- Email: [email protected]
Dependence of tunneling current through a single molecule of phenylene oligomers on the molecular length
Ultramicroscopy 97 (2003) 19–26
Dependence of tunneling current through a single molecule of phenylene oligomers on the molecular length Satoshi Wakamatsu, Shintaro Fujii, Uichi Akiba, Masamichi Fujihira* Department of Biomolecular Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8501, Japan Received 8 July 2002; received in revised form 17 December 2002
Abstract The electrical properties of single phenylene oligomers were studied in terms of the dependence of the tunneling current on the length of the oligomers using self-assembling techniques and scanning tunneling microscopy (STM). It is important to isolate single molecules in an insulating matrix for the measurement of the conductivity of the single molecule. We demonstrate here a novel self-assembled monolayer (SAM) matrix appropriate for isolation of the single molecules. A bicyclo[2.2.2]octane derivative was used for a SAM matrix, in which the single molecules were inserted at molecular lattice defects. The isolated single molecules of phenylene oligomers inserted in the SAM matrix were observed as protrusions in STM topography using a constant current mode. We measured the topographic heights of the molecular protrusions using STM and estimated the decay constant, b; of the tunneling current through the single phenylene oligomers using a bilayer tunnel junction model. r 2003 Elsevier Science B.V. All rights reserved. Keywords: Phenylene oligomer; Bicyclo[2.2.2]octane derivative; Self-assembled monolayer; Scanning tunneling microscopy
1. Introduction Electronic tunneling through a single molecule is one of the most fundamental and important properties for molecular devices. It has been investigated with molecular resolution using scanning tunneling microscopy (STM) [1–7]. STM can form a bilayer junction where the tunneling electrons flow through a self-assembled monolayer (SAM) and a vacuum gap between a metal tip and a metal substrate surface. In the case of STM, it is difficult to know the distance between the tip and *Corresponding author. Tel.: +81-45-924-5784; fax: +81-45924-5817. E-mail address: [email protected] (M. Fujihira).
the sample surface because the thickness and the conductivity of the sample are convoluted. Bumm et al. estimated the conductance of alkanethiol SAMs from STM data using a bilayer tunnel junction model [3]. The tunneling probability of the composite bilayer junction is the product of the tunneling probabilities of the individual layers, Tt ¼ Tfilm Tgap ; where Tt is the total tunneling probability, Tfilm and Tgap are the tunneling probabilities of a SAM and a vacuum gap, respectively. On the other hand, conducting atomic force microscopy (AFM) enabled us to contact the AFM tip with the sample surface under a controlled load. This method has been used widely to characterize the electrical properties of SAMs [8–16].
0304-3991/03/$ - see front matter r 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3991(03)00026-3
20
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
Recent self-assembling techniques allowed us to measure electronic tunneling through a single molecule embedded in an insulating matrix using STM and conducting AFM [1,2,4–7,9,11,12,16]. Immersion of a preformed host SAM in a dilute solution of guest molecules resulted in isolating them in the host SAM. The isolated molecules were considered to be inserted at defects of the preformed matrices, such as step edges of the Au(1 1 1) substrate surface and structural domain boundaries of the host SAMs [1,2]. Isolated conductive molecules appear as protrusions in an STM topographic image using a constant current mode. In most cases, an alkanethiol SAM was used as the host matrix. Bumm et al. inferred that many of the like features at structural domain boundaries of the alkanethiol SAM matrix are individual molecules, because access for insertion is limited at the structural domain boundaries surrounded by the close-packed film [1,2]. Cui et al. reported electronic conduction through an isolated single molecule sandwiched between an Au nanoparticle and an Au(1 1 1) surface, in which either end of the isolated molecule was chemically bonded [11]. They observed that I V measurements on the Au nanoparticles binding to the isolated molecules produced five distinct families of curves. The curves corresponded to multiples of a fundamental curve, which was ascribed to the I V curve of the single molecule. This implies that the isolated molecules in the SAM matrix are not always the single molecules. Therefore, when we used the alkanethiol SAM as the host matrix, it is difficult to judge whether individual or bundled molecules were isolated at the defect sites. We used a bicyclo[2.2.2]octane derivative (BCO) SAM as the host matrix [17] and a series of phenylene oligomers, C6H5(C6H4)m1CH2SH (m ¼ 1; 2; 3), as the conductive guest molecules. The BCO SAM has unique properties that typical alkanethiol SAMs do not have. First, the BCO SAM [17] has no structural domain boundary [1,2]. Second, the BCO SAM has molecular lattice defects, which are defects of single molecules described later. The size of the molecular lattice defect is considered to be that of the single BCO molecule. Because the BCO SAM matrix has no structural domain boundary, inserted guest
molecules exist at the molecular lattice defects in the matrix, so that single guest molecules are certainly isolated in the BCO SAM matrix. In this study, we investigated electronic tunneling through isolated single molecules of phenylene oligomers embedded in the BCO SAM matrices using STM. We succeeded in quantitative measurement for electronic tunneling through the single molecule of phenylene oligomers. We estimated the dependence of the tunneling current on the molecular length of phenylene oligomers, i.e. the decay constant b: This was accomplished by measuring the heights of protrusions of individual phenylene oligomers in STM images and by analyzing the heights using the bilayer tunnel junction model.
2. Experimental We prepared three types of phenylene oligomers (benzenemethanethiol; BMT, 4-biphenylmethanethiol; BPMT, and [1,10 :40 ,100 -terphenyl]-4methanethiol; TPMT) and a bicyclo[2.2.2]octane derivative (bis(bicyclo[2.2.2]octylmethyl) disulfide; BCO) shown in Fig. 1. BMT was purchased from Tokyo Chemical Industry. BPMT, TPMT, and BCO were synthesized according to the literature [17–19]. Au(1 1 1) substrates were prepared by thermal evaporation of Au onto freshly cleaved
BMT SH
BPMT SH
TPMT SH
BCO
S
S
Fig. 1. Chemical structures of BMT, BPMT, TPMT, and BCO.
