1: \documentstyle[multicol,aps,prl,epsf]{revtex}

2: 

3: \pagestyle{empty}

4: \begin{document}

5: 

6: \draft

7: 

8: \title{Conductance of a single molecule anchored by an isocyanide substituent to gold 

9: electrodes}

10: 

11: \author{Manabu Kiguchi, Shinichi Miura, Kenji Hara, Masaya Sawamura, Kei Murakoshi}

12: 

13: \address{Division of Chemistry, Graduate School of Science, Hokkaido University, Sapporo, 060-0810, 

14: Japan}

15: 

16: 

17: \date{\today}

18: 

19: \maketitle

20: 

21: \begin{abstract}

22: 

23: The effect of anchoring group on the electrical conductance of a single molecule 

24: bridging two Au electrodes was studied using di-substituted (isocyanide (CN-), thiol (S-) or 

25: cyanide (NC-)) benzene. The conductance of a single Au/1,4-diisocyanobenzene/Au junction 

26: anchored by isocyanide via a C atom (junction with the Au-CN bond) was 

27: $3 \times 10 ^{-3} G_{0}$ ($2e^{2}/h$). 

28: The value was comparable to  $4 \times 10 ^{-3} G_{0}$ of a single Au/1,4-benzenedithiol/Au junction with 

29: the Au-S bond. The Au/1,4-dicyanobenzene/Au molecular junction with the Au-NC bond did 

30: not show well-defined conductance values. The metal-molecule bond strength was estimated 

31: by the distance over which the molecular junction was stretched before breakdown. The 

32: stretched length of the molecular junction with the Au-CN bond was comparable to that of the 

33: Au junction, indicating that the Au-CN bond was stronger than the Au-Au bond. 

34: \end{abstract}

35: 

36: \medskip

37: 

38: 

39: 

40: 

41: \begin{multicols}{2}

42: \narrowtext

43: 

44: Recently, molecular electronic devices have attracted attention, since these devices offer the 

45: possibility for extreme scale down and ultra-small energy consumption of logic units. 

46: Building the device with a single molecule demands an understanding of the electron 

47: transport properties through single molecule junctions connecting two metal electrodes. The 

48: electron transport characteristics through a single molecule have been studied using 

49: mechanically controllable break junctions \cite{1} and scanning tunneling microscopy (STM) \cite{2,3,4}. 

50: With regard to anchoring groups of the molecular junction, the Au-S bond is used to connect 

51: molecules to metal electrodes. It is because the Au-S bond is a strong bond, which leads to the 

52: effective hybridization of the molecular and metal orbitals. The effective hybridization of the 

53: orbitals is essential for the high conductance of the molecular junction. On the other hand, the 

54: high conductance of the molecular junction also requires the condition of energy matching 

55: between the Fermi level of metal electrodes and conduction orbital of the molecule \cite{5}. A strong 

56: metal-molecule bond and a smaller energy difference are essential for the molecular junction 

57: to show high conductance. The energy of the molecular orbital varies with the end group of 

58: the molecule. It is important to develop new end groups other than a thiol to enable more 

59: stable and higher conductance single molecular junctions.

60: 

61: Among various end groups, isocyanideend groups are attractive. Isocyanides bind to 

62: transition metals via a strong coordination bond by effective donation to metal and back 

63: donation from metals. The strength of the metal-CN bond is comparable to that of the metal-S 

64: bond. Because of this effective interaction between the isocyanide group and metals, air stable 

65: isocyanide SAM films are formed on Au \cite{6}, as with the case of thiol SAMs \cite{7}. This strong 

66: metal-CN bond may contribute to increase the conductance of the junction. Recent theoretical 

67: calculations show promise for a molecular junction based on an isocyanide group \cite{8}. The 

68: conductance of the molecular junction binding a benzene molecule to Au atoms by isocyanide 

69: coordination (the molecular junction with the Au-CN bond) is close to that of the molecular 

70: junction with the Au-S bond. In the case of Ni electrodes, the conductance of molecular 

71: junctions with isocyanide end groups is higher than those of thiols \cite{8}. Despite these interests, 

72: the conductance of a single molecule with isocyanide end groups has not been studied 

73: adequately \cite{9}. 

