. Kim, B. Ozyilmaz, J.-H. Ahn, B. H. Hong, S. Iijima, Nat. Nanotechnol. 2010, 5, 574–578.
small 2014, DOI: 10.1002/smll.201400434
[33] A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, S. Roth, A. K. Geim, Phys. Rev. Lett. 2006, 97, 187401. [34] G. M. Scarpello, D. Ritelli, Meccanica 2006, 41, 519–527. [35] N. Petrone, I. Meric, J. Hone, K. L. Shepard, Nano Lett. 2013, 13, 121–125. [36] G. Sberveglieri, S. Groppelli, P. Nelli, Sens. Actuators B: Chem. 1991, 4, 457–461. [37] Y. Liu, E. Koep, M. A. Liu, Chem. Mater. 2005, 17, 3997–4000.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: February 17, 2014 Revised: April 1, 2014 Published online:
www.small-journal.com
7
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers Hongkyw Choi, Jin Sik Choi, Jin-Soo Kim, Jong-Ho Choe, Kwang Hyo Chung, Jin-Wook Shin, Jin Tae Kim, Doo-Hyeb Youn, Ki-Chul Kim, Jeong-Ik Lee, Sung-Yool Choi, Philip Kim, Choon-Gi Choi,* and Young-Jun Yu*
Graphene
leading to high surface-to-volume ratio and outstanding conductivity is applied for gas molecule sensing with fully utilizing its unique transparent and flexible functionalities which cannot be expected from solid-state gas sensors. In order to attain a fast response and rapid recovering time, the flexible sensors also require integrated flexible and transparent heaters. Here, large-scale flexible and transparent gas molecule sensor devices, integrated with a graphene sensing channel and a graphene transparent heater for fast recovering operation, are demonstrated. This combined all-graphene device structure enables an overall device optical transmittance that exceeds 90% and reliable sensing performance with a bending strain of less than 1.4%. In particular, it is possible to classify the fast (≈14 s) and slow (≈95 s) response due to sp2-carbon bonding and disorders on graphene and the self-integrated graphene heater leads to the rapid recovery (≈11 s) of a 2 cm × 2 cm sized sensor with reproducible sensing cycles, including full recovery steps without significant signal degradation under exposure to NO2 gas.
1. Introduction Utilizing the electrical response to adsorbed molecules on sp2-bonded carbon networks, field effect transistors (FETs) based on graphene and reduced graphene oxide (rGO) have
H. Choi, Dr. J. S. Choi, Dr. J.-S. Kim, Dr. J.-H. Choe, Dr. K. H. Chung, Dr. J. T. Kim, Dr. D.-H. Youn, Prof. K.-C. Kim, Dr. C.-G. Choi, Dr. Y.-J. Yu Creative Research Center for Graphene Electronics Electronics and Telecommunications Research Institute (ETRI) 218 Gajeong-ro, Yuseong-gu, Daejeon 305–700, Korea E-mail: [email protected]; [email protected] H. Choi, Dr. C.-G. Choi Department of Advanced Device Technology University of Science and Technology (UST) Daejeon 305–333, Korea J.-W. Shin, Dr. J.-I. Lee OLED Research Center, Electronics and Telecommunications Research Institute (ETRI) 218 Gajeong-ro, Yuseong-gu, Daejeon 305–700, Korea
been used in gas sensor applications.[1–20] Molecular sensors using graphene and rGO have demonstrated a substantial enhancement of their sensitivity, a reduced 1/f noise level, and selectivity of the target gas molecules.[1–20] Recently, similar gas sensor applications have been extended to other
Prof. K.-C. Kim Department of Advanced Chemical Engineering Mokwon University Daejeon 302–729, Korea Prof. S.-Y. Choi Department of Electrical and Engineering and Graphene Research Center Korea Advanced Institute of Science and Technology (KAIST) Daejeon 305–701, Korea Prof. P. Kim Department of Physics Columbia University New York, NY 10027, USA