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
s
s
s
s
s
s
s
Au (111) (a)
s
s
s
s
s
s
Au (111) (b)
s
21
s
s
s
s
Au (111) (c)
Fig. 2. Schematic representations of three types of SAMs in which single phenylene oligomers of (a) BMT, (b) BPMT, and (c) TPMT are isolated in BCO SAM matrices on Au[1 1 1], respectively.
Fig. 3. STM images of three type of SAMs on Au[1 1 1]. All STM images were obtained with a tungsten tip using a constant current mode (30 pA) at a sample bias voltage of +1.0 V. (a) An STM image of the BCO SAM (15 15 nm2 in size). Single Au atom deep pits and a hexagonally close-packed lattice were observed. The BCO SAM has no structural domain boundary. Molecular lattice defects are highlighted by circles. (b) An STM image of a typical alkanethiol SAM composed of decanethiol (15 15 nm2 in size). The structural domain boundaries and the single Au atom deep pits were observed. (c) An STM image of a decanethiol SAM matrix in which guest BPMT molecules were inserted (40 40 nm2 in size). Bright spots were assigned as the inserted BPMT molecules.
mica that was heated in vacuum to 300 C [7]. The substrates were removed from the chamber and immersed immediately in thiol solutions. Three types of SAMs (Figs. 2(a)–(c)) containing isolated single molecules were prepared in a twostep process [1]. First, SAM matrices of BCO were prepared on the Au substrates by immersion for 24 h or longer in a 0.1 mM ethanol solution of BCO. Second, after thorough rinsing of the SAM matrices with ethanol, the BCO SAM matrices were immersed for a few minutes in a 1 mM ethanol solution of BMT or BPMT, or in a 0.1 mM tetrahydrofuran (THF) solution of TPMT. Then, the resulting two-component SAMs were rinsed thoroughly with the same solvents as those used for the second step, and dried. STM images were obtained with a tungsten tip using a constant current mode (30 pA) at the sample bias voltage of +1.0 V under an ultra high vacuum (UHV) condition (o5 108 Pa) (a JEOL
JSPM-4500S/TM-51021 UHV scanning tunneling microscope). Current vs. separation between the tip and the sample (I S) characteristics were measured in UHV by first setting an initial gap separation with an initial tunneling current (30 pA), second disconnecting the feedback loop of the z piezoelectric regulation, and finally moving the tip towards the sample surface.
3. Results and discussion Fig. 3(a) shows a constant-current STM image of a 15 15 nm2 area of the BCO SAM on Au[1 1 1]. All images in Fig. 3 were recorded with a sample bias voltage of +1.0 V and a tunneling current of 30 pA. In the same way as in an n-alkanethiol SAM (Fig. 3(b)), single Au atom deep pits and a hexagonally close-packed lattice are seen in the BCO SAM shown in Fig. 3(a) [17].
22
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
A typical alkanethiol SAM matrix contains a variety of characteristic defect sites, such as the structural domain boundaries [1,2,20] and the single Au atom deep pits [20], as shown in Fig. 3(b). On the other hand, owing to the bulkiness of the BCO moieties, they are expected to be oriented perpendicularly to the substrate surface. Thus the formation of domain structure due to the tilt of adsorbates can be avoided, so that the structural domain boundaries were not observed in Fig. 3(a) [17]. In addition, molecular lattice defects, which are defects of single BCO molecules and highlighted by the circles in Fig. 3(a), were clearly observed on the BCO SAM surface. The molecular lattice defects are unique to the BCO SAM. They were not observed on the typical alkanethiol SAM surfaces as shown in Fig. 3(b). Bumm et al. suggested that exposure of one component SAM to a solution containing guest molecules resulted in the molecules inserted in SAM defect sites [1]. At these sites, the guest molecules can approach to the underlying gold atoms so that the sulfur head groups can chemisorb to the gold surface. Therefore, when the guest molecules are more conductive than the molecules forming the SAM matrix, STM images show that the guest molecules protrude from the SAM matrix, even if the physical height of the guest molecules is lower than that of the SAM matrix as shown in Fig. 3(c) [2,4,7]. In Fig. 3(c),
the guest molecule BPMT was inserted in a decanethiol SAM matrix. The length of the BPMT molecules is considered to be slightly shorter than that of the decanethiol SAM matrix. The bright spots in Fig. 3(c) were assigned as the inserted BPMT molecules. However, because the STM tip–sample interaction convolves electronic and physical structure [1,2], it is difficult to judge from the STM images whether individual or bundled molecules were isolated at the defect sites. We used the BCO SAM matrix as a host to isolate phenylene oligomers in the present study. As described above, the BCO SAM has no structural domain boundary so that the inserted phenylene oligomers exist at the substrate step edges, at the edge of the single Au atom deep pits, and the molecular lattice defects. Among them, only each molecular lattice defect of the last ones can always allow a single molecule to be isolated due to the size of the molecular lattice defect. Namely, single guest molecules (not bundles of molecules) are isolated at the molecular lattice defects in the BCO SAM matrix. STM images of the two-component SAMs containing the single molecules of BMT, BPMT, and TPMT are shown in Figs. 4(a)–(c), respectively. The images shown in Fig. 4 were recorded with the sample bias of +1.0 V and the tunneling current of 30 pA. Some bright spots were observed in each STM image. We assigned the bright spots in a terrace as the single molecules of phenylene
Fig. 4. STM images (20 20 nm2 in size) and topographic cross sections of the two-component SAMs containing the single molecules of (a) BMT, (b) BPMT, and (c) TPMT. All STM images were obtained with a tungsten tip using a constant current mode (30 pA) at a sample bias voltage of +1.0 V. Bright spots were assigned as single molecules. The topographic cross sections were extracted from the STM images on scans indicated by black lines.
