Probing Electronic Doping of Single-Walled Carbon Nanotubes by

Publication Date (Web): November 13, 2012 ... of SWNT-based gas-sensing devices or the stability of individual SWNT-based field effect transistors (FE...
1 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/JPCL

Probing Electronic Doping of Single-Walled Carbon Nanotubes by Gaseous Ammonia with Dielectric Force Microscopy Jie Zhang,†,‡ Wei Lu,‡ Yize Stephanie Li,‡ Di Lu,‡ Ting Zhang,‡ Xiaoping Wang,† and Liwei Chen*,‡ †

Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei Anhui 230026, China ‡ i-LAB, Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences, Suzhou, Jiangsu 215123, China S Supporting Information *

ABSTRACT: The electronic properties of single-walled carbon nanotubes (SWNTs) are sensitive to the gas molecules adsorbed on nanotube sidewalls. It is imperative to investigate the interaction between SWNTs and gas molecules in order to understand the mechanism of SWNT-based gas-sensing devices or the stability of individual SWNT-based field effect transistors (FETs). To avoid the Schottky barrier at the metal/SWNT contact, which dominates the performance of SWNT-based FETs, we utilize a contactless technique, dielectric force microscopy (DFM), to study the intrinsic interaction between SWNTs and gaseous ammonia molecules. Results show that gaseous ammonia affects the conductivity of semiconducting SWNTs but not metallic SWNTs. Semiconducting SWNTs, which are p-type doped in air, show suppressed hole concentration in ammonia gas and are even inverted to n-type doping in some cases. SECTION: Physical Processes in Nanomaterials and Nanostructures

S

charge carrier type and concentration in SWNTs are greatly altered upon exposure to ammonia. The change of the SWNT electronic property is nonuniform, with significant tube-to-tube variation, and results from chemisorption-induced doping rather than substrate gating effects. Our previous work established a DFM-based method for contactless probing of SWNT metallicity.13 The dielectric response of metallic and semiconducting materials is dominated by the motion of charge carries, which suggests that both dielectric response and electric conductivity are determined by the carrier concentration and mobility. Thus, the low-frequency dielectric response of SWNTs measured by the DFM technique indirectly reflects their conductivity. When a gate bias voltage is applied to the DFM tip to locally modulate the charge carrier concentration, the DFM signal of semiconducting SWNTs varies significantly in response to the gate voltage. For example, p-type semiconducting SWNTs exhibit carrier accumulation and strong DFM signal under negative local gate bias voltage but exhibit carrier depletion and weak DFM signal under positive local gate bias. On the other hand, metallic tubes remain nearly invariant under different gate bias. A gate modulation ratio was defined as the ratio of the SWNT DFM signal under two different local gate voltages, for example, under 2 V of local gate bias (S2V) to that under −2 V of local gate (S−2V) to identify the SWNT metallicity: metallic SWNTs

ingle-walled carbon nanotubes (SWNTs) exhibit extraordinary mechanical, thermal and electrical properties due to their unique one-dimensional all-carbon structure.1,2 For many proposed applications such as those in nanoelectronic and optoelectronic devices, the interaction between SWNTs and gaseous molecules is critically important. When individually utilized, SWNTs are composed entirely of surface atoms, and thus, the electronic properties of SWNTs are extremely sensitive to the gaseous environment.3,4 On one hand, this may impose serious concerns on individual SWNT device stability, but on the other hand, it may also be exploited in gassensing devices.5−7 In both cases, it is crucial to understand how the interaction between SWNTs and gaseous molecules affects the electronic properties of SWNTs. Generally speaking, electronic properties of nanomaterials are investigated via field effect transistor (FET) device transport measurements.8,9 However, individual SWNT FET devices have been shown to be dominated by nanotube−electrode contacts rather than SWNTs themselves.10,11 The interaction between gas molecules and SWNT/metal contacts modulates the height of the Schottky barrier at the contact and strongly affects the device performance.12 Therefore, electronic interactions between SWNTs and gaseous molecules cannot be readily examined. Here, we present a dielectric force microscopy (DFM)13 investigation on the effect of gaseous ammonia on SWNTs. DFM is a proximal probe based technique that is capable of contactless probing of electronic properties with nanometerscaled spatial resolution. We identify mobile charge carrier type in SWNTs with DFM and observe the changes upon interaction with ammonia gas. Our results show that the © 2012 American Chemical Society