DOI: 10.1002/smll.201400434 small 2014, DOI: 10.1002/smll.201400434
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
full papers
H. Choi et al.
two-dimensional (2D) materials, such as molybdenum disulfide (MoS2).[21–25] For these two-dimensional systems relying on the adsorption of gas molecules on the field effect transistors channel, recovering the sensors’ responsibility required the complete desorption of the adsorbed molecules, which is usually achieved by heating the device beyond the desorption temperature. For flexible and transparent gas sensing applications using these 2D materials, however, this heating process for the recovery step is often accompanied with a technical challenge due to the limited performance of the microfabricated heaters in the gas sensory applications.[11–14] Moreover, a conventional sensor with this type of micro-sized heater could not be supported for large sized and low-cost, design-free applications employing 2D materials. Although there are previous reports as the graphene application for flexible and transparent heaters which exceeds to temperature of 150 °C,[26–29] the optimization of the both graphene sensor and heater combination has been required for high performance molecules sensor. In this work, we demonstrate the centimeter scale transparent gas sensors with either laterally or vertically built-in graphene heaters on a flexible substrate. This integration of a sensor and heater device
consisting of graphene exhibits high sensitivity, high optical transmittance, high bending strain resistance and a rapid recovery time.
2. Results and Discussion The graphene in this experiment was synthesized by chemical vapor deposition (CVD),[30–32] and a polyethersulfone (PES) substrate was employed as a transparent and flexible supporting layer (see Experimental Section). In our initial design, indeed the graphene sensing channel also served as a heater. However, since the typical recovery process requires a large heating power, the sensory characteristic of graphene channels often altered significantly after the heating process due to the large resistance variation of the channel. We speculate that the contacts degradation is responsible for such alteration owing to the large current flow during the heating and recovery process. Therefore, we separate the heater and sensory parts of the device as in our current design to achieve a more reliable device operation. Figure 1a illustrates the preparation of the lateral gas molecule sensor-heater
Figure 1. a) Schematic diagram of fabrication for laterally positioned graphene sensor and heater on flexible transparent substrate. b,c) Optical microscope image of b) transparent and c) flexible SLG sensor channel-BLG heater on PES substrate. d) Optical microscope images of laser patterned interface between graphene sensor and heater. Here because of high optical transmittance of SLG and BLG (Tr > 90%), sufficient contrast between laser patterned interface and graphene on PES substrate was not guaranteed. Thus this optical microscope image was measured on SiO2 substrate.
2 www.small-journal.com
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
Figure 2. a) The represented Raman spectroscopy of SLG, BLG, and FLG channels. b) The represented statistical analysis of intensity ratio between Raman D and G peak (ID/IG) of SLG from spatial distribution images of ID/IG in inset. The scale bar in inset is 20 µm. Here, because sufficient Raman scattering signal of graphene on PES substrate was not guaranteed due to high optical transmittance of PES substrate, these Raman spectra were measured on SiO2 substrate. c) The relative resistance variation ΔR/R0 of SLG (green dots), BLG (blue dots), and FLG (red dots) channels with 6 mm width and 20 mm length as a function of time under NO2 gas (40 ppm). Here black lines are fitted line for ΔR/R0 and large variation of ΔR/R0 of SLG is −51%.
association. Upon the transfer of graphene onto the PES substrate, single-layer graphene (SLG) for the channel and bilayer graphene (BLG) for the heater were fabricated at the center and sides of the PES substrate, respectively as shown in Figures 1b and c. The graphene channel and heaters were contacted with Cr/Au (≈10 nm/40 nm thickness) electrodes. Here, the optical transmittance (Tr) values of the SLG and BLG are remained to be higher than 90% and they exhibit a similar surface roughness condition (see Supporting Information (SI), Figure S1). In order to rule out undesired connections between the gas sensing channel and the heater, we utilized a laser patterning method (see Experimental section). Figure 1d shows optical images of the disconnected interface between the channel and the heater created by laser lithography. The quality and number of the graphene layers are crucial for gas sensor applications, because the response time and sensitivity of graphene gas sensors have been determined by the binding sites on the channel surface. While the low-energy binding sites consists of the sp2-bonded carbon providing a rapid response, the high-energy binding sites such as vacancies and oxygen functional groups produce a slow sensor response.