74: 

75: In the present study, we have measured the electrical conductance of several 

76: di-substituted (-S, -NC, -CN) benzenes bridging two Au electrodes. As for the di-substituted 

77: benzene, there are only a few reports on the conductance of 1,4-benzenedithiol bridging two 

78: Au electrodes \cite{1,3,4}. Based on the bond strength of the anchoring group \cite{10}, 

79: the conductance of a single molecule was discussed using a simple tunneling model \cite{5}.

80: 

81: A sharp tip was used, comprising a Au wire (diameter $\sim$0.25 mm, $>$99.9$\%$). The substrate 

82: was Au(111), prepared by a flame annealing and quenching method \cite{11}. The measurements 

83: were carried out in tetraethyleneglycol dimethyl ether (tetraglyme) or toluene solutions since 

84: isocyanides could be hydrolyzed under aqueous acidic or basic conditions \cite{12}. The solution 

85: contained 1mM concentrations of 1,4-diisocyanobenzene, 1,4-benzenedithiol, or 

86: 1,4-dicyanobenzene. A Au STM tip was repeatedly moved into and out of contact with a Au 

87: (111) substrate at a rate of 50 nm/s in the solution. Each molecule was terminated with two 

88: isocyanide, thiol, or cyanide end groups bridging the tip and substrate electrodes via the 

89: respective anchoring groups, as shown in Fig. 1. Conductance was measured during the 

90: breaking process under an applied bias of 20 mV between the tip and substrate. All statistical 

91: data was obtained from a large number (over 1000) of individual conductance traces.

92: 

93: Figure 2 (a,b) show the typical conductance traces of Au point contacts broken in the  

94: tetraglyme solution containing 1 mM of 1,4-diisocyanobenzene. The conductance decreased 

95: in a stepwise fashion with each step occurring at an integer multiple of 

96: $2 \times 10 ^{-3} G_{0}$ $\sim$ $3 \times 10 ^{-3} G_{0}$. 

97: The last plateau before the contact break showed a value of $2 \times 10 ^{-3} G_{0}$. 

98: The corresponding 

99: histogram shows a feature at  $3 \times 10 ^{-3} G_{0}$($\pm 1 \times 10 ^{-3} G_{0}$ ). 

100: These steps in the conductance trace 

101: likely originate from the formation of a stable, single Au/1,4-diisocyanobenzene/Au junction, 

102: possibly binding a benzene molecule to Au atoms by isocyanide coordination via C atoms 

103: (molecular junction with the Au-CN bond). The conductance behavior in the absence of 

104: 1,4-diisocyanobenzene molecules was studied. Neither steps nor peaks were observed in the 

105: same conductance regime (see Fig. 2 (a,b)). The effect of the solvent was also checked using 

106: toluene. The inset of Fig. 2 (b) shows the conductance histogram observed in toluene solution 

107: containing 1 mM of 1,4-diisocyanobenzene. The conductance histogram shows a feature near 

108: $3 \times 10 ^{-3} G_{0}$, as is the case with tetraglyme. Based on these experimental results, the 

109: conductance of a single molecular junction with the Au-CN bond was determined to be 

110: $3 \times 10 ^{-3} G_{0}$($\pm 1 \times 10 ^{-3} G_{0}$ ).

111: 

112: The conductance of 1,4-dicyanobenzene and 1,4-benzenedithiol is discussed. In the 

113: Au/1,4-dicyanobenzene/Au junction (molecular junction with the Au-NC bond), a molecule 

114: binds to Au atoms via the N atom of the cyanide group, while a molecule binds to Au atoms 

115: via the S atom in the Au/1,4-benzenedithiol/Au molecular junction (molecular junction with 

116: the Au-S bond). Figure 2 (c-f) show the conductance traces and histograms of Au point 

117: contacts broken in the tetraglyme solution containing 1mM of 1,4-dicyanobenzene or 

118: 1,4-benzenedithiol. As for 1,4-dicyanobenzene, small conductance steps were observed in the 

119: conductance trace. The difference in conductance between the steps was smaller than 

120: $5 \times 10 ^{-4} G_{0}$. The conductance histogram does not show a clear feature above 

121: $1 \times 10 ^{-3} G_{0}$. Thus, 

122: the conductance of the molecular junction with the Au-NC bond would be smaller than 

123: $1 \times 10 ^{-3} G_{0}$. Although another possibility that the molecule could not make stable and 

124: reproducible junctions should be also considered for the interpretation on this conductance 

125: behavior, estimated value may reflect relatively small conductivity of the junction. For 

126: 1,4-benzenedithiol, steps occurring at an integer multiple of $4 \times 10 ^{-3} G_{0}$ 

127: were observed in the conductance traces, and the conductance histogram shows a feature 

128: near $4 \times 10 ^{-3} G_{0}$. 