Received: October 10, 2012 Accepted: November 13, 2012 Published: November 13, 2012 3509

dx.doi.org/10.1021/jz301622a | J. Phys. Chem. Lett. 2012, 3, 3509−3512

The Journal of Physical Chemistry Letters

Letter

Figure 1. Schematic illustration of the experimental setup of EFM and DFM.

exhibited S2V/S−2V close to 1; semiconducting SWNTs (typically p-type doped in air) showed S2V/S−2V less than 1; and n-type semiconductors (e.g., ZnO nanowires in air) showed S2V/S−2V greater than 1.13 As shown in Figure 1, 12 individual SWNTs synthesized with the laser ablation method (Supporting Information) were examined for their interaction with gaseous ammonia using electrostatic force microscopy (EFM) and DFM techniques in this study. The electronic nature of the SWNTs, that is, metallic or semiconducting, was first investigated in a pure nitrogen environment. The gate modulation ratio of these SWNTs is listed in Table S2 (Supporting Information). Six out of the 12 SWNTs were found to exhibit metallic behavior with a gate modulation ratio S2V/S−2V in the 0.9−1.5 range. The other six SWNTs showed p-type semiconducting behavior, that is, stronger DFM signal under negative gate bias than that under positive gate bias, with a gate modulation ratio S2V/S−2V in the 0.1−0.3 range. These observations are consistent with previous results and constitute the basis for studying SWNT interaction with NH3. DFM imaging in different concentrations of NH3 environments reveals that the electronic conductivity of metallic SWNTs does not change much upon interaction with gaseous ammonia. Figure 2a shows a matrix of DFM images under different bias and different NH3 concentrations from a single SWNT, whose topographical image is presented in Figure S1a in the Supporting Information. The SWNT in Figure 2a consists of an intratube junction as manifested in different gate response across the junction in the middle of the tube. The left side section (enclosed in the red square box) is a metallic SWNT section because it displays roughly unchanged DFM signal from −2 to 2 V gate bias in a nitrogen environment (Figure 2a, row A). In rows B, C, D, and E, the concentration of NH3 is increased to 5, 50, and 100% and then purged back to 0%, respectively. The DFM signal of this tube remains rather stable at either gate bias change or NH3 concentration. The third line of the Table S2 (Supporting Information) quantitatively shows that the gate modulation ratio (S2V/ S−2V) of this tube remains close to 1 under different NH3 concentrations. The gate modulation ratio of the other five metallic SWNTs exhibits similar behavior (Supporting Information). Therefore, the dielectric responses of metallic SWNTs are essentially independent of the gate bias and NH3 molecule. Semiconducting SWNTs exhibit more complicated behavior upon interacting with ammonia. In one type of response to NH3, charge carriers in the SWNT change from holes to

Figure 2. DFM images of SWNTs under different concentrations of NH3 and varied local gate voltages. (a) A SWNT with a metallic section on the left and a semiconducting section on the right, in which the metallic section shows little change upon exposure to increased concentration of NH3 (row A: pure N2; row B: 5% NH3; row C: 50% NH3; row D: 100% NH3; row E: purged and reimaged in N2), while the seconducting section shows suppressed conductivity and even carrier type inversion under high NH3 concentrations (rows C and D). (b) A semiconducting SWNT shows suppressed conductivity but no carrier inversion under increased concentration of NH3 (row F: pure N2; row G: 5% NH3; row H: 50% NH3; row I: 100% NH3; row J: purged and reimaged in N2). The scale bar in row A is 500 nm.