[2,4] The Raman spectra in Figure 2a exhibit fingerprints reflecting the elevation of the number of graphene layers from SLG to few-layer graphene (FLG) channels with the broadening of the 2D peak and the decrease in the intensity ratio (I2D/IG) between G and 2D.[33] Moreover, the increasing intensity (ID) of Raman D-peak originating from the graphene edge as well as the vacancies and oxygen functional groups on the graphene appears at FLG. In Figure 2b, the spatial distribution between low- and highenergy binding sites on surface of SLG channel reflected in the Raman mapping leads to the statistical analysis results of ID/IG within range of 0.2–0.5. Figure 2c shows the relative resistance variation (ΔR/R0) of SLG, BLG, and FLG sensors as a function of time under an NO2 gas environment (40 ppm), where ΔR and R0 are the resistance variation and initial resistance of the graphene sensor channel, respectively. Although a variation of ΔR/R0 in all graphene sensors emerges due to the adsorbed molecules, different slopes of ΔR/R0 under each graphene condition are observed. Upon small 2014, DOI: 10.1002/smll.201400434
the fitting of ΔR/R0 as a function of time during exposure to NO2 gas by exponential decay, and the response time constant (τ) can be determined from the time evolution of ΔR/R0. Here, we extracted τ from the exponential equation as ΔR/R0 (t) = ΔR/R0 (t0) exp[–(t – t0)/τ], where t0 is the time of changing the curvature of ΔR/R0, the negative sign indicates the exponential decay for NO2 gas, respectively and depending on the molecules binding site, the time constant can be distinguished as either a fast (τf) or a slow (τs) response ascribing to the molecules adsorbed at low-energy or high-energy bonding sites, respectively. Since the ΔR/R0 leads to either multi- (τf and τs) or single-variation (τs) as a function of time, we employed multi- or single-exponential equations, respectively. For example, we could observe that τs for BLG and FLG is more dominant than it is in SLG. Although there are the adsorbing molecules at the low-energy binding sites on surface channel, the encapsulated inner layers in BLG and FLG can lead to a sustaining channel conductance. This degrades sensor response, resulting slower sensing responses for BLG and FLG are mainly exhibited with τs values of 60 s at a ΔR/R0 ratio of −32% and 104 s with −31%, respectively. On the other hand, SLG shows a rapid sensing response with a τf value of 14 s up to an ΔR/R0 ratio of −20% at the initiatory stage and with the ΔR/R0 ratio exceeding −51% with a slow response with a τs of 95 s. Furthermore, the low-ratio of structural disorders (ID/IG < 0.1), that is, the high-ratio of the low-energy binding sites on graphene channel allows us to improve sensor sensitivity and response speed. (see SI, Figure S2) Consequently, the SLG channel with minor structural disorders led to an enhancement of the sensor sensitivity and response. At room temperature, the full recovery time of molecules adsorbed onto 2 cm × 2 cm sized graphene in air is observed to be longer than around several thousand seconds, as shown in Figure 3a. In order to reduce such a long recovery time of the sensors, we employed a graphene heating layer for fast and reproducible sensing performance. Figure 3b shows the recovery time constant (τr) as derived by exponential fitting (≈1 – exp(–t/τr)) of the recovery curves after applying different heating temperatures (inset of Figure 3b). When the graphene heater is activated to increase the channel
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
3
full papers
H. Choi et al.
Figure 3. a) The relative resistance variation ΔR/R0 of SLG channel as a function of time under NO2 gas (40 ppm) including recovery step without heating. Here yellow area in 1800–5000 s indicates recovery steps with turning NO2 gas off. b) Recovering time constant τr as a function of heater temperature. Inset : the recovering curves of the ΔR/R0 as a function of time under different temperature from room temperature to 250 °C. c) Temperature distribution along transverse (x-axis) and longitudinal (y-axis) direction of sensor-heater device structured as laterally intercalated SLG sensor channel (6 mm width) between BLG heaters (7 mm width) with applied 1.7 W of electric power. Here the red dot and blue dot are temperature profiles of thermal image in inset along x-axis and y-axis with origin at center on channel, respectively. Inset : Spatial temperature distribution of graphene heaters (7 mm width) which intercalate 6 mm width graphene sensor with applied 1.7 W. Here three broken squares indicate center channel and side heaters area, respectively. The scale bar is 7 mm. d) The relative resistance variation ΔR/R0 leading to saturation on −39% of SLG channels as a function of time under NO2 gas (40 ppm) including recovery step with Ton ≈ 165 °C heating. Inset : the recovering curves of ΔR/R0 of SLG (black dots), BLG (blue dots) and FLG (red dots) channels under NO2 gas (40 ppm) with Ton as a function of time.
