Electronic Conductance of Platinum Atomic Contact in a Nitrogen

Article pubs.acs.org/JPCC

Electronic Conductance of Platinum Atomic Contact in a Nitrogen Atmosphere Satoshi Kaneko,† Jinjiang Zhang,‡ Jianwei Zhao,*,‡ and Manabu Kiguchi*,† †

Department of Chemistry, Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1 W4-10 Ookayama, Meguro-ku, Tokyo 152-8551, Japan ‡ State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, People’s Republic of China ABSTRACT: We have investigated the platinum atomic contact in a nitrogen atmosphere using mechanically controllable break junction technique and theoretical calculation. Conductance, dI/dV measurement, and theoretical calculation revealed a single N2 molecule placed between platinum electrodes, where the N2 molecular axis was parallel to the junction axis. The conductance of the single N2 molecular junction was 1 G0 (=2e2/h), which was comparable to that of metal atomic contacts indicating the strong interaction between the nitrogen molecule and Pt electrodes. Theoretical calculation supported the high conductivity of the molecular junction and revealed that two dominant channels contributed to the electron transport through the junction. The two dominant channels were from the coupling between orbitals (dxz and dyz) of Pt atom and orbitals (px and py) of N atom.

■

INTRODUCTION There has been considerable interest in metal nanocontacts because of their potential application to ultrasmall electronic devices.1−9 Currently, interaction between metal nanocontacts and foreign atoms or molecules is one of the hot topics. Interesting interaction and adsorption properties could appear for the metal nanocontacts because the atomic and electronic structures of the metal nanocontacts are different from those of bulk metals because of the decrease in the number of the neighboring atoms. The previously reported transmission electron microscopy (TEM) study revealed the formation of the Au atomic wire in ultrahigh vacuum (UHV). The interatomic distance between Au atoms was extremely long (0.35−0.4 nm) compared to that of bulk (0.25 nm).3 The long interatomic distance was explained by the formation of a zigzag structure or by the presence of impurity atoms, such as C, H, or O in the nanocontact. The theoretical calculation showed that the metal nanowire could be stabilized by the adsorption of atoms or molecules on the metal nanowire or incorporation of the atoms or molecules into the nanowire. The formation of Ag and Co atomic wires was reported in the presence of hydrogen at 4 K, although clean silver and cobalt do not form atomic wires.4−6 In the case of Pt nanocontacts, the formation of the single hydrogen molecular junction was revealed by the point contact spectroscopy (PCS), inelastic electron tunneling spectroscopy (IETS), shot noise measurements, and theoretical calculation.7,8 In solution, the interesting interaction between hydrogen and the Au atomic contact was observed during the hydrogen evolution reaction.9 The hydrogen atom adsorbed on the Au monoatomic contact at the hydrogen evolution potential. There are little experimental results which directly show the existence of hydrogen on the Au flat surface during © XXXX American Chemical Society

the hydrogen evolution reaction. Currently, interaction between metal nanocontact and nitrogen has attracted a lot of interest.10−12 When N2 molecules interact with transition metals, the reactivity of N2 molecules is enhanced, (e.g., ammonium synthetic reaction10). The reactivity of N2 could be increased further for the metal nanocontact because of the decrease in the number of the neighboring atoms. The interaction between Cu nanocontact and nitrogen has been investigated using ab-initio calculation.11,12 The calculation result showed a tendency of N and N2 to make strong bonds in Cu nanocontacts leading to the formation of longer Cu atomic wires. Although the interesting interaction between metal nanocontact and nitrogen was proposed, there has been little experimental investigation on this topic. In this study, we have investigated the interaction between nitrogen and Pt nanocontact at 10 K. The Pt electrode was more reactive than the Cu electrode because of the higher density of states at the Fermi level.13−17 The atomic and electronic structures of Pt nanocontacts in a nitrogen atmosphere have been investigated using mechanically controllable break junction (MCBJ)18 and theoretical calculation.19−21 The conductance and vibration spectroscopy measurement combined with the theoretical study revealed the formation of the single N2 molecular junction. The conductance of the single N2 molecular junction was 1 G0, which was close to metal atomic contact. Received: February 13, 2013 Revised: March 25, 2013

A

dx.doi.org/10.1021/jp401562d | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