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
oligomers inserted at the molecular lattice defects of the BCO SAM matrices. The difference in size and shape of the bright spots can be ascribed to thermal motion of the isolated single phenylene oligomers. Topographic cross sections extracted from the above three STM images along the black lines are also shown in Figs. 4(a)–(c), respectively. These lines pass over the isolated molecules protruding from the matrices. The different heights of the protrusions for the same phenylene oligomer were observed in each STM image, although they were isolated in the same SAM matrices. When we repeated imaging the same area, the observed height of one protrusion located in the same position was changed also with time. We considered that these phenomena are caused by change in the tilt of the molecular axis of the isolated single phenylene oligomer [21]. In Fig. 5, histograms of values of the STM topographic height from the surroundings show that the heights of protrusions for (a) BMT, (b) BPMT, and (c) TPMT range over 0.10–0.22,
Fig. 5. Histograms of values of the STM topographic heights of protrusions from the surroundings for (a) BMT, (b) BPMT, and (c) TPMT. Heights of the protrusions of BMT, BPMT, and TPMT range from 0.10 to 0.22 nm, from 0.14 to 0.42 nm, and from 0.20 to 0.68 nm, respectively.
23
Fig. 6. A representative I2S characteristic on the BCO SAM surface. A change in slope at z ¼ 0:2470.02 nm was considered to be caused by the contact of the tip with the top of the SAM. A linear fit in the vacuum region gives a line with the decay constant of 11.470.5 nm1.
0.14–0.42, and 0.20–0.68 nm, respectively. All of the data in the histograms, i.e. counts vs. topographic height, were extracted from more than 100 images, which contain the bright spots corresponding to the single molecules, for each phenylene oligomer. In Fig. 4, the cross sections were drawn across the highest protrusions among observed ones for each guest molecule. Here, it is reasonable that the isolated single molecules corresponding to the highest protrusions are tilted perpendicularly to the substrates. The highest protrusions of BMT, BPMT, and TPMT are 0.22, 0.42, and 0.68 nm, respectively. Differences in the height of the single molecular protrusions perpendicular to the substrates are 0.20 nm between BMT and BPMT, and 0.26 nm between BPMT and TPMT. Fig. 6 shows a representative I S characteristic above the BCO SAM surface. A near exponential dependence of the current as a function of the z-position was observed. In this measurement, the initial z-position (z ¼ 0 nm) was set by feed back control under the constant current
24
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
mode STM with 30 pA. Then, the tip was moved towards the sample surface under the fixed sample bias voltage at +1.0 V after disconnecting the feedback. When the tip moved close to the SAM, a sudden change in slope was observed at z ¼ 0:2470.02 nm. We assume that this change in slope was caused by the contact of the tip with the top of the SAM. A fitted straight line for the plots in a vacuum region was drawn on the graph. The decay constant of the tunneling current through the vacuum above the BCO SAM surface was 11.470.5 nm1. Using these results described above, we estimated the dependence of the tunneling current on the length of phenylene oligomers at the same sample bias voltage of +1.0 V. We assumed that contribution by the tunneling current through space (chain-to-chain coupling) in the SAM is negligible [22]. We obtained the value of b for the tunneling current through the single phenylene oligomers using the bilayer tunnel junction model according to the following procedure. (i) In Fig. 7,
Fig. 7. A semi-logarithmic graph of the tunneling current vs. separation between the tip and the sample. The I2S characteristic shown in Fig. 6 was replotted on this graph. The contact point of the tip with the BCO SAM was defined newly as the position at z ¼ 0 nm. Open squares are the STM topographic heights (see text) at the tunneling current of 30 pA above the BCO SAM matrix for the isolated single phenylene oligomers of BMT, BPMT, and TPMT. The height of the tip on the BCO SAM matrix under 30 pA was also plotted as an open square. Filled circles are the supposed contact points of the tip with the isolated single phenylene oligomers of BMT, BPMT, and TPMT, respectively. A linear fit of the filled circles gives a b value of 5.370.2 nm1.
a semi-logarithmic graph of the tunneling current vs. separation between the tip and the BCO SAM surface was plotted. The I2S characteristic shown in Fig. 6 was replotted on this graph. The contact point of the tip with the BCO SAM surface was defined newly as the position at z ¼ 0 nm. (ii) The measured heights of the isolated single phenylene oligomers of BMT, BPMT, and TPMT (the open squares) above the BCO SAM matrix in Fig. 4 were plotted on the dotted line corresponding to the constant tunneling current of 30 pA. The height of the tip on the BCO SAM matrix under 30 pA was also plotted. Namely, the STM topographic heights of the isolated single phenylene oligomers of BMT, BPMT, and TPMT above BCO were 0.22, 0.42 and 0.68 nm, respectively. If we could measure I2S curves above the isolated single molecules, we would expect that the I2S curves in vacuum fall on thin solid lines in Fig. 7 with the same decay constant as 11.4 nm1 above the BCO SAM matrix. It was, however, difficult to measure the I2S curves above the isolated single molecules because of the drift of the tip position. (iii) The molecular length of one phenylene unit is B0.43 nm [14,23,24]. If phenylene oligomers are oriented perpendicularly to the surface, the physical height of one phenylene unit is 0.43 nm. The physical height of BMT is determined to be slightly lower (B0.02 nm) than that of BCO from molecular models. The lines at each terminal z-position of phenylene oligomers, BMT, BPMT, and TPMT, were drawn as the dashed lines. (iv) The filled circles are plotted at the points of intersections between the supposed I2S curves in vacuum and the dashed lines for each phenylene oligomer. The filled circles are the points, where the tip is just in contact with the end of the isolate single phenylene oligomers. Thus the tunneling currents of these points correspond to what would be measured at the contacts. The linear fit (a thick solid line) of the filled circles gives a b value of 5.370.2 nm1. This value of b is the decay constant of tunneling current through the single molecule of phenylene oligomers at the sample bias voltage of +1.0 V. In our previous study [15], we measured the I2V characteristics of the phenylene oligomer SAMs using conducting AFM with Au-coated tips
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
modified chemically by decanethiol. The value of b ¼ 5:570.6 nm1 was obtained at the sample bias voltage of +1.0 V. The b value of 5.3 nm1 in the present work is slightly lower than that of 5.5 nm1 obtained using conducting AFM. The difference between these two values of b might be due to the difference in the barrier heights for electronic tunneling, because we used the tungsten STM tips in this study and the Au-coated AFM tips in the previous work. The work functions of Au (polycrystalline) and W (polycrystalline) are 5.1 and 4.5 eV, respectively [25,26]. The difference of the work functions causes the difference in the barrier heights for electronic tunneling. In our case, because the tip was served as the electron donor electrode due to the negative tip bias voltage, the lower work function of the tip electrode explains the lower barrier height for electron tunneling. However, in our previous study, we assumed that hole tunneling was dominant through these phenylene oligomer SAMs using bilayer junctions formed by conducting AFM with the chemically modified Au-coated tips [15]. In this present STM system, it is difficult to understand the mechanism of the hole tunneling through a vacuum gap. Therefore, here we can only suggest that the value of b ¼ 5:3 nm1 is reasonable, compared to that of b ¼ 5:5 nm1 measured by conducting AFM for the SAMs of phenylene oligomers. We are now investigating the effect of the work functions of metal tips on STM and conducting AFM.