temperature T, τr starts to decrease rapidly. For T < 100 °C, τr > 100 s, but for T < 250 °C, τr ≈ 11 s, two orders of magnitude shorter than that of room temperature. Here, in order to investigate the direct thermal influence from the heater to the channel, we utilized a vertically stacked SLG channel and a FLG heater junction (2 cm × 2 cm size) as isolated by a PES supporting layer (see SI, Figure S3). With the lateral sensor channel and heater combination on the PES substrate, we employed two BLG heaters positioned beside the SLG sensor channel to reduce the full recovery time, as shown at Figure 1. Here, because the heating temperature (T = 100 °C) and the electric power (0.6 W) of SLG are an insufficient, we employ BLG for stiffening heater performance (see SI, Figure S4). Although the BLG in this work leads to a sheet resistance range of ≈500 Ω sq–1, larger than the stacked FLG heater with a doping treatment as reported previously,[26–29] the highest temperature by the BLG heater reached about 250 °C under an applied bias voltage VDS = 30 V corresponding to the power consumption 1.7 W. With this maximum high temperature of the sensor recovery operation, we achieve 27 s of the rising time and 55 s of the decay times under turn-on and turn-off conditions, respectively (see SI, Figure S4). Because thermal energy is provided along a lateral path from the heater to the sensor, the effective temperature
4 www.small-journal.com
at the sensor is lower than that at the heater. In this experiment, we set the limit, T = 250 °C empirically considering the thermal limitation for sustaining the PES substrate without melting. Therefore, in order to increase the temperature on the channel by laterally positioned two heaters with an upper-bound temperature of about 250 °C, the width of the channel should be decreased to narrow the gap between the heaters. In a simulation of the thermal propagation between the center channel and the side heaters by the heat diffusion equation of κ∇2T = 0 to describe the steady-state temperature profile, we found that a channel width of 6 mm leads to a sufficient temperature of 155 °C on a channel as supplied by heaters at 250 °C, where κ and T are the thermal conductivity and the temperature, respectively (see SI, Figure S5). Indeed, decreasing the distance between the heaters as well as channel width increased the temperature, but this modification increase the resistance of the graphene channel. Considering these two competing factors, we optimized the device design to produce the channel resistance ≤ 2 kΩ with the channel width ≈6 mm. Furthermore, the ΔR/R0 of the channels which are wider than 4 mm could be converged on sufficient sensitivity higher than −20%. (see SI, Figure S6) Figure 3c shows a thermal image and temperature line profiles along the transverse (x-axis) and longitudinal (y-axis) directions
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
Figure 4. a) The relative resistance variation ΔR/R0 of SLG channels as a function of time including recovery step with 100–165 °C heating under different NO2 gas concentration from 40 to 0.5 ppm. b) The ΔR/R0 of SLG channels as a function of time including recovery step by 100–165 °C heating under NO2 (1 ppm) with 0 to 1.4% bending strain. Inset : Schematic diagram of strained graphene sensor in NO2 gas condition. Here arrows exhibit uniaxial tensile strain direction (y-axis). c) The ΔR/R0 of sensor at initial, saturated and fully recovered conditions denoted as 1) off, 2) on, and 3) reset at (b), respectively, under 1 ppm NO2 gas as a function of strain from 0 to 1.4%. Inset: Optical image of bended graphene gas sensor-heater junction under 1.2% bending strain.
