Bonding Asymmetry and Adatoms in Low-Density ... - ACS Publications

ARTICLE pubs.acs.org/JPCC

Bonding Asymmetry and Adatoms in Low-Density Self-Assembled Monolayers of Dithiols on Au(111) Aisyah M. Sharif,† D. Noel Buckley,† Manfred Buck,‡ and Christophe Silien*,† † ‡

Department of Physics and Energy, and Materials and Surface Science Institute, University of Limerick, Limerick, Ireland EaStCHEM School of Chemistry, University of St Andrews, St Andrews KY16 9ST, United Kingdom ABSTRACT: The involvement of Au adatoms in the interfacial structure of monothiols on Au(111) is now established, but the same issue is virtually unexplored in the case of dithiols. To gain more insight into the latter, we have studied by scanning tunneling microscopy the self-assembly of two prototypic symmetric dithiols (1,6-hexanedithiol and biphenyl-4,40 -dimethanethiol) from dilute aqueous solutions and correlated their growth with the deconstruction of the Au(111) herringbone pattern known to produce adatoms. For both molecules, we determine an initial low-density monolayer where the molecules are lying down and paired by 0.45 Å tall protrusions, assigned to Au adatoms. The other terminal group is imaged differently, revealing a strong asymmetry in the dithiol bonding. The formation of vacancy islands and, thus, the extraction of additional adatoms from terraces are detected only after substantial molecular rearrangement and loss of bonding asymmetry.

1. INTRODUCTION Thiol monolayers have introduced an unprecedented flexibility in surface functionalization1
3 and are today among the most widely investigated self-assembled monolayers (SAMs) on metals. The development of molecular electronics also implicates thiols such as the dithiols, with their ability to bond to two metal electrodes.4 However, if the formation of highly ordered saturated films of standing-up monothiol is well-established, the alkanethiol SAMs on Au(111) being the leading example,1
3 comparable organization in saturated dithiol monolayers remains challenging5
8 since intermolecular bonding9 and attachment of both thiol moieties onto the substrate10
12 are significant sources of defects. Moreover, akin to monothiols, the self-assembly of dithiols involves at low coverage molecules that are adsorbed lying down on the substrate,1
3 and at saturation coverage a majority of standing-up molecules,5
8 with thus a change in orientation that should be thermodynamically preferred since it maximizes the density of Au thiolates but that can be also kinetically hindered.10 A comprehensive understanding of the interface between dithiols and Au(111) at the atomic and molecular level is essential for rationalizing these observations and ensuring a reliable exploitation of the dithiols. Yet, it is only very recently that a picture has emerged for the comparatively simpler case of monothiol adsorption, in which thiol bonding on Au(111) is seen as involving Au adatoms.13
19 With respect to the ubiquitous saturated alkanethiol SAMs, there remains a debate on whether the thiol to adatom ratio is 1:1 or 2:1.16
19 At coverage below saturation, scanning tunneling microscopy (STM) revealed dimers where one adatom bonds two monothiols that are lying down on an otherwise un-reconstructed surface.16 The adatoms are initially supplied by deconstruction of r