Gas Detection with Vertical InAs Nanowire Arrays - Nano Letters (ACS

May 26, 2010 - ... indicted for chemical releases. A grand jury in Houston indicted Arkema and two of its executives on criminal charges that they ...
0 downloads 0 Views 1MB Size
pubs.acs.org/NanoLett

Gas Detection with Vertical InAs Nanowire Arrays Peter Offermans,* Mercedes Crego-Calama, and Sywert H. Brongersma Holst Centre/IMEC-NL, PO Box 8550, 5605 KN Eindhoven, The Netherlands ABSTRACT Nanowire-based devices show great promise for next generation (bio)chemical sensors as evidenced by the large volume of high-quality publications. Here, a nanoscale gas sensing device is presented, based on gold-free grown vertical InAs nanowire arrays. The nanowires are contacted Ohmically in their as-grown locations using an air bridge construction, leaving the nanowire surface free for gas adsorption. Noise measurements were performed to determine the measurement resolution for gas detection. These devices are sensitive to NO2 concentrations well below 100 ppb at room temperature. NO2 exposure leads to both a reduction in carrier density and electron mobility. KEYWORDS Nanowire, InAs, sensor, gas, vertical geometry, NO2

emiconductor nanowires offer a promising platform for high-performance sensing devices that employ direct electrical readout. For example, silicon nanowire-based electronic devices have successfully demonstrated label-free sensing of biomolecules in liquids.1 Similar to conventional ion-sensitive field-effect transistors, these devices exhibit a conductance change in response to binding-induced variations in the electric field or potential at the nanowire surface.2,3 In contrast to nanowires in an electrolytic environment, nanowire-based sensing in a gaseous ambient generally involves charge transfer from or to adsorbed molecules or a gas-induced modification of the height of Schottky barriers formed at their contacts. For example, a large enhancement in the sensitivity to gases and biochemicals was recently demonstrated by replacing one of the Ohmic contacts to a ZnO nanowire by a Schottky contact.4,5 Other excellent examples of gas sensing nanostructures include carbon nanotubes6-9 and various metaloxide nanowires.10-12 In contrast, III/V-based nanowires have received relatively little attention for gas sensing. However, growth control, with regards both to the ability to tune electronic properties and to positioning and growth orientation, has reached a high degree of perfection.13-16 Furthermore, examples of III-V nanowire contacting in asgrown locations, abandoning pick-and-place methods17 have been demonstrated.18 However, for gas sensing, there is still a need for a vertical contacting scheme that allows gas adsorption and surface modification enabling tailoring with sensing molecules that are gas specific19,20 (Figure 1). InAs is a promising material for gas sensing, because it exhibits an electron accumulation layer at the surface, which renders it sensitive to accumulated charges or dipoles.19-21 At the same time, InAs allows relatively easy fabrication of

S

Ohmic contacts due to its small band gap. Additionally, surface modification toward selective binding of a variety of gases can build on considerable expertise on functionalization of planar InAs substrates.22 Recently, the response of a horizontal single InAs nanowire field-effect transistor to saturated alcoholic vapors was demonstrated.23 Here, we demonstrate for the first time gas sensing with vertically integrated InAs nanowire arrays. These devices are sensitive to parts-per-billion levels of NO2, which is a strong electron acceptor and known to interact with the InAs surface.19,20 Furthermore, detection of NO2 finds important applications for monitoring environmental pollution resulting from combustion or automotive emission.24 High-quality vertical InAs nanowire arrays can be obtained on InP by two distinctly different approaches. The first depends on nanopatterning of gold islands (20-50 nm diameter) that control the locations where nanowire growth is initiated and continues by the vapor-liquid-solid mechanism.25,26 InAs nanowires can also be fabricated

* To whom correspondence should be addressed, [email protected]. Received for review: 02/13/2010 Published on Web: 05/26/2010

FIGURE 1. Schematic of nanosensor concept. Illustration of selective adsorption of gas molecules on a contacted vertical nanowire surface.