of the laterally aligned sensor-heater structure with a 6-mm width for the SLG channel and a 7-mm width for the BLG heaters with an applied electric power of 1.7 W. Because the supplied thermal energy from the heaters is mainly focused on the center of the channel and given that there is some degradation of the temperature toward the edges, the temperature profiles along the x- and y-axes with the origin at the center of the channel exhibits a temperature deviation between 100 °C to 165 °C under the supplied heater temperature of 250 °C. Thus, this heating system results in a temperature sufficiently higher than 100 °C for the rapid recovery step. Consequently, as shown in Figure 3d, this sensor-heater device can complete the full recovery step from ΔR/R0 of −39% within 20 s for a 2 cm × 0.6 cm sensor channel in both NO2 gas environment (40 ppm) and the recovery curves of BLG and FLG also exhibit similar scale as SLG (see inset of Figure 3d). Furthermore, the initially hole-doped CVD graphene channels by undesired residues on surface lead to higher sensing sensitivity for the electron acceptor molecules than donor molecules (see SI, Figure S7).[11,12] Utilizing this lateral graphene sensor-heater combined device, the sensitivity variation under different NO2 gas concentrations (40–0.5 ppm) is measured, as shown in Figure 4a. A feasible reduction of ΔR/R0 from −40 to −12% is observed while decreasing the NO2 gas concentration from 40 to 1 ppm, and the high sensitivity of our sensors is guaranteed even at the low concentration of a ΔR/R0 ratio of −10% for 0.5 ppm. Moreover, each sensing cycle under a different NO2 concentration include a full recovery step, and we are able to measure the gas sensing result rapidly without signal degradation. Turning our attention to the verification of the flexibility of our graphene sensor, we also investigate the sensing performance under different bending conditions. With an applied uniaxial tensile strain along the y-axis, as shown in the inset of Figure 4b, gas sensing and recovery cycles are measured in Figure 4b. Although there is slight variation in the decay curves as the bending strain increased from 0 to 1.4%, the sensing curvature is not distorted. Note that the bending strain (ε) was extracted from the bending geometry equation as ε = (δp + δg)/2Rc × (1 + 2η + χη2)/(1 + χη + χη + χη2), where η = δg/δp, χ = Yg/Yp, and Rc is bending radius, and small 2014, DOI: 10.1002/smll.201400434
δp, δg, and Yp, Yg are thickness and Young’s modulus for PES substrate and graphene, respectively.[26,34,35] In Figure 4c, we display ΔR/R0 ratio as a function of strain at three different conditions corresponding the (1) off, (2) on, and (3) reset in Figure 4b. We find that ΔR/R0 exhibit a deviation within ≈5% under a strain level of 1.4%. Repeated straining between flat and 1.0% bending, produces a device performance change less than 10%, presumably due to the residual static charging on channel surface (see SI, Figure S8). Furthermore, we also observe reproducible results from a vertically stacked SLG channel and a FLG heater junction device (see SI, Figure S3); hence, this flexible and transparent graphene sensor-heater device can be extended to variously scaled applications with unlimited sensor designs.
3. Conclusion We demonstrate a flexible and transparent graphene gas molecule sensor integrated with a graphene heater. Laterally positioned sensor and heaters on a PES substrate with transparency (Tr > 90%) are fabricated to exhibit high-performance gas molecules sensing with a ΔR/R0 ratio of 10% with 0.5 ppm NO2 gas and a rapid response and recovery time smaller than 20 s for a centimeter-scale sensor channel. The realization of large scale transparent gas sensor on a flexible substrate is a novel combination that has never been realized as an integrated system. Upon comparing with conventional solid-state oxide gas sensor (SnO2) which operates with response time of 50–90 s and recovery time of 26 s to 8 min at 250 °C,[36,37] SLG channel-BLG heater junction sensor exhibits significantly fast-response and recovery performance. In addition, the flexibility of the graphene sensor-heater combined devices are guaranteed with a gas sensing sensitivity ratio ΔR/R0 of 12% for 1 ppm NO2 gas under a bending strain rate of 1.4%. This flexible and transparent all graphene sensor and heater scheme presents the possibility for an unconstrained size and design of sensor applications such as smart window and wearable nose-sensors application with transparent flexible 2D materials.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