■

Article

EXPERIMENTAL SECTION Experiments have been performed using mechanically controllable break junction (MCBJ).18 A notched Pt wire (0.1 mm in diameter, 10 mm in length) was used for the electrode. After the wire was fixed on elastic substrate (5 × 25 mm, t = 1 mm) by epoxy adhesive (Stycast 2850FT), the substrate was mounted in a three-point bending configuration inside a vacuum chamber that was pumped to a pressure below 2 Pa. After a cryogenic vacuum was attained by cooling to about 10 K, the notched part of the wire was broken by bending of the substrate. Using piezo element, the bending was controlled with atomic-scale resolution. Many cycles of breaking and relaxing the Pt contacts enabled us to fabricate atomic contacts repeatedly. N2 gas (>99.9995 vol %, several μmol) was introduced into the chamber through a heated capillary. DC two-point voltage-biased conductance was monitored during the breaking process. dI/dV spectrum was measured using lockin technique by fixing the movement of the piezo element. The conductance was monitored using an AC modulation of 1 mV amplitude and a frequency of 7.777 kHz while sweeping the DC bias between −80 and +80 mV.

Figure 1. (a) Conductance traces before (black) and after (red) introduction of N2 gas measured at the bias voltage of 100 mV. (b) Conductance histogram constructed from 1000 conductance traces before (black) and after (red) introduction of N2 gas. The bin size was 0.004 G0. The count was normalized by the number of total breaking processes.

■

ultrahigh vacuum at low temperature.2 Electrical conductance through a metal atomic contact is expressed by G = 2e2/h∑ Ti, where Ti is the transmission coefficient of the ith conductance channel, e is the electron charge, and h is Plank’s constant. As several channels contribute to the conductance in transition metal, Pt atomic contact showed conductance of 1.5−2 G018 depending on the atomic configuration of the contact. In the presence of nitrogen, a plateau was observed at 1 G0 in addition to the 1.5 G0 plateau. A conductance histogram constructed from a thousand of conductance traces showed both 1 G0 and 1.5 G0 peaks (see Figure 1b). The appearance of the 1 G0 peak indicated the formation of the stable atomic contact showing conductance of 1 G0. To investigate the atomic configuration of the atomic contact showing 1 G0, vibration spectroscopy for the single molecule junction was measured. The vibration spectrum (d2I/dV2 curve) was obtained by numerical derivation of the differential conductance (dI/dV) curve of the contact. The dI/dV curve was measured as a function of voltage across the contact while the electrodes’ separation was fixed. Figure 2a represents a typical dI/dV curve measured at conductance around 1 G0. Symmetric conductance enhancement was observed around ±40 meV, which originated from electron−vibration interaction in contact.7,13,14,26

THEORETICAL CALCULATIONS The inset of Figure 3 shows the structure models investigated in this study. The left and right Pt electrodes were constructed from metal bulk oriented in the (111) direction and atomically sharp tip. The tips were modeled as four-atom tetrahedrons. In the model named “atomic” (Pt−N), the atomic nitrogen was placed between Pt electrodes. The other models (Pt−N2) represent the N2 molecule placed between Pt electrodes with its molecular axis vertical (“bridge”, “vertical”) or parallel (“parallel”) to the junction axis. While the N2 molecule was placed on the line of two top Pt atoms in the vertical configuration, the N2 molecule was placed above the Pt−Pt bond in the bridge configuration. The geometry of each model was optimized with first-principle computer code Atomistix Toolkit (ATK) software22−24 during which the bulk of electrodes was fixed, while the N atom (or N2 molecule) and the two tips were relaxed. The total energy of the system was calculated as a function of the electrode distances, where the distance was the length between two Pt atoms at the top of the tetrahedrons. The electron-transport behavior of the models at stable distances was also investigated. During these calculations, double-ξ with polarization basis set and single-ξ with polarization basis set were chosen for N atoms and Pt atoms, respectively. Models only consisting of a N atom (or N2 molecule) and two four-atom Pt tetrahedrons were used for the calculation of vibration energy before geometry optimization was carried out at b3lyp level with lanl2dz basis set with Gaussian software.25 During the geometry optimization, each four-atom Pt tetrahedron was kept as a group, while the distance between two tetrahedrons and the conformation of the N atom (or N2 molecule) were relaxed.