© 2010 American Chemical Society

2412

DOI: 10.1021/nl1005405 | Nano Lett. 2010, 10, 2412–2415

FIGURE 2. TEM imaging of an as-grown InAs nanowire. The images show an InAs nanowire grown by LP-MOVPE without catalyst particle on an InP(111)B substrate, using a 1.5 nm thick SiOx initiation layer. The nanowire has a predominantly zinc blende crystal structure with some wurtzite segments. The nanowire is covered by a 2 nm thick amorphous oxide layer and has a flat top facet.

without using Au or other metal particles as catalyst, by using a SiOx initiation27 or a SiN masking layer.28 The latter approach allows high yield vertical InAs nanowire growth on Si(111) substrates as has recently been demonstrated by Tomioka et al.28 In this study, InAs nanowires were grown using lowpressure metalorganic vapor-phase epitaxy (LP-MOVPE) with trimethylindium (TMI) and arsine (AsH3) as precursor materials, transported in a flow of H2 gas. As substrates we used n-doped epitaxy-ready InP(111)B wafers onto which a 1.2 nm thin SiOx initiation layer was evaporated, which facilitates nanowire nucleation. The substrates were heated to the growth temperature of 600 °C in a H2 atmosphere, and the nanowires were grown in 90 s. Notably, the nanowires appear untapered and have a length of about 3 µm and a thickness of 50-100 nm. The crystal structure of nanowires grown in this manner is predominantly zinc blende but contains considerable parts with wurtzite segments27 as can be seen from the transmission electron microscopy (TEM) image in Figure 2. Furthermore, a 2 nm thick In-rich amorphous oxide layer (InxOy) was observed on the nanowire surface. After nanowire growth (Figure 3a), an 80 nm thick Si3N4 layer was conformally deposited onto the nanowires and patterned into predetermined areas, varying from 120 × 120 µm2 down to 30 × 30 µm2, outside of which the nanowires were subsequently etched away (Figure 3b). Electrical connections were made by an air bridge construction. As the electrical contacts have to be isolated from the InP substrate, an 80 nm thick layer of Si3N4 was again deposited, covering the entire substrate. The nanowires were then fully embedded in a resist, which was etched back using an oxygen plasma, to a height of about 2 µm. By applying a CF4 plasma to the exposed part of the nanowires, only the Si3N4 covering the top of the nanowires was removed, enabling it to be contacted by the air bridge (Figure 3c-e). As a support for the top contact, resist islands were patterned onto the nanowire arrays and etched back to a height of about 2 µm. After a short buffered hydrofluoric acid (BHF) etch, contacts were made by sputter deposition of 10 nm Ti followed by 1 µm of gold (which obviously will need to be © 2010 American Chemical Society

FIGURE 3. Scanning electron microscopy (SEM) images of the nanosensor during different stages of fabrication. (a-f) SEM images showing as-grown nanowires (a), patterned nanowire arrays (b), nanowire arrays contacted to larger bond pads by an air bridge (c), nanowires partially covered by Si3N4 (d), a side view of an air bridge contact (e), and a close-up of a contacted nanowire still partially covered by Si3N4 (f).

replaced by another metal when implementing this approach on a silicon substrate). Subsequently, air bridges were formed by patterning the contact metals using optical lithography and wet etching. Finally, all remaining resist was removed and the structures were dried in isopropanol vapor in order to prevent stiction. By using a high acceleration voltage, the close-up in Figure 3f clearly shows the Si3N4 covering a contacted nanowire. After the nanowires were in contact with the air bridges, the I-V characteristics of the structures were measured by probing a pair of arrays on their top contacts (Figure 4, bottom right). As shown in Figure 4 (top left), the current increases linearly as a function of the applied voltage, indicating the successful formation of Ohmic contacts. To verify that the measured current is actually due to the contacted nanowires, and not due to, e.g., shorts from the contact pads to the conductive substrate, the current is plotted as a function of the array’s surface area (Figure 4a). From the resulting linear dependence, a resistance of 20 kΩ for a single nanowire is obtained, which is comparable to that of gold catalyzed MOVPE grown nanowires reported by Thelander et al.18 To estimate the inherent noise level of the nanowires, the noise spectrum of the nanowire array (with R ∼ 1 kΩ) was determined by measuring the low frequency noise on the nanowire output voltage Vn while forcing a bias current (20 µA) and calculating the low frequency resistor noise using 2413

DOI: 10.1021/nl1005405 | Nano Lett. 2010, 10, 2412-–2415

γR )

γV IB2

(Ω2 /Hz)

(1)

where γ is the noise spectral density per unit hertz. The noise spectrum of vertically connected nanowire arrays was determined using an Agilent 35670A dynamic signal analyzer in the frequency range of 0.7-1600 kHz (Figure 4b). The normalized nanowire noise spectrum is independent of bias current and shows a clear 1/f behavior which fits well to

γR )

α 2 (Ω /Hz) f

FIGURE 5. Gas detection with vertical InAs nanowire arrays. Response of a nanowire array with approximately 200 nanowires to varying concentrations of NO2 in N2. The inset shows a saturated response to 9 ppm NO2/N2.