4. Conclusions We made the novel host BCO SAMs, in which the single phenylene oligomers were inserted as isolated single molecules. The BCO SAMs have no structural domain boundary so that the guest molecules are inserted at the molecular lattice defects of the BCO SAM matrices. These SAMs allowed us to investigate the electrical properties of the single isolated phenylene oligomers using STM. In this way, we succeeded in estimating the dependence of the tunneling current through the single phenylene oligomers on the molecular length by STM. The reasonable value of
25
b ¼ 5:3 nm1 for the single phenylene oligomers at the sample bias voltage of +1.0 V was obtained from the STM measurement of the molecular protrusion.
Acknowledgements This work was supported by Grant-in-Aid for Creative Scientific Research on ‘‘Devices on molecular and DNA levels’’ (No. 13GS0017) from the Ministry of Education, Science, Sports, and Culture of Japan.
References [1] L.A. Bumm, J.J. Arnold, M.T. Cygan, T.D. Dunbar, T.P. Burgin, L. Jones II, D.L. Allara, J.M. Tour, P.S. Weiss, Science 271 (1996) 1705. [2] M.T. Cygan, T.D. Dunbar, J.J. Arnold, L.A. Bumm, N.F. Shedlock, T.P. Burgin, L. Jones II, D.L. Allara, J.M. Tour, P.S. Weiss, J. Am. Chem. Soc. 120 (1998) 2721. [3] L.A. Bumm, J.J. Arnold, T.D. Dunbar, D.L. Allara, P.S. Weiss, J. Phys. Chem. B 103 (1999) 8122. [4] T.D. Dunbar, M.T. Cygan, L.A. Bumm, G.S. McCarty, T.P. Burgin, W.A. Reinerth, L. Jones II, J.J. Jackiw, J.M. Tour, P.S. Weiss, D.L. Allara, J. Phys. Chem. B 104 (2000) 4880. [5] Z.J. Donhauser, B.A. Mantooth, K.F. Kelly, L.A. Bumm, J.D. Monnell, J.J. Stapleton, D.W. Price Jr., A.M. Rawlett, D.L. Allara, J.M. Tour, P.S. Weiss, Science 292 (2001) 2303. [6] T. Ishida, W. Mizutani, N. Choi, U. Akiba, M. Fujihira, H. Tokumoto, J. Phys. Chem. B 104 (2000) 11680. [7] S. Wakamatsu, U. Akiba, M. Fujihira, Colloid Surf. A 198–200 (2002) 785. [8] M. Salmeron, G. Neubauer, A. Folch, M. Tomitori, D.F. Ogletree, P. Sautet, Langmuir 9 (1993) 3600. [9] G. Leatherman, E.N. Durantini, D. Gust, T.A. Moore, A.L. Moore, S. Stone, Z. Zhou, P. Rez, Y.Z. Liu, S.M. Lindsay, J. Phys. Chem. B 103 (1999) 4006. [10] D.J. Wold, C.D. Frisbie, J. Am. Chem. Soc. 123 (2001) 5549. [11] X.D. Cui, A. Primak, X. Zarate, J. Tomfohr, O.F. Sankey, A.L. Moore, T.A. Moore, D. Gust, G. Harris, S.M. Lindsay, Science 294 (2001) 571. [12] H. Sakaguchi, A. Hirai, F. Iwata, A. Sasaki, T. Nagamura, Appl. Phys. Lett. 79 (2001) 3708. [13] X.D. Cui, X. Zarate, J. Tomfohr, O.F. Sankey, A. Primak, A.L. Moore, T.A. Moore, D. Gust, G. Harris, S.M. Lindsay, Nanotechnology 13 (2002) 5. [14] D.J. Wold, R. Haag, M.A. Rampi, C.D. Frisbie, J. Phys. Chem. B 106 (2002) 2813.
26
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
[15] S. Wakamatsu, U. Akiba, M. Fujihira, Jpn. J. Appl. Phys. 41 (2002) 4998. [16] T. Ishida, W. Mizutani, Y. Aya, H. Ogiso, S. Sasaki, H. Tokumoto, J. Phys. Chem. B 106 (2002) 5886. [17] S. Fujii, U. Akiba, M. Fujihira, Chem. Commun. (2001) 1688. [18] Y.-T. Tao, C.-C. Wu, J.-Y. Eu, W.-L. Lin, K.-C. Wu, C.-H. Chen, Langmuir 13 (1997) 4018. . J. Am. Chem. Soc. 120 [19] H.-J. Himmel, A. Terfort, C. Woll, (1998) 12069.
[20] G.E. Poirier, Chem. Rev. 97 (1997) 1117. [21] S. Wakamatsu, S. Fujii, U. Akiba, M. Fujihira, Nanotechnology 14 (2003) 258. [22] K. Slowinski, R.V. Chamberlain, C.J. Miller, M. Majda, J. Am. Chem. Soc. 119 (1997) 11910. [23] Y. Kim, C.M. Lieber, Inorg. Chem. 28 (1989) 3990. [24] A. Helms, D. Heiler, G. McLendon, J. Am. Chem. Soc. 114 (1992) 6227. [25] D.E. Eastman, Phys. Rev. B 2 (1970) 1. [26] L. Apker, E. Taft, J. Dickey, Phys. Rev. 73 (1948) 46.