electrons, while in the other type of response, holes in the SWNT are depleted, but charge carriers are not inverted to electrons. The right side section of the tube in Figure 2a (enclosed in the black square box, labeled as tube A) is a p-type doped semiconducting SWNT in nitrogen gas, indicated by the decreasing DFM signal with increasing gate voltage (row A) and a gate modulation ratio S2V/S−2V of 0.2. Upon exposure to 5% ammonia (row B), the DFM signal becomes indistinguishable from the background, suggesting nearly depleted hole concentration of the tube under this particular gaseous environment. Further increase of ammonia concentration to 50 and 100% (rows C and D) brings back the DFM signal, but intriguingly, the DFM signal now increases with increasing gate voltage. The gate modulation ratio S2V/S−2V of tube A in 50 and 100% ammonia reaches 2.5 and 6.5, respectively, indicating that charge carrier type in this SWNT has been inverted to electrons under a high concentration of gaseous ammonia. These results are consistent with previous observations from SWNT FET experiments. For example, Kong et al. and Brandley et al. both observed that the conductance of individual semiconducting SWNTs decreased after exposure to 1% NH3 in Ar (equivalent to 0.45 mM) or a 60 mM NH3 aqueous solution.3,14 In Brandley’s work, semiconducting SWNTs showed n-type doped behavior when the concentration of ammonia was larger than 0.1 M in aqueous solution. Kong et al. did not 3510

dx.doi.org/10.1021/jz301622a | J. Phys. Chem. Lett. 2012, 3, 3509−3512

The Journal of Physical Chemistry Letters

Letter

observe the inversion from p-type to n-type probably due to the low concentration of NH3. In our measurements, the inversion from p-type to n-type happens at 50% NH3 concentration (∼22 mM) or higher. Figure 2b shows another type of response of a semiconducting SWNT to NH3. The nanotube in Figure 2b, which is labeled as tube B, exhibits p-type doped behavior in nitrogen gas (row F) with a gate modulation ratio S2V/S−2V of 0.3. In 5% NH3 (row G), the DFM signal also shows significant reduction, but at high NH3 concentration of 50 and 100% (rows H and I), the DFM signal does not show clear n-type behavior, with S2V/ S−2V ratios of 0.4 and 1.7, respectively. This indicates a significant change in the carrier concentration but not quite a clear inversion of carrier type yet. Among the six semiconducting SWNTs investigated in this study, we found two tubes whose charge carrier type is clearly inverted to electrons under a high concentration of ammonia, and the other four tubes show suppressed DFM signal, that is, depleted carrier concentration but not quite inverted carrier type. All semiconducting SWNTs return to their original p-type doping property after purging with nitrogen and exposing to air (rows E and J in Figure 2a and b), indicating that the interaction with gaseous ammonia is reversible. The other four semiconducting SWNTs exhibit the dielectric response similar to that of either tube A or B. The data shown above are the first clarification of the SWNT conductivity change in gaseous environments without interference of metal contacts. In particular, the interaction of the SWNT with ammonia gas results in suppression of hole concentration and even electron doping. The extent to which the electronic doping reaches varies significantly from one tube to another. In some semiconducting SWNTs, the hole concentration is suppressed in NH3, while in some other semiconducting SWNTs, holes are depleted, and charge carrier type is further inverted to electrons. Two possible mechanisms have been proposed to account for the electronic doping of semiconducting SWNTs upon exposure to ammonia.3,15 First, ammonia may adsorb on the SiO2 surface of the substrate, which changes the surface potential as well as the effective “gating” of the substrate.3 Second, ammonia molecules may directly adsorb on SWNTs and result in charge transfer and electronic doping. In order to understand the effect of NH3-induced surface gating, we study the surface potential of the SiO2 substrate under different concentrations of NH3 using the EFM technique. Simultaneous with the DFM measurements, the contact potential difference (CPD) between the conducting tip and the sample is obtained by recording the DC voltage needed to zero out the EFM signal. We find that the surface potential of the Ti−Pt-coated DFM tip changes very little in NH3 from that in N2 (Table S1 in Supporting Information). Therefore, the CPD change in different gaseous environments is approximately the surface potential change of the sample. We analyze the surface potential of the SiO2 substrate around the two semiconducting SWNTs in Figure 2a and b, which are referred to as tubes A and B hereafter, respectively. As shown in Figure 3, a decrease in the SiO2 substrate surface potential of about 200 mV is observed in both cases. Importantly, the decrease in surface potential is finished at 10% NH3 concentration, and the surface potential of SiO2 is almost a constant when the NH3 concentration is increased from 10 to 100%. In contrast, major changes in conductivity or dielectric response of tubes A and B occur under the NH3 concentration range beyond 10%.