(2)

with R ) 0.45. By integrating the resistor noise over the applied frequency range using

∆R2 )

∫ γR df

Next, the applicability of this device for gas sensing is determined by testing its response to NO2 exposure in an N2 environment. Using a custom-built MDC probe station with gas-flow capability at atmospheric pressure, the nanowire devices can be exposed to varying NO2 concentrations supplied from a permeation tube into a nitrogen carrier gas flow. Prior to exposure, the protective Si3N4 covering the nanowires was removed using a CF4 plasma. The response to NO2 was measured at room temperature while cycling between pure N2 and NO2 concentrations in the range of 115 ppb to 1.7 ppm using a 1 h cycling time. As shown in Figure 5, the response (1 - ∆I/I0) increases during NO2 exposure, while it recovers by flushing with N2. The onset of the response is almost immediate, and concentrations as low as 115 ppb (∆R ∼ 5%) can easily be detected within 10 min with a signal-to-noise ratio >10. The reduction in device current during gas exposure may be attributed to the expected role of NO2 as an electron acceptor, which would reduce the electron density in the surface electron accumulation layer by charge transfer; however, the electron mobility may also be affected.23 To investigate this further, the change in electron field-effect mobility during gas exposure was determined (see Supporting Information). Both the carrier density and the (apparent) electron field-effect mobility29 decrease during gas exposure. This is in contrast to the findings in ref 23 for the sensing of various saturated vapors. We tentatively attribute these effects to a reduced surface electron accumulation, and increased charging and scattering at InAs/InxOy interface states during NO2 exposure.29-31 Further studies32 on the sensing mechanism as well as the concentration dependence of both the response time and rate are needed to clarify their relation InAs/InxOy interface. In conclusion, gas sensing is demonstrated for the first time with vertical InAs nanowire arrays grown without gold nucleation, contacted Ohmically in as-grown locations, using an integration flow that leaves their surface accessible, and by showing their high sensitivity to levels of NO2 below 100 ppb. Sensitivity and selectivity toward other gases is under investigation and will be accomplished by tailoring the

(3)

an obtainable measurement resolution of ∼2 Ω is determined, which is 0.2% of the device resistance.

FIGURE 4. Electrical characterization of vertical InAs nanowire arrays. (a) Linear dependence of current vs nanowire array size with measurement configuration (bottom right) and linear I-V demonstrating Ohmic contacts (top left). (b) Noise spectrum of a nanowire array with approximately 200 nanowires showing 1/f behavior. © 2010 American Chemical Society

2414

DOI: 10.1021/nl1005405 | Nano Lett. 2010, 10, 2412-–2415

(14) Dayeh, S. A.; Soci, C.; Bao, X.-Y.; Wang, D. Nano Today 2009, 4, 347–358. (15) Li, H.-Y.; Wunnicke, O.; Borgstro¨m, M. T.; Immink, W. G. G.; Van Weert, M. H. M.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Lett. 2007, 7, 1144–1148. (16) Mohan, P.; Motohisa, J.; Fukui, T. Appl. Phys. Lett. 2006, 88, No. 013110. (17) Huang, Y.; Duan, X.; Wei, Q.; Lieber, C. Science 2001, 291, 630– 633. (18) Thelander, C.; Rehnstedt, C.; Froberg, L.; Lind, E.; Martensson, T.; Caroff, P.; Lowgren, T.; Ohlsson, B.; Samuelson, L.; Wernersson, L. IEEE Trans. Electron Devices 2008, 55, 3030–3036. (19) Garcia, M. A.; Losurdo, M.; Wolter, S. D.; Kim, T. H.; Lampert, W. V.; Bonaventura, J.; Bruno, G.; Giangregorio, M.; Brown, A. J. Vac. Sci. Technol., B 2007, 25, 1504–1510. (20) Garcia, M. A.; Losurdo, M.; Wolter, S. D.; Lampert, W. V.; Bonaventura, J.; Bruno, G.; Yi, C.; Brown, A. S. Sens. Lett. 2008, 6, 627–634. (21) Petrovykh, D.; Yang, M.; Whitman, L. Surf. Sci. 2003, 523, 231– 240. (22) Knoben, W.; Brongersma, S. H.; Crego-Calama, M. J. Phys. Chem. C 2009, 113, 18331–18340. (23) Du, J.; Liang, D.; Tang, H.; Gao, X. P. Nano Lett. 2009, 9, 4348– 4351. (24) Vaseashta, A.; Vaclavikova, M.; Vaseashta, S.; Gallios, G.; Roy, P.; Pummakarnchana, O. Sci. Technol. Adv. Mater. 2007, 8, 47–59. (25) Martensson, T.; Carlberg, P.; Borgstrom, M.; Montelius, L.; Seifert, W.; Samuelson, L. Nano Lett. 2004, 4, 699–702. (26) Persson, A. I.; Fro¨berg, L. E.; Samuelson, L.; Linke, H. Nanotechnology 2009, 20, 225304. (27) Mandl, B.; Stangl, J.; Martensson, T.; Mikkelsen, A.; Eriksson, J.; Karlsson, L.; Bauer, G.; Samuelson, L.; Seifert, W. Nano Lett. 2006, 6, 1817–1821. (28) Tomioka, K.; Motohisa, J.; Hara, S.; Fukui, T. Nano Lett. 2008, 8, 3475–3480. (29) Dayeh, S. A.; Soci, C.; Yu, P. K. L.; Yu, E. T.; Wang, D. Appl. Phys. Lett. 2007, 90, 162112. (30) Ford, A. C.; Ho, J. C.; Chueh, Y. L.; Tseng, Y. C.; Fan, Z.; Guo, J.; Bokor, J.; Javey, A. Nano Lett. 2009, 9, 360–365. (31) Koley, G.; Cai, Z. IEEE Sens. 2008, 118, 118–121. (32) A detailed comparison of the gas response of nanowire devices in the vertical and horizontal geometry is under investigation and will be published elsewhere. (33) Tanzer, T.; Bohn, P.; Roshchin, I.; Greene, L.; Klem, J. Appl. Phys. Lett. 1999, 75, 2794–2796. (34) Hang, Q.; Wang, F.; Carpenter, P.; Zemlyanov, D.; Zakharov, D.; Stach, E.; Buhro, W.; Janes, D. Nano Lett. 2008, 8, 49–55.