Dependence of tunneling current through a single molecule of phenylene oligomers on the molecular length Satoshi Wakamatsu, Shintaro Fujii, Uichi Akiba, Masamichi Fujihira* Department of Biomolecular Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8501, Japan Received 8 July 2002; received in revised form 17 December 2002
Abstract The electrical properties of single phenylene oligomers were studied in terms of the dependence of the tunneling current on the length of the oligomers using self-assembling techniques and scanning tunneling microscopy (STM). It is important to isolate single molecules in an insulating matrix for the measurement of the conductivity of the single molecule. We demonstrate here a novel self-assembled monolayer (SAM) matrix appropriate for isolation of the single molecules. A bicyclo[2.2.2]octane derivative was used for a SAM matrix, in which the single molecules were inserted at molecular lattice defects. The isolated single molecules of phenylene oligomers inserted in the SAM matrix were observed as protrusions in STM topography using a constant current mode. We measured the topographic heights of the molecular protrusions using STM and estimated the decay constant, b; of the tunneling current through the single phenylene oligomers using a bilayer tunnel junction model. r 2003 Elsevier Science B.V. All rights reserved. Keywords: Phenylene oligomer; Bicyclo[2.2.2]octane derivative; Self-assembled monolayer; Scanning tunneling microscopy
1. Introduction Electronic tunneling through a single molecule is one of the most fundamental and important properties for molecular devices. It has been investigated with molecular resolution using scanning tunneling microscopy (STM) [1–7]. STM can form a bilayer junction where the tunneling electrons flow through a self-assembled monolayer (SAM) and a vacuum gap between a metal tip and a metal substrate surface. In the case of STM, it is difficult to know the distance between the tip and *Corresponding author. Tel.: +81-45-924-5784; fax: +81-45924-5817. E-mail address: [email protected] (M. Fujihira).
the sample surface because the thickness and the conductivity of the sample are convoluted. Bumm et al. estimated the conductance of alkanethiol SAMs from STM data using a bilayer tunnel junction model [3]. The tunneling probability of the composite bilayer junction is the product of the tunneling probabilities of the individual layers, Tt ¼ Tfilm Tgap ; where Tt is the total tunneling probability, Tfilm and Tgap are the tunneling probabilities of a SAM and a vacuum gap, respectively. On the other hand, conducting atomic force microscopy (AFM) enabled us to contact the AFM tip with the sample surface under a controlled load. This method has been used widely to characterize the electrical properties of SAMs [8–16].
0304-3991/03/$ - see front matter r 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-3991(03)00026-3
20
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
Recent self-assembling techniques allowed us to measure electronic tunneling through a single molecule embedded in an insulating matrix using STM and conducting AFM [1,2,4–7,9,11,12,16]. Immersion of a preformed host SAM in a dilute solution of guest molecules resulted in isolating them in the host SAM. The isolated molecules were considered to be inserted at defects of the preformed matrices, such as step edges of the Au(1 1 1) substrate surface and structural domain boundaries of the host SAMs [1,2]. Isolated conductive molecules appear as protrusions in an STM topographic image using a constant current mode. In most cases, an alkanethiol SAM was used as the host matrix. Bumm et al. inferred that many of the like features at structural domain boundaries of the alkanethiol SAM matrix are individual molecules, because access for insertion is limited at the structural domain boundaries surrounded by the close-packed film [1,2]. Cui et al. reported electronic conduction through an isolated single molecule sandwiched between an Au nanoparticle and an Au(1 1 1) surface, in which either end of the isolated molecule was chemically bonded [11]. They observed that I V measurements on the Au nanoparticles binding to the isolated molecules produced five distinct families of curves. The curves corresponded to multiples of a fundamental curve, which was ascribed to the I V curve of the single molecule. This implies that the isolated molecules in the SAM matrix are not always the single molecules. Therefore, when we used the alkanethiol SAM as the host matrix, it is difficult to judge whether individual or bundled molecules were isolated at the defect sites. We used a bicyclo[2.2.2]octane derivative (BCO) SAM as the host matrix [17] and a series of phenylene oligomers, C6H5(C6H4)m1CH2SH (m ¼ 1; 2; 3), as the conductive guest molecules. The BCO SAM has unique properties that typical alkanethiol SAMs do not have. First, the BCO SAM [17] has no structural domain boundary [1,2]. Second, the BCO SAM has molecular lattice defects, which are defects of single molecules described later. The size of the molecular lattice defect is considered to be that of the single BCO molecule. Because the BCO SAM matrix has no structural domain boundary, inserted guest
molecules exist at the molecular lattice defects in the matrix, so that single guest molecules are certainly isolated in the BCO SAM matrix. In this study, we investigated electronic tunneling through isolated single molecules of phenylene oligomers embedded in the BCO SAM matrices using STM. We succeeded in quantitative measurement for electronic tunneling through the single molecule of phenylene oligomers. We estimated the dependence of the tunneling current on the molecular length of phenylene oligomers, i.e. the decay constant b: This was accomplished by measuring the heights of protrusions of individual phenylene oligomers in STM images and by analyzing the heights using the bilayer tunnel junction model.
2. Experimental We prepared three types of phenylene oligomers (benzenemethanethiol; BMT, 4-biphenylmethanethiol; BPMT, and [1,10 :40 ,100 -terphenyl]-4methanethiol; TPMT) and a bicyclo[2.2.2]octane derivative (bis(bicyclo[2.2.2]octylmethyl) disulfide; BCO) shown in Fig. 1. BMT was purchased from Tokyo Chemical Industry. BPMT, TPMT, and BCO were synthesized according to the literature [17–19]. Au(1 1 1) substrates were prepared by thermal evaporation of Au onto freshly cleaved
BMT SH
BPMT SH
TPMT SH
BCO
S
S
Fig. 1. Chemical structures of BMT, BPMT, TPMT, and BCO.
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
s
s
s
s
s
s
s
Au (111) (a)
s
s
s
s
s
s
Au (111) (b)
s
21
s
s
s
s
Au (111) (c)
Fig. 2. Schematic representations of three types of SAMs in which single phenylene oligomers of (a) BMT, (b) BPMT, and (c) TPMT are isolated in BCO SAM matrices on Au[1 1 1], respectively.