5
full papers
H. Choi et al.
4. Experimental Section Graphene Growth by Chemical Vapor Deposition and Transferring on PES Substrate: Single layer graphene (SLG) film was grown by chemical vapor deposition (CVD) on 25 µm thick Cu foil (99.999% Cu foil from Alfa Aesar). After pre-annealing of Cu foil for 20 min, graphene was grown at temperature of 1000 °C with a gas mixture of CH4 (30 sccm) and H2 (10 sccm) under ambient pressure (3.9 × 10−1 Torr) for 20 min. After graphene growth, rapid cooling step (cooling rate: ≈32 °C min–1) is followed under Ar (100 sccm) gas environment. In order to transfer graphene on 200 µm thick polyethersulfone (PES) substrate, graphene on Cu foil was spincoated with Poly(methyl methacrylate) (PMMA) supporting layer. Then the backside Cu foil was etched by 0.1 M ammonium persulfate solution. Finally the rinsed “PMMA/graphene” films were transferred on selected area of PES substrate and then PMMA was removed by acetone. Employing this transfer process, we prepared the BLG on a PES substrate by layer-by-layer stacking of SLGs. The few-layer graphene (FLG) employed for channel and heater layers was grown on Ni catalyst by CVD, similar to process of SLG on Cu foil, and thickness of these FLG (5–10 layers) were confirmed by tunneling electron microscopy. Graphene Sensor Device Fabrication and Characterizations: The graphene channel and heaters were contacted with Cr/Au (≈10 nm/40 nm thickness) electrodes. The spatial Raman scattering distribution excited by 532 nm laser was employed to inspect layer number and quality of graphene. In order to disconnect between graphene sensor and heater area, confocal focusing and scanning system with infrared laser (1064 nm of wavelength, 2 ns of pulse width, 50 kHz of pulse repetition rate and 0.3 W of power) leading to 30 µm resolution were employed. Chamber which is capable of enclosing graphene gas sensor with controllable NO2 (NH3) gas concentration from 40 to 0.5 ppm by mass flow controller (MFC) was utilized for measurement of gas sensing sensitivity. Gas molecules sensing properties reflecting the resistance variation of graphene channel were measured with 2-probe geometry by standard lock-in techniques. The thermal images and temperature of graphene heaters operated by electrical power supply were measured with infrared camera system (Fluke Thermal imager, Ti100) and the k-type thermocouple was employed for the cross-checked heater temperature.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements H.C. and J.S.C. contributed equally to this work. This work was supported by the Creative Research Program of the ETRI (14ZE1110), Korea and a grant (2011–0031660) from the Center for Advanced Soft Electronics under the Global Frontier Research Program of the Ministry of Science, ICT & Future Planning, Korea. S.Y.C. acknowledges the research grants from Nano-Material Technology
6 www.small-journal.com
Development Program (2012M3A7B4049807) and Global Frontier Research Center for Advanced Soft Electronics (2011-0031640). P.K. acknowledges Nano-Material Technology Development Program (2012M3A7B4049966).
[1] F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson, K. S. Novoselov, Nat. Mater. 2007, 6, 652–655. [2] Y. Dan, Y. Lu, N. J. Kybert, Z. Luo, A. T. C. Johnson, Nano Lett. 2009, 9, 1472–1475. [3] S. Rumyantsev, G. Liu, M. S. Shur, R. A. Potyrailo, A. A. Balandin, Nano Lett. 2012, 12, 2294–2298. [4] B. Kumar, K. Min, M. Bashirzadeh, A. B. Farimani, M.-H. Bae, D. Estrada, Y. D. Kim, P. Yasaei, Y. D. Park, E. Pop, N. R. Aluru, A. Salehi-Khojin, Nano Lett. 2013, 13, 1962–1968. [5] J. T. Robinson, F. K. Perkins, E. C. Snow, Z. Wei, P. E. Sheehan, Nano Lett. 2008, 8, 3137–3140. [6] J. Kong, N. R. Franklin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, H. Dai, Science 2000, 278, 622–625. [7] Y. Guo, B. Wu, H. Liu, Y. Ma, Y. Yang, J. Zheng, G. Yu, Y. Liu, Adv. Mater. 