■

RESULTS AND DISCUSSION Figure 1a shows typical conductance traces representing conductance change of the Pt contacts during the breaking process of the contact. Stepwise fashion represented the atomic structural transition of the contact. The plateau at 1.5 G0 (2e2/ h) corresponded to the conductance of a Pt atomic contact, which agreed with the previously reported result obtained in

Figure 2. (a) Typical dI/dV spectrum of the Pt−N2 junction showing conductance close to 1 G0. (b) Histogram of the vibration energy of the Pt−N2 junction constructed from 41 dI/dV curves. B

dx.doi.org/10.1021/jp401562d | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

Article

45.0 meV) around 40 meV. These vibration modes are depicted in Figure 4. On the basis of the experimental and theoretical

To accurately determine the vibration energy, 41 differential conductance spectra were collected for the contact in the conductance regime of 0.05∼1.2 G0 resulting in the energy histogram presented in Figure 2b. The distribution could be fitted by a Gaussian function, whose peak center was 43 meV and whose full width at half-maximum was 8 meV. As the energy of 43 meV was larger than the phonon energy of Pt27 and much smaller than that of the internal vibration modes of N2,28 the vibration mode of 43 meV would correspond to the vibration mode between Pt electrode and nitrogen. The existence of nitrogen in the Pt contact was revealed by the vibration spectroscopy. Further discussion about the atomic structure and the vibration mode is given in the next theoretical calculation part. With a view to investigating the structure of the molecular junction, theoretical calculations were performed for four possible structures (see inset in Figure 3). Figure 3 represents

Figure 4. Three possible vibration modes of the Pt−N2 junction with the parallel configuration.

calculation, the Pt−N2 system showing conductance close to 1 G0 was assigned to the parallel configuration. Theoretical calculation also revealed that the N−N bond was stretched for the Pt−N2 junction. While the N−N bond was 1.09 Å for the isolated N2 molecule, the N−N bond was 1.16 Å in the molecular junction because of the interaction between the molecule and the metal electrodes. To get a better understanding of the electron-transport mechanism through the Pt−N2 junction with the parallel configuration, we analyzed the transmission coefficient at the Fermi level and found that there were two dominant channels which transferred electrons through the junction. Their transmission probabilities were 0.458 and 0.454, respectively. From transmission spectrum (Figure 5a), we could roughly

Figure 5. (a) Transmission spectrum at zero bias and (b) two dominant channels of the peak at −0.1 eV for the Pt−N2 junction with the parallel configuration. Figure 3. Total energy of the (a) Pt−N and (b) Pt−N2 junctions as a function of the distance between electrodes.

predict that these two dominant channels originate from the tail of the broad peak at −0.1 eV. Figure 5b shows the main channels of −0.1 eV. We could find that these π-type channels were from the effective coupling between orbitals (dxz and dyz) of Pt atom and orbitals (px and py) of N atom.29 Finally, we briefly comment on the shape of dI/dV curves. While symmetric conductance enhancement was observed in dI/dV curves like Figure 2a, conductance suppression was also observed in dI/dV curves. Figure 6 shows the distribution of the conductance enhancement and suppression in dI/dV curves for the Pt−N2 junctions as a function of conductance and energy in the conductance range from 0.05 G0 to 1.2 G0. Conductance enhancement was observed more frequently than conductance suppression. There was no correlation between conductance and shape of dI/dV curves. The previous

the total energy of the junctions as a function of the distances between the electrodes. The stable states appeared at 3.65, 2.50, 4.29, and 5.03 Å for the atomic, bridge, vertical, and parallel configurations, respectively. The bridge and parallel configurations were much more stable than that of the vertical configuration. The conductance was calculated to be 1.34, 2.91, 1.12, and 0.98 G0 for the atomic, bridge, vertical, and parallel configurations, respectively. We then calculated the vibration energy for each junction. Table 1 shows the summary of the vibration modes around 40 meV, which was observed in the present study. For the atomic and vertical configurations, there were no vibration modes around 40 meV. For the parallel configuration, there were three vibration modes (42.2, 44.1,

Table 1. Summary of the Calculation Results of the Pt−N and Pt−N2 Junctions model

atomic

bridge

vertical

parallel

most stable distance (Å) conductance at 0 V (G0) vibration modes around 40 meV

3.65 1.34

2.50 2.91 37.4, 44.1 meV

4.29 1.12

5.03 0.98 42.2, 44.1, 45.0 meV

C

dx.doi.org/10.1021/jp401562d | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