Fig. 3. STM images of three type of SAMs on Au[1 1 1]. All STM images were obtained with a tungsten tip using a constant current mode (30 pA) at a sample bias voltage of +1.0 V. (a) An STM image of the BCO SAM (15 15 nm2 in size). Single Au atom deep pits and a hexagonally close-packed lattice were observed. The BCO SAM has no structural domain boundary. Molecular lattice defects are highlighted by circles. (b) An STM image of a typical alkanethiol SAM composed of decanethiol (15 15 nm2 in size). The structural domain boundaries and the single Au atom deep pits were observed. (c) An STM image of a decanethiol SAM matrix in which guest BPMT molecules were inserted (40 40 nm2 in size). Bright spots were assigned as the inserted BPMT molecules.
mica that was heated in vacuum to 300 C [7]. The substrates were removed from the chamber and immersed immediately in thiol solutions. Three types of SAMs (Figs. 2(a)–(c)) containing isolated single molecules were prepared in a twostep process [1]. First, SAM matrices of BCO were prepared on the Au substrates by immersion for 24 h or longer in a 0.1 mM ethanol solution of BCO. Second, after thorough rinsing of the SAM matrices with ethanol, the BCO SAM matrices were immersed for a few minutes in a 1 mM ethanol solution of BMT or BPMT, or in a 0.1 mM tetrahydrofuran (THF) solution of TPMT. Then, the resulting two-component SAMs were rinsed thoroughly with the same solvents as those used for the second step, and dried. STM images were obtained with a tungsten tip using a constant current mode (30 pA) at the sample bias voltage of +1.0 V under an ultra high vacuum (UHV) condition (o5 108 Pa) (a JEOL
JSPM-4500S/TM-51021 UHV scanning tunneling microscope). Current vs. separation between the tip and the sample (I S) characteristics were measured in UHV by first setting an initial gap separation with an initial tunneling current (30 pA), second disconnecting the feedback loop of the z piezoelectric regulation, and finally moving the tip towards the sample surface.
3. Results and discussion Fig. 3(a) shows a constant-current STM image of a 15 15 nm2 area of the BCO SAM on Au[1 1 1]. All images in Fig. 3 were recorded with a sample bias voltage of +1.0 V and a tunneling current of 30 pA. In the same way as in an n-alkanethiol SAM (Fig. 3(b)), single Au atom deep pits and a hexagonally close-packed lattice are seen in the BCO SAM shown in Fig. 3(a) [17].
22
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
A typical alkanethiol SAM matrix contains a variety of characteristic defect sites, such as the structural domain boundaries [1,2,20] and the single Au atom deep pits [20], as shown in Fig. 3(b). On the other hand, owing to the bulkiness of the BCO moieties, they are expected to be oriented perpendicularly to the substrate surface. Thus the formation of domain structure due to the tilt of adsorbates can be avoided, so that the structural domain boundaries were not observed in Fig. 3(a) [17]. In addition, molecular lattice defects, which are defects of single BCO molecules and highlighted by the circles in Fig. 3(a), were clearly observed on the BCO SAM surface. The molecular lattice defects are unique to the BCO SAM. They were not observed on the typical alkanethiol SAM surfaces as shown in Fig. 3(b). Bumm et al. suggested that exposure of one component SAM to a solution containing guest molecules resulted in the molecules inserted in SAM defect sites [1]. At these sites, the guest molecules can approach to the underlying gold atoms so that the sulfur head groups can chemisorb to the gold surface. Therefore, when the guest molecules are more conductive than the molecules forming the SAM matrix, STM images show that the guest molecules protrude from the SAM matrix, even if the physical height of the guest molecules is lower than that of the SAM matrix as shown in Fig. 3(c) [2,4,7]. In Fig. 3(c),
the guest molecule BPMT was inserted in a decanethiol SAM matrix. The length of the BPMT molecules is considered to be slightly shorter than that of the decanethiol SAM matrix. The bright spots in Fig. 3(c) were assigned as the inserted BPMT molecules. However, because the STM tip–sample interaction convolves electronic and physical structure [1,2], it is difficult to judge from the STM images whether individual or bundled molecules were isolated at the defect sites. We used the BCO SAM matrix as a host to isolate phenylene oligomers in the present study. As described above, the BCO SAM has no structural domain boundary so that the inserted phenylene oligomers exist at the substrate step edges, at the edge of the single Au atom deep pits, and the molecular lattice defects. Among them, only each molecular lattice defect of the last ones can always allow a single molecule to be isolated due to the size of the molecular lattice defect. Namely, single guest molecules (not bundles of molecules) are isolated at the molecular lattice defects in the BCO SAM matrix. STM images of the two-component SAMs containing the single molecules of BMT, BPMT, and TPMT are shown in Figs. 4(a)–(c), respectively. The images shown in Fig. 4 were recorded with the sample bias of +1.0 V and the tunneling current of 30 pA. Some bright spots were observed in each STM image. We assigned the bright spots in a terrace as the single molecules of phenylene
Fig. 4. STM images (20 20 nm2 in size) and topographic cross sections of the two-component SAMs containing the single molecules of (a) BMT, (b) BPMT, and (c) TPMT. All STM images were obtained with a tungsten tip using a constant current mode (30 pA) at a sample bias voltage of +1.0 V. Bright spots were assigned as single molecules. The topographic cross sections were extracted from the STM images on scans indicated by black lines.