129: Thus, the conductance of a single molecular junction with the Au-S bond was determined to be   

130: $4 \times 10 ^{-3} G_{0}$($\pm 1 \times 10 ^{-3} G_{0}$ ). 

131: The conductance of a single molecular junction with the Au-S bond was also 

132: studied by other groups, and it ranges from  $1.1 \times 10 ^{-2} G_{0}$ \cite{13} 

133: to $4.1 \times 10 ^{-4} G_{0}$ \cite{1}. Our result of 

134: $4 \times 10 ^{-3} G_{0}$ falls within this range of these values. 

135: Since the different conductance values for 

136: the same molecular junction by orders of magnitude may be due to the different experimental 

137: conditions, it is difficult to compare the conductance of single molecular junctions of different 

138: molecules measured under different conditions. In the present case, the conductance of a 

139: single molecular junction is discussed based on our results measured under the same 

140: condition. Statistical analysis of the present data showed that the conductance values of a 

141: single molecular junction measured under the same condition could be determined to be in a 

142: precision of 30 $\%$. Our results showed that the conductance of the molecular junction with the 

143: Au-CN bond was comparable to the junction with the Au-S bond. The conductance of the 

144: molecular junction with the Au-NC bond was much smaller than the junction with the Au-S 

145: and Au-CN bonds.

146: 

147: Our experimental results on the conductance of a single molecular junction is discussed 

148: with a simple tunneling model and theoretical calculation results. In the simple tunneling 

149: model \cite{5}, the conductance of a single molecular junction is given 

150: by $A(\beta G \rho)^{2}$, where $A$, $\beta$, $G$, $\rho$

151: are the constant, hopping integral between the molecule and metal, Green's function, and 

152: local density of states (LDOS) at the Fermi energy of the contact metal atom, respectively.   

153: $\beta$ increases with the bond strength. $G$ decreases with the energy difference between the Fermi 

154: level and conduction orbital. Therefore, a strong metal-molecule bond (large $\beta$) and small 

155: energy difference (large $G$) are essential for a molecular junction with high conductance. In 

156: the case of the molecular junction with the Au-CN bond, the bond strength and the energy 

157: difference between the Fermi level and conduction orbital are similar to that with the Au-S 

158: bond (similar $\beta$ and $G$) \cite{14,15}. Thus, the conductance of the molecular junction with the 

159: Au-CN bond would be close to that of the Au-S bond. Theoretical calculations show that the 

160: molecular junction with the Au-S bond and Au-CN bond have similar conductance values \cite{8}. 

161: Our experimental results agree well with those theoretical estimates. On the other hand, the 

162: strength of the Au-NC bond is much smaller than that of the Au-CN or Au-S bonds, which 

163: yields a small for the Au-NC bond. Since the energy difference between the Fermi level and 

164: conduction orbital of 1,4-dicyanobenzene is close to that of 1,4-diisocyanobenzene \cite{13,14}, the 

165: conductance of the molecular junction with the Au-NC bond would be much smaller than that 

166: with the Au-CN bond due to a small $\beta$. Theoretical calculations shows that the conductance 

167: of the molecular junction with the metal-NC bond is three times smaller than that with the 

168: metal-S bond \cite{14,15}. 