■

REFERENCES

(1) Yanson, A. I.; Rubio Bollinger, G.; van den Brom, H. E.; Agraït, N.; van Ruitenbeek, J. M. Formation and Manipulation of a Metallic Wire of Single Gold Atoms. Nature 1998, 395, 783−785. (2) Smit, R. H. M.; Untiedt, C.; Yanson, A. I.; van Ruitenbeek, J. M. Common Origin for Surface Reconstruction and the Formation of Chain of Metal Atoms. Phys. Rev. Lett. 2001, 26, 266102. (3) Ohnishi, H.; Kondo, Y.; Takayanagi, K. Quantized Conductance through Individual Rows of Suspended Gold Atoms. Nature 1998, 395, 780−783. (4) Bahan, S. R.; Jacobsen, K. W. Chain Formation of Metal Atoms. Phys. Rev. Lett. 2001, 87, 266101. (5) Thijssen, W. H. A.; Strange, M.; aan de Brugh, J. M. J.; van Ruitenbeek, J. M. Formation and Properties of Metal-Oxygen Atomic Chains. New. J. Phys. 2008, 10, 033005. (6) Nakazumi, T.; Kiguchi, M. Formation of Co Atomic Wire in Hydrogen Atmosphere. J. Phys. Chem. Lett. 2010, 1, 923−926. (7) Smit, R. H. M.; Noat, Y.; Untiedt, C.; Lang, N. D.; van Hemert, M. C.; van Ruitenbeek, J. M. Measurement of the Conductance of a Hydrogen Molecule. Nature 2002, 419, 906−909. (8) Djukic, D.; van Ruitenbeek, J. M. Shot Noise Measurements on a Single Molecule. Nano Lett. 2006, 6, 789−793. (9) Kiguchi, M.; Konishi, T.; Hasegawa, K.; Shidara, S.; Murakoshi, K. Three Reversible States Controlled on a Gold Monoatomic Contact by the Electrochemical Potential. Phys. Rev. B 2008, 77, 245421. (10) Aika, K. Activation of Dinitrogen on Metal Surfaces. How Can We Break NN Bond? Hyomen (Rev. Top. Surf. Sci. Technol. Avantgarde) 1985, 23, 104−114. (11) Amorim, E. P. M.; da Silva, E. Z. Mechanochemistry in Cu Nanowires: N and N2 Enhancing the Atomic Chain Formation. Phys. Rev. B 2010, 82, 153403. (12) Amorim, E. P. M.; da Silva, E. Z. Effect of Light Impurities on the Electronic Structure of Copper Nanowires. Phys. Rev. B 2012, 85, 155407. (13) Kiguchi, M.; Tal, O.; Wohlthat, S.; Pauly, F.; Krieger, M.; Djukic, D.; Cuevas, J. C.; van Ruitenbeek, J. M. Highly Conductive Molecular Junctions Based on Direct Binding of Benzene to Platinum Electrodes. Phys. Rev. Lett. 2008, 101, 046801. (14) Kaneko, S.; Nakazumi, T.; Kiguchi, M. Fabrication of a WellDefined Single Benzene Molecule Junction Using Ag Electrodes. J. Phys. Chem. Lett. 2010, 1, 3520−3523. (15) Kiguchi, M.; Miura, S.; Hara, K.; Sawamura, M.; Murakoshi, K. Conductance of Single 1,4-Disubstituted Benzene Molecules Anchored to Pt Electrodes. Appl. Phys. Lett. 2007, 91, 053110. (16) Kiguchi, M.; Nakamura, H.; Takahashi, Y.; Takahashi, T.; Ohto, T. Effect of Anchoring Group Position on Formation and Conductance of a Single Disubstituted Benzene Molecule Bridging Au Electrodes: Change of Conductive Molecular Orbital and Electron Pathway. J. Phys. Chem. C 2010, 114, 22254−22261. (17) Kiguchi, M.; Nakazumi, T.; Hashimoto, K.; Murakoshi, K. Atomic Motion in Single H2 and D2 Molecule Junction Induced by Phonon Excitation. Phys. Rev. B 2010, 81, 045420. (18) Agraït, N.; Yeyati, A. L.; van Ruitenbeek, J. M. Quantum Properties of Atomic-Sized Conductors. Phys. Rep. 2003, 377, 81−279. (19) Wang, N.; Liu, H.; Zhao, J.; Cui, Y.; Xu, Z.; Ye, Y.; Kiguchi, M.; Murakoshi, K. Theoretical Investigation on the Electron Transport Path through the Porphyrin Molecules and Chemisorption of CO. J. Phys. Chem. C 2009, 113, 7416−7423. (20) Stokbro, K.; Taylor, J.; Brandbyge, M.; Mozos, J. L.; Ordejon, P. Theoretical Study of the Nonlinear Conductance of Di-Thiol Benzene Coupled to Au(1 1 1) Surfaces via Thiol and Thiolate Bonds. Comput. Mater. Sci. 2003, 27, 151. (21) Lee, Y. J.; Brandbyge, M.; Puska, M. J.; Taylor, J.; Stokbro, K.; Nieminen, R. M. Electron Transport through Monovalent Atomic Wires. Phys. Rev. B 2004, 69, 125409. (22) Atomistix ToolKit; QuantumWise A/S, www.quan-tumwise. com.