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
oligomers inserted at the molecular lattice defects of the BCO SAM matrices. The difference in size and shape of the bright spots can be ascribed to thermal motion of the isolated single phenylene oligomers. Topographic cross sections extracted from the above three STM images along the black lines are also shown in Figs. 4(a)–(c), respectively. These lines pass over the isolated molecules protruding from the matrices. The different heights of the protrusions for the same phenylene oligomer were observed in each STM image, although they were isolated in the same SAM matrices. When we repeated imaging the same area, the observed height of one protrusion located in the same position was changed also with time. We considered that these phenomena are caused by change in the tilt of the molecular axis of the isolated single phenylene oligomer [21]. In Fig. 5, histograms of values of the STM topographic height from the surroundings show that the heights of protrusions for (a) BMT, (b) BPMT, and (c) TPMT range over 0.10–0.22,
Fig. 5. Histograms of values of the STM topographic heights of protrusions from the surroundings for (a) BMT, (b) BPMT, and (c) TPMT. Heights of the protrusions of BMT, BPMT, and TPMT range from 0.10 to 0.22 nm, from 0.14 to 0.42 nm, and from 0.20 to 0.68 nm, respectively.
23
Fig. 6. A representative I2S characteristic on the BCO SAM surface. A change in slope at z ¼ 0:2470.02 nm was considered to be caused by the contact of the tip with the top of the SAM. A linear fit in the vacuum region gives a line with the decay constant of 11.470.5 nm1.
0.14–0.42, and 0.20–0.68 nm, respectively. All of the data in the histograms, i.e. counts vs. topographic height, were extracted from more than 100 images, which contain the bright spots corresponding to the single molecules, for each phenylene oligomer. In Fig. 4, the cross sections were drawn across the highest protrusions among observed ones for each guest molecule. Here, it is reasonable that the isolated single molecules corresponding to the highest protrusions are tilted perpendicularly to the substrates. The highest protrusions of BMT, BPMT, and TPMT are 0.22, 0.42, and 0.68 nm, respectively. Differences in the height of the single molecular protrusions perpendicular to the substrates are 0.20 nm between BMT and BPMT, and 0.26 nm between BPMT and TPMT. Fig. 6 shows a representative I S characteristic above the BCO SAM surface. A near exponential dependence of the current as a function of the z-position was observed. In this measurement, the initial z-position (z ¼ 0 nm) was set by feed back control under the constant current
24
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
mode STM with 30 pA. Then, the tip was moved towards the sample surface under the fixed sample bias voltage at +1.0 V after disconnecting the feedback. When the tip moved close to the SAM, a sudden change in slope was observed at z ¼ 0:2470.02 nm. We assume that this change in slope was caused by the contact of the tip with the top of the SAM. A fitted straight line for the plots in a vacuum region was drawn on the graph. The decay constant of the tunneling current through the vacuum above the BCO SAM surface was 11.470.5 nm1. Using these results described above, we estimated the dependence of the tunneling current on the length of phenylene oligomers at the same sample bias voltage of +1.0 V. We assumed that contribution by the tunneling current through space (chain-to-chain coupling) in the SAM is negligible [22]. We obtained the value of b for the tunneling current through the single phenylene oligomers using the bilayer tunnel junction model according to the following procedure. (i) In Fig. 7,
Fig. 7. A semi-logarithmic graph of the tunneling current vs. separation between the tip and the sample. The I2S characteristic shown in Fig. 6 was replotted on this graph. The contact point of the tip with the BCO SAM was defined newly as the position at z ¼ 0 nm. Open squares are the STM topographic heights (see text) at the tunneling current of 30 pA above the BCO SAM matrix for the isolated single phenylene oligomers of BMT, BPMT, and TPMT. The height of the tip on the BCO SAM matrix under 30 pA was also plotted as an open square. Filled circles are the supposed contact points of the tip with the isolated single phenylene oligomers of BMT, BPMT, and TPMT, respectively. A linear fit of the filled circles gives a b value of 5.370.2 nm1.
a semi-logarithmic graph of the tunneling current vs. separation between the tip and the BCO SAM surface was plotted. The I2S characteristic shown in Fig. 6 was replotted on this graph. The contact point of the tip with the BCO SAM surface was defined newly as the position at z ¼ 0 nm. (ii) The measured heights of the isolated single phenylene oligomers of BMT, BPMT, and TPMT (the open squares) above the BCO SAM matrix in Fig. 4 were plotted on the dotted line corresponding to the constant tunneling current of 30 pA. The height of the tip on the BCO SAM matrix under 30 pA was also plotted. Namely, the STM topographic heights of the isolated single phenylene oligomers of BMT, BPMT, and TPMT above BCO were 0.22, 0.42 and 0.68 nm, respectively. If we could measure I2S curves above the isolated single molecules, we would expect that the I2S curves in vacuum fall on thin solid lines in Fig. 7 with the same decay constant as 11.4 nm1 above the BCO SAM matrix. It was, however, difficult to measure the I2S curves above the isolated single molecules because of the drift of the tip position. (iii) The molecular length of one phenylene unit is B0.43 nm [14,23,24]. If phenylene oligomers are oriented perpendicularly to the surface, the physical height of one phenylene unit is 0.43 nm. The physical height of BMT is determined to be slightly lower (B0.02 nm) than that of BCO from molecular models. The lines at each terminal z-position of phenylene oligomers, BMT, BPMT, and TPMT, were drawn as the dashed lines. (iv) The filled circles are plotted at the points of intersections between the supposed I2S curves in vacuum and the dashed lines for each phenylene oligomer. The filled circles are the points, where the tip is just in contact with the end of the isolate single phenylene oligomers. Thus the tunneling currents of these points correspond to what would be measured at the contacts. The linear fit (a thick solid line) of the filled circles gives a b value of 5.370.2 nm1. This value of b is the decay constant of tunneling current through the single molecule of phenylene oligomers at the sample bias voltage of +1.0 V. In our previous study [15], we measured the I2V characteristics of the phenylene oligomer SAMs using conducting AFM with Au-coated tips
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
modified chemically by decanethiol. The value of b ¼ 5:570.6 nm1 was obtained at the sample bias voltage of +1.0 V. The b value of 5.3 nm1 in the present work is slightly lower than that of 5.5 nm1 obtained using conducting AFM. The difference between these two values of b might be due to the difference in the barrier heights for electronic tunneling, because we used the tungsten STM tips in this study and the Au-coated AFM tips in the previous work. The work functions of Au (polycrystalline) and W (polycrystalline) are 5.1 and 4.5 eV, respectively [25,26]. The difference of the work functions causes the difference in the barrier heights for electronic tunneling. In our case, because the tip was served as the electron donor electrode due to the negative tip bias voltage, the lower work function of the tip electrode explains the lower barrier height for electron tunneling. However, in our previous study, we assumed that hole tunneling was dominant through these phenylene oligomer SAMs using bilayer junctions formed by conducting AFM with the chemically modified Au-coated tips [15]. In this present STM system, it is difficult to understand the mechanism of the hole tunneling through a vacuum gap. Therefore, here we can only suggest that the value of b ¼ 5:3 nm1 is reasonable, compared to that of b ¼ 5:5 nm1 measured by conducting AFM for the SAMs of phenylene oligomers. We are now investigating the effect of the work functions of metal tips on STM and conducting AFM.