169: 

170: Relatively strong binding between Au and C atoms in Au-CN binding was also 

171: demonstrated by conductance traces. Figure 3 (a,b) show the typical conductance traces of Au 

172: point contacts broken in the presence of 1,4-diisocyanobenzene. The length of the last 

173: conductance plateau of the molecular junction (0.16 nm in Fig. 3 (b)) was comparable to that 

174: of the Au junction (0.12 nm in Fig. 3 (a)). Since the last plateau corresponds to the single 

175: molecular or metal atom junction \cite{11}, the length of the last plateau is the distance over which a 

176: molecular or metal junction can be stretched before breakdown (breakdown distance). To 

177: evaluate characteristics of the breakdown distance quantitatively, the distributions of the 

178: breakdown distance of the Au junctions and molecular junction are shown in Fig. 3 (c). The 

179: breakdown distance was defined as the distance between the points at which the conductance 

180: dropped below 1.2 $G_{0}$ and 0.8 $G_{0}$ for the Au junctions, and  

181: $3.2 \times 10 ^{-3} G_{0}$ and  $2.0 \times 10 ^{-3} G_{0}$ 

182: for the molecular junctions, respectively. The average length for the Au junctions and the 

183: molecular junction was 0.078 ($\pm$0.008) nm and 0.073 ($\pm$0.007) nm, respectively. Two 

184: breakdown distances agreed well with each other. 

185: 

186: The breakdown can take place at one of the three bonds, Au-CN, Au-Au, or an 

187: intramolecular bond in the Au/1,4-diisocyanobenzene/Au molecular junction. The respective 

188: bonds could show different breakdown distances because of the difference in the bond 

189: strength. Comparable breakdown distances indicates that breakdown in the molecular junction 

190: took place at the Au-Au bond rather than the Au-CN bond, which is compelling proof that the 

191: Au-CN bond was stronger than the Au-Au metal bond in the molecular junction with the 

192: Au-CN bond. In the case of the Au/1,4-benzenedithiol/Au molecular junction, the breakdown 

193: distance (0.10 nm in Fig. 2(c)) was also comparable to that of the Au junction, indicating that 

194: the Au-S bond was also stronger than the Au-Au bond. The relatively short length of the 

195: plateau (0.02 nm see Fig. 2 (e)) of the Au/1,4-dicyanobenzene/Au junction indicates that the 

196: Au-NC bond was weaker than the Au-Au bond. Similar behavior was also observed in the 

197: system of 1,8-octanedithiol bonded to two Au electrodes by force measurement using atomic 

198: force microscopy (AFM) \cite{16}. The force required to break the molecular junctions was close to 

199: that of the Au junctions, indicating that the strength of the Au-S bond is higher than that of the 

200: Au-Au metal bond. 

201: 

202: A well-defined molecular junction with a strong Au-CN bond was fabricated in the 

203: present study. As for the di-substituted benzene, the conductance of the molecular junction 

204: with the Au-CN bond was comparable to that with the Au-S bond. The energy of the 

205: molecular orbital varies with the linker group and the end group of the molecule. In the case 

206: of 1,4-diisocyanobenzene, the linker and end group are defined as benzene ring (C$_{6}$H$_{4}$) and 

207: isocyanide (CN), respectively. In some linker groups, the energy difference between the Fermi 

208: level of the metal electrodes and conduction orbital of the molecule could be smaller for 

209: isocyanide end groups than thiols. Since the bond strength of the metal-CN bond is 

210: comparable to that of the metal-S bond, the conductance of the molecular junction with the 

211: metal-CN bond could be higher than that with the metal-S bond in some combinations of the 

212: metals and molecules. In developing an electronic device with a single molecule, it will be 

213: important to study electron transport through a single molecule with a high degree of 

214: functionality, such as in the photoswitching effect and single molecule magnetism. By using 

215: the best end group for the individual molecule, the performance of the device would be 

216: drastically improved. The isocyanide end group would be a suitable candidate.

217: 

218: In conclusion, we studied the conductance of a single 1,4-diisocyanobenzene, 

219: 1,4-dicyanobenzene or 1,4-benzenedithiol molecule bridging two Au electrodes. The 

220: conductance of the molecular junction with the Au-CN bond was comparable to that with the 

221: Au-S bond. The plateau length analysis showed that the Au-CN coordination bond was 

222: stronger than Au-Au metal bond. The isocyanide end group is a promising alternative to the 

223: thiol. This work was partially supported by a Grant-in-Aid for Scientific Research A (No. 

224: 16205026) and Grant-in-Aid for Scientific Research on Priority Areas (No. 17069001) from 

225: MEXT.

226: 

227: \begin{references}

228: \bibitem{1} M. A. Reed, C. Zhou, C. J. Muller, T. P. Burgin, J. M. Tour, Science {\bf 278}, 252 (1997). 