Figure 3. Substrate surface potential and DFM signal under different NH3 concentrations for tubes A and B (the semiconducting SWNTs in Figure 2a and b, respectively). Black squares are the CPD of the SiO2 substrate near the tubes, and red circles are the DFM signals of tubes A and B. The solid symbols represent the data point of the tubes after purging with N2, exposing to air, and reimaging in N2.

After purging with nitrogen and exposing to air, the surface potential of SiO2 remains at a low level, indicating strong NH3 adsorption on SiO2, while the dielectric responses of tubes A and B are restored to their original states under the same conditions. Another point to notice is that the negatively shifted surface potential of SiO2 shall result in hole accumulation, similar in effect to a negative gate voltage on the SWNT FET. However, what was experimentally observed for semiconducting SWNTs was hole depletion or even carrier type inversion to electrons in some cases. Therefore, the surface potential of SiO2 or the surface gating effect is not the dominating factor for the electrical property of the s-SWNT.8 Several theoretical works have pointed out that the interaction between NH3 molecules and defect-free SWNTs is weak physisorption, and the electron transfer between NH3 and the SWNT is about 0.04 electron per NH3 molecule.6,14,16 The amount of the transferred charge is so small that the density of state (DOS) of the SWNT changes very little in an ammonia environment.16 Therefore, it becomes an interesting topic to account for experimentally observed s-SWNT sensitivity to ammonia. Feng Xue et al. studied ammonia adsorption on SWNTs by infrared spectra and theoretical calculations.17 The interaction between NH3 molecules and the (5,5) SWNT was calculated using the SCC-DFTB method. Their theoretical results revealed that the interaction between NH3 and the defect-free SWNT (or defective SWNT containing missing carbon atoms) is very weak, with an interaction energy of −12.1 kJ/mol (or −15.9 kJ/mol). However, the interaction energy increased to −40.2 kJ/mol when oxygen molecules coadsorbed on the SWNT. Experimental results from Feng Xue et al. also demonstrated that the NH3 binds to an oxidized SWNT significantly better than that 3511

dx.doi.org/10.1021/jz301622a | J. Phys. Chem. Lett. 2012, 3, 3509−3512

The Journal of Physical Chemistry Letters

Letter

(3) Kong, J.; Franklin, N. R.; Zhou, C. W.; Chapline, M. G.; Peng, S.; Cho, K. J.; Dai, H. J. Nanotube Molecular Wires as Chemical Sensors. Science 2000, 287, 622−625. (4) Langlet, R.; Arab, M.; Picaud, F.; Devel, M.; Girardet, C. Influence of Molecular Adsorption on the Dielectric Properties of a Single Wall Nanotube: A Model Sensor. J. Chem. Phys. 2004, 121, 9655−9665. (5) Li, J.; Lu, Y. J.; Ye, Q.; Cinke, M.; Han, J.; Meyyappan, M. Carbon Nanotube Sensors for Gas and Organic Vapor Detection. Nano Lett. 2003, 3, 929−933. (6) Zhao, J. J.; Buldum, A.; Han, J.; Lu, J. P. Gas Molecule Adsorption in Carbon Nanotubes and Nanotube Bundles. Nanotechnology 2002, 13, 195−200. (7) Goldoni, A.; Larciprete, R.; Petaccia, L.; Lizzit, S. Single-Wall Carbon Nanotube Interaction with Gases: Sample Contaminants and Environmental Monitoring. J. Am. Chem. Soc. 2003, 125, 11329− 11333. (8) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. RoomTemperature Transistor Based on a Single Carbon Nanotube. Nature 1998, 393, 49−52. (9) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666−669. (10) Martel, R.; Derycke, V.; Lavoie, C.; Appenzeller, J.; Chan, K. K.; Tersoff, J.; Avouris, P. Ambipolar Electrical Transport in Semiconducting Single-Wall Carbon Nanotubes. Phys. Rev. Lett. 2001, 87. (11) Derycke, V.; Martel, R.; Appenzeller, J.; Avouris, P. Controlling Doping and Carrier Injection in Carbon Nanotube Transistors. Appl. Phys. Lett. 2002, 80, 2773−2775. (12) Cui, X. D.; Freitag, M.; Martel, R.; Brus, L.; Avouris, P. Controlling Energy-Level Alignments at Carbon Nanotube/Au Contacts. Nano Lett. 2003, 3, 783−787. (13) Lu, W.; Zhang, J.; Li, Y. S.; Chen, Q.; Wang, X. P.; Hassanien, A.; Chen, L. W. Contactless Characterization of Electronic Properties of Nanomaterials Using Dielectric Force Microscopy. J. Phys. Chem. C 2012, 116, 7158−7163. (14) Bradley, K.; Gabriel, J. C. P.; Briman, M.; Star, A.; Gruner, G. Charge Transfer from Ammonia Physisorbed on Nanotubes. Phys. Rev. Lett. 2003, 91. (15) Peng, N.; Zhang, Q.; Chow, C. L.; Tan, O. K.; Marzari, N. Sensing Mechanisms for Carbon Nanotube Based NH3 Gas Detection. Nano Lett. 2009, 9, 1626−1630. (16) Chang, H.; Lee, J. D.; Lee, S. M.; Lee, Y. H. Adsorption of NH3 and NO2 Molecules on Carbon Nanotubes. Appl. Phys. Lett. 2001, 79, 3863−3865. (17) Feng, X.; Irle, S.; Witek, H.; Morokuma, K.; Vidic, R.; Borguet, E. Sensitivity of Ammonia Interaction with Single-Walled Carbon Nanotube Bundles to the Presence of Defect Sites and Functionalities. J. Am. Chem. Soc. 2005, 127, 10533−10538.