2011, 23, 4626–4630. [8] W. Li, X. Geng, Y. Guo, J. Rong, Y. Gong, L. Wu, X. Zhang, P. Li, J. Xu, G. Cheng, M. Sun, L. Liu, ACS Nano 2011, 5, 6955–6961. [9] W. Yuan, A. Liu, L. Huang, C. Li, G. Shi, Adv. Mater. 2013, 25, 766–771. [10] M. G. Chung, D.-H. Kim, D. K. Seo, T. Kim, H. U. Im, H. M. Lee, J.-B. Yoo, S.-H. Hong, T. J. Kang, Y. H. Kim, Sens. Actuators B: Chem. 2012, 169, 387–392. [11] F. Yavari, Z. Chen, A. V. Thomas, W. Ren, H.-M. Cheng, N. Koratkar, Sci. Rep. 2011, 1, 166. [12] J. D. Fowler, M. J. Allen, V. C. Tung, Y. Yang, R. B. Kaner, B. H. Weiller, ACS Nano 2009, 3, 301. [13] H. Y. Jeong, D.-S. Lee, H. K. Choi, D. H. Lee, J.-E. Kim, J. Y. Lee, W. J. Lee, S. O. Kim, S.-Y. Choi, Appl. Phys. Lett. 2010, 96, 213105. [14] S. Some, Y. Xu, Y. Kim, Y. Yoon, H. Qin, A. Kulkarni, T. Kim, H. Lee, Sci. Rep. 2013, 3, 1868. [15] A. A. Balandin, Nat. Nanotechnol. 2013, 10, 549–555. [16] X. Huang, X. Qi, F. Boey, H. Zhang, Chem. Soc. Rev. 2012, 41, 666–686. [17] X. Huang, Z. Yin, S. Wu, X. Qi, Q. He, Q. Zhang, Q. Yan, F. Boey, H. Zhang, Small 2011, 7, 1876–1902. [18] Q. He, S. Wu, Z. Yin, H. Zhang, Chem. Sci. 2012, 3, 1764– 1772. [19] H. Li, J. Wu, X. Qi, Q. He, C. Liusman, G. Lu, X. Zhou, H. Zhang, Small 2013, 9, 382–386. [20] J. Wu, H. Li, X. Qi, Q. He, B. Xu, H. Zhang, Small 2014, DOI: 10.1002/smll.201303637. [21] H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu, D. W. H. Fam, A. L. Y. Tok, Q. Zhang, H. Zhang, Small 2012, 8, 63–67. [22] Q. He, Z. Zeng, Z. Yin, H. Li, S. Wu, X. Huang, H. Zhang, Small 2012, 8, 2994–2999. [23] J. D. Late, Y.-K. Huang, B. Liu, J. Acharya, S. N. Shirodkar, J. Luo, A. Yan, D. Charles, U. V. Waghmare, V. P. Dravid, C. N. R. Rao, ACS Nano 2013, 7, 4879–4891. [24] F. K. Perkins, A. L. Friedman, E. Cobas, P. M. Campbell, G. G. Jernigan, B. T. Jonker, Nano Lett. 2013, 13, 668–673. [25] X. Huang, Z. Zeng, H. Zhang, Chem. Soc. Rev. 2013, 42, 1934–1946. [26] J. Kang, H. Kim, K. S. Kim, S.-K. Lee, S. Bae, J.-H Ahn, Y.-J. Kim, J.-B. Choi, B. H. Hong, Nano Lett. 2011, 11, 5154–5158. [27] D. Sui, Y. Huang, L. Huang, J. Liang, Y. Ma, Y. Chen, Small 2011, 7, 3186–3192. [28] J. J. Bae, S. C. Lim, G. H. Han, Y. W. Jo, D. L. Doung, E. S. Kim, S. J. Chae, T. Q. Huy, N. V. Luan, Y. H. Lee, Adv. Funct. Mater. 2012, 22, 4819–4826.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2014, DOI: 10.1002/smll.201400434
Flexible and Transparent Gas Molecule Sensor Integrated with Sensing and Heating Graphene Layers
[29] H. Kim, Y. Kim, K. S. Kim, H. Y. Jeong, A. R. Jang, S. H. Han, D. H. Yoon, K. S. Suh, H. S. Shin, T. Y. Kim, W. S. Yang, ACS Nano 2013, 7, 5769–5776. [30] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J.-H Ahn, P. Kim, J.-Y. Choi, B. H. Hong, Nature 2009, 457, 706–710. [31] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Velamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo, R. S. Ruoff, Science 2009, 324, 1312–1314. [32] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H. R. Kim, Y. I. Song, Y.-J. Kim, K. S. Kim, B. Ozyilmaz, J.-H. Ahn, B. H. Hong, S. Iijima, Nat. Nanotechnol. 2010, 5, 574–578.
small 2014, DOI: 10.1002/smll.201400434
[33] A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, S. Roth, A. K. Geim, Phys. Rev. Lett. 2006, 97, 187401. [34] G. M. Scarpello, D. Ritelli, Meccanica 2006, 41, 519–527. [35] N. Petrone, I. Meric, J. Hone, K. L. Shepard, Nano Lett. 2013, 13, 121–125. [36] G. Sberveglieri, S. Groppelli, P. Nelli, Sens. Actuators B: Chem. 1991, 4, 457–461. [37] Y. Liu, E. Koep, M. A. Liu, Chem. Mater. 2005, 17, 3997–4000.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: February 17, 2014 Revised: April 1, 2014 Published online:
www.small-journal.com
7