Figure 6. Distribution of the conductance enhancement and suppression in dI/dV spectra for the Pt−N2 system as a function of conductance and energy in the conductance range from 0.05 G0 to 1.2 G0.

theoretical study predicted that conductance enhancement was observed if the transmission probability was below 0.5, and conductance suppression was observed otherwise.30 This phenomena was experimentally observed in H2O31 and benzendithiole.32 The discrepancy might be explained by the difference in the number of the channels. In such a system like H2O molecule, electron transport occurs through a single channel, and the molecule−metal coupling is symmetric.30−32 On the other hand, in the Pt−N2 system, more than one channel contributed to the electron transport as we have shown above. The existence of multichannels may affect the shape of dI/dV curves. Theoretical calculation has predicted that conductance enhancement could be induced by nontransmitting d channels which coupled to the transmitting s channel via transverse modes in the Pt−H2 system.32

■

CONCLUSION The electronic conductance of platinum atomic contact in the nitrogen atmosphere has been investigated with experimental technique and theoretical calculation. A single N2 molecule was placed between platinum electrodes in parallel where N−N bond distance was elongated. The conductance of the single N2 molecular junction was 1 G0 which was close to the metal atomic contact. Two dominant channels originating from the coupling between N2 molecule and electrodes contributed to the electron transport.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (M.K.), [email protected] (J.Z.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by Grants-in-Aid for Scientific Research in Innovative Areas (No. 23111706) and Grant-in-Aid for Scientific Research (A) (No. 24245027) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) and the National Natural Science Foundation of China (No. 51071084 and 21273113). S.K. was supported by Grant-in-Aid for JSPS Research Fellow. D

dx.doi.org/10.1021/jp401562d | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

Article

(23) Brandbyge, M.; Mozos, J.-L.; Ordejón, P.; Taylor, J.; Stokbro, K. Density-Functional Method for Nonequilibrium Electron Transport. Phys. Rev. B 2002, 65, 165401. (24) Taylor, J.; Guo, H.; Wang, J. Ab initio Modeling of Quantum Transport Properties of Molecular Electronic Devices. Phys. Rev. B 2001, 63, 245407. (25) Frisch, M. J. et al. Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004. (26) Stipe, B. C.; Rezaei, M. A.; Ho, W. Single-Molecule Vibrational Spectroscopy and Microscopy. Science 1998, 280, 1732−1735. (27) Khotkevich, A. V. Modern State of Point Contact Spectroscopy of Electron-Phonon Interaction in Transition Metals. Physica B 1996, 218, 31−34. (28) Huber, K. P.; Herzberg, G. Molecular Spectra and Molecular Structure; IV. Constants of Diatomic Molecules; van Nostrand Reinhold: New York, 1979. (29) Bruch, L. W.; Nabar, R. P.; Mavrikakis, M. Weak Molecular Chemisorption of N(2)/Pt(111). J. Phys.: Condens. Matter 2009, 21, 264009. (30) Paulsson, M.; Frederiksen, T.; Ueba, H.; Lorente, N.; Brandbyge, M. Unified Description of Inelastic Propensity Rules for Electron Transport through Nanoscale Junctions. Phys. Rev. Lett. 2008, 100, 226604. (31) Tal, O.; Krieger, M.; Leerink, B.; van Ruitenbeek, J. M. ElectronVibration Interaction in Single-Molecule Junctions: From Contact to Tunneling Regime. Phys. Rev. Lett. 2008, 100, 196804. (32) Kim, Y.; Pietsch, T.; Erbe, A.; Belzig, W.; Scheer, E. Benzenedithiol: A Broad-Range Single-Channel Molecular Conductor. Nano Lett. 2011, 11, 3734−3738.

E

dx.doi.org/10.1021/jp401562d | J. Phys. Chem. C XXXX, XXX, XXX−XXX