to a pristine SWNT. Similarly, Bradley et al. found that NH3 molecules cause changes in the SWNT electrical property when water molecules coadsorb on the SWNT.14 Therefore, an emerging theme from the literature is that semiconducting SWNTs can be electron-doped by NH3 via charge transfer only with the presence of functional groups or coadsorbed O2/water at the tube sidewall. This may well account for the tube-to-tube variation observed in our experiments, especially in those semiconducting SWNTs. This direct experimental investigation of the SWNT conductivity change in different gaseous environments may have important implications in carbonnanotube-based gas-sensing applications. The results help to explain large variations in device performance of individual SWNT-based sensors and suggest that controlled functionalization will be critically important in thin-film-based carbon nanotube sensors. In conclusion, we investigate the electronic doping of SWNTs by NH3 using the DFM technique. While the electrical property of metallic SWNTs is not affected by NH3, that of semiconducting SWNTs is dramatically altered upon exposure to NH3. The hole doping level in semiconducting SWNTs is significantly reduced upon NH3 exposure, and some SWNTs can even be inverted to n-type doping. Tube-to-tube variation among semiconducting SWNTs is ascribed to differences in structure, defects, and functionalization on nanotube sidewalls. This work demonstrates that the DFM technique is not only a convenient method for characterization of electronic properties of nanomaterials but also a very unique one due to the spatial mapping capability and the contactless nature. We believe that the DFM method will help to address various materials and device characterization challenges in the future.

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental section, supporting results, and discussion in figures and tables including topographic images of SWNTs, discussion on the influence of ammonia for the tip and substrate, the gate modulation ratio of the 12 SWNTs measured in this work, and the discussion on the relationship between SWNT diameter and charge-transfer property with ammonia. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (Grants 10904105, 21103222, and 11005150), the Knowledge Innovation Program of the Chinese Academy of Sciences (No. KJCX2-YW-H21), and the National Basic Research Program of China (2010CB934700). L.C. thanks the Chinese Academy of Science “Hundred Talents” program.

■

REFERENCES

(1) Avouris, P. Molecular Electronics with Carbon Nanotubes. Acc. Chem. Res. 2002, 35, 1026−1034. (2) Dai, H. J. Carbon Nanotubes: Synthesis, Integration, and Properties. Acc. Chem. Res. 2002, 35, 1035−1044. 3512

dx.doi.org/10.1021/jz301622a | J. Phys. Chem. Lett. 2012, 3, 3509−3512