surface of the nanowires with sensing molecules that selectively bind gases. For example, redox-active molecules such as metalloporphyrins and phthalocyanines are known to form charge-transfer complexes with electron accepting or reducing gases. Functionalization of the InAs surface has already been achieved in-house by employing thiol-based molecular linkage22 and by binding directly to the native oxide. Surface passivation,33,34 in combination with functionalization, may also improve response times. Thus, these nanowire-based devices form a promising platform for highperformance nanosensors that employ direct electrical readout. Supporting Information Available. I/Vg characteristics before and after NO2 exposure measured in a horizontal nanowire geometry. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)

Stern, E.; Klemic, J. F.; Routenberg, D. A.; Wyrembak, P. N.; Turner-Evans, D. B.; Hamilton, A. D.; LaVan, D. A.; Fahmy, T. M.; Reed, M. A. Nature 2007, 445, 519–522. Cui, Y.; Wei, Q.; Park, H.; Lieber, C. M. Science 2001, 293, 1289– 1292. Nair, P. R.; Alam, M. A. Nano Lett. 2008, 8, 1281–1285. Yeh, P. H.; Li, Z.; Wang, Z. L. Adv. Mater. 2009, 21, 4975–4978. Wei, T. Y.; Yeh, P. H.; Lu, S. Y.; Wang, Z. L. J. Am. Chem. Soc. 2009, 131, 17690–17695. Kong, J.; Franklin, N. R.; Zhou, C. W.; Chapline, M. G.; Peng, S.; Cho, K. J.; Dai, H. J. Science 2000, 287, 622–625. Collins, P.; Bradley, K.; Ishigami, M.; Zettl, A. Science 2000, 287, 1801–1804. Li, J.; Lu, Y.; Ye, Q.; Cinke, M.; Han, J.; Meyyappan, M. Nano Lett. 2003, 3, 929–933. Heinze, S.; Tersoff, J.; Martel, R.; Derycke, V.; Appenzeller, J.; Avouris, P. Phys. Rev. Lett. 2002, 89, 106801. Comini, E. Anal. Chim. Acta 2006, 568, 28–40. Kolmakov, A.; Klenov, D.; Lilach, Y.; Stemmer, S.; Moskovits, M. Nano Lett. 2005, 5, 667–673. Zhang, D.; Liu, Z; Li, C.; Tang, T.; Liu, X.; Han, S.; Lei, B.; Zhou, C. Nano Lett. 2004, 4, 1919–1924. Bjork, M.; Thelander, C.; Hansen, A.; Jensen, L.; Larsson, M.; Wallenberg, L.; Samuelson, L. Nano Lett. 2004, 4, 1621–1625.

© 2010 American Chemical Society

2415

DOI: 10.1021/nl1005405 | Nano Lett. 2010, 10, 2412-–2415