4. Conclusions We made the novel host BCO SAMs, in which the single phenylene oligomers were inserted as isolated single molecules. The BCO SAMs have no structural domain boundary so that the guest molecules are inserted at the molecular lattice defects of the BCO SAM matrices. These SAMs allowed us to investigate the electrical properties of the single isolated phenylene oligomers using STM. In this way, we succeeded in estimating the dependence of the tunneling current through the single phenylene oligomers on the molecular length by STM. The reasonable value of
25
b ¼ 5:3 nm1 for the single phenylene oligomers at the sample bias voltage of +1.0 V was obtained from the STM measurement of the molecular protrusion.
Acknowledgements This work was supported by Grant-in-Aid for Creative Scientific Research on ‘‘Devices on molecular and DNA levels’’ (No. 13GS0017) from the Ministry of Education, Science, Sports, and Culture of Japan.
References [1] L.A. Bumm, J.J. Arnold, M.T. Cygan, T.D. Dunbar, T.P. Burgin, L. Jones II, D.L. Allara, J.M. Tour, P.S. Weiss, Science 271 (1996) 1705. [2] M.T. Cygan, T.D. Dunbar, J.J. Arnold, L.A. Bumm, N.F. Shedlock, T.P. Burgin, L. Jones II, D.L. Allara, J.M. Tour, P.S. Weiss, J. Am. Chem. Soc. 120 (1998) 2721. [3] L.A. Bumm, J.J. Arnold, T.D. Dunbar, D.L. Allara, P.S. Weiss, J. Phys. Chem. B 103 (1999) 8122. [4] T.D. Dunbar, M.T. Cygan, L.A. Bumm, G.S. McCarty, T.P. Burgin, W.A. Reinerth, L. Jones II, J.J. Jackiw, J.M. Tour, P.S. Weiss, D.L. Allara, J. Phys. Chem. B 104 (2000) 4880. [5] Z.J. Donhauser, B.A. Mantooth, K.F. Kelly, L.A. Bumm, J.D. Monnell, J.J. Stapleton, D.W. Price Jr., A.M. Rawlett, D.L. Allara, J.M. Tour, P.S. Weiss, Science 292 (2001) 2303. [6] T. Ishida, W. Mizutani, N. Choi, U. Akiba, M. Fujihira, H. Tokumoto, J. Phys. Chem. B 104 (2000) 11680. [7] S. Wakamatsu, U. Akiba, M. Fujihira, Colloid Surf. A 198–200 (2002) 785. [8] M. Salmeron, G. Neubauer, A. Folch, M. Tomitori, D.F. Ogletree, P. Sautet, Langmuir 9 (1993) 3600. [9] G. Leatherman, E.N. Durantini, D. Gust, T.A. Moore, A.L. Moore, S. Stone, Z. Zhou, P. Rez, Y.Z. Liu, S.M. Lindsay, J. Phys. Chem. B 103 (1999) 4006. [10] D.J. Wold, C.D. Frisbie, J. Am. Chem. Soc. 123 (2001) 5549. [11] X.D. Cui, A. Primak, X. Zarate, J. Tomfohr, O.F. Sankey, A.L. Moore, T.A. Moore, D. Gust, G. Harris, S.M. Lindsay, Science 294 (2001) 571. [12] H. Sakaguchi, A. Hirai, F. Iwata, A. Sasaki, T. Nagamura, Appl. Phys. Lett. 79 (2001) 3708. [13] X.D. Cui, X. Zarate, J. Tomfohr, O.F. Sankey, A. Primak, A.L. Moore, T.A. Moore, D. Gust, G. Harris, S.M. Lindsay, Nanotechnology 13 (2002) 5. [14] D.J. Wold, R. Haag, M.A. Rampi, C.D. Frisbie, J. Phys. Chem. B 106 (2002) 2813.
26
S. Wakamatsu et al. / Ultramicroscopy 97 (2003) 19–26
[15] S. Wakamatsu, U. Akiba, M. Fujihira, Jpn. J. Appl. Phys. 41 (2002) 4998. [16] T. Ishida, W. Mizutani, Y. Aya, H. Ogiso, S. Sasaki, H. Tokumoto, J. Phys. Chem. B 106 (2002) 5886. [17] S. Fujii, U. Akiba, M. Fujihira, Chem. Commun. (2001) 1688. [18] Y.-T. Tao, C.-C. Wu, J.-Y. Eu, W.-L. Lin, K.-C. Wu, C.-H. Chen, Langmuir 13 (1997) 4018. . J. Am. Chem. Soc. 120 [19] H.-J. Himmel, A. Terfort, C. Woll, (1998) 12069.
[20] G.E. Poirier, Chem. Rev. 97 (1997) 1117. [21] S. Wakamatsu, S. Fujii, U. Akiba, M. Fujihira, Nanotechnology 14 (2003) 258. [22] K. Slowinski, R.V. Chamberlain, C.J. Miller, M. Majda, J. Am. Chem. Soc. 119 (1997) 11910. [23] Y. Kim, C.M. Lieber, Inorg. Chem. 28 (1989) 3990. [24] A. Helms, D. Heiler, G. McLendon, J. Am. Chem. Soc. 114 (1992) 6227. [25] D.E. Eastman, Phys. Rev. B 2 (1970) 1. [26] L. Apker, E. Taft, J. Dickey, Phys. Rev. 73 (1948) 46.