229: \bibitem{2} B. Xu and N.J. Tao, Science {\bf 301}, 1221 (2003).

230: \bibitem{3} J. Ulrich, D. Esrail, W. Pontius, L. Venkataraman, D. Millar, and L. H. Doerrer, J. Phys. 

231: Chem. B {\bf 110}, 2462(2006). 

232: \bibitem{4} X. Xiao, B. Xu, and N. J. Tao, Nano Lett. {\bf 4}, 267 (2004).

233: \bibitem{5} T. Tada and K. Yoshizawa, CHEMPHYSCHEM {\bf 12}, 1035 (2002).

234: \bibitem{6} D.L. DuBose, R.E. Robinson, T.C. Holovics, D.R. Moody, E. C. Weintrob, C.L. Berrie, and 

235: M.V. Barybin, Langmuir {\bf 22}, 4599 (2006).

236: \bibitem{7} K. Hara, K. Iwahashi, Y. Kanamori, S. Naito, S. Takakusagi, K. Uosaki, and M. Sawamura, 

237: Chem. Lett. {\bf 35}, 870 (2006).

238: \bibitem{8} J.M. Seminario, CE. De La Cruz, and PA Derosa, J. Am. Chem. Soc. {\bf 123}, 5616 (2001).  

239: \bibitem{9} J. Chen, L.C. Calvet, M.A. Reed, D.W. Carr, D.S. Grubisha, D.W. Bennett, Chem. Phys. 

240: Lett. {\bf 313}, 741 (1999).

241: \bibitem{10} S. K. Chastain and W. R. Mason, Inorg. Chem. {\bf 21}, 3717 (1982).

242: \bibitem{11} M. Kiguchi, T. Konishi, and K. Murakoshi, Phys. Rev. B {\bf 73}, 125406 (2006).

243: \bibitem{12} I. D. Cunningham, G. J. Buist and S. R. Arkle, 

244: J. Chem. Soc., Perkin Trans. {\bf 2}, 589 (1991).

245: \bibitem{13} J. He, O. Sankey, M. Lee, N. Tao, X. Lic and S. Lindsay, Faraday Discussion {\bf 131}, 145 

246: (2006).

247: \bibitem{14} N. D. Lang and Ph. Avouris, Phys. Rev. B {\bf 64}, 125323 (2001).

248: \bibitem{15} Y. Xue and M. A. Ratner, Phys. Rev. B {\bf 69}, 85403(2004).

249: \bibitem{16} B. Xu, X. Xiao, and N. J. Tao, J. Am. Chem. Soc. {\bf 125}, 16164 (2003).

250: \end{references}

251: 

252: \begin{figure}

253: \begin{center}

254: \leavevmode\epsfxsize=70mm \epsfbox{fig1.eps}

255: \caption{

256: Schematic illustration of a molecule junction, in which a single molecule is bonded to 

257: two Au electrodes.}

258: \label{fig1}

259: \end{center}

260: \end{figure}

261: 

262: \begin{figure}

263: \begin{center}

264: \leavevmode\epsfxsize=85mm \epsfbox{fig2.eps}

265: \caption{

266: Typical conductance trace and histogram of Au point contacts broken in the tetraglyme 

267: solution consisting of (a,b) 1,4-diisocyanobenzene, (c,d) 1,4-dicyanobenzene and (e,f) 

268: 1,4-benzenedithiol. The dotted-line shows the result in the absence of molecules. The inset of 

269: (b) shows the conductance histogram in the toluene solution containing 1 mM of 

270: 1,4-diisocyanobenzene.}

271: \label{fig2}

272: \end{center}

273: \end{figure}

274: 

275: \begin{figure}

276: \begin{center}

277: \leavevmode\epsfxsize=70mm \epsfbox{fig3.eps}

278: \caption{Typical conductance trace of Au junctions (a) and 1,4-diisocyanobenzene junctions in 

279: tetraglyme solution containing 1 mM of 1,4-diisocyanobenzene. (c) The distribution of 

280: lengths for the last conductance plateau for the Au junctions (box), and molecular junctions 

281: (star).}

282: \label{fig3}

283: \end{center}

284: \end{figure}

285: 

286: 

287: \end{multicols}

288: \end{document} 

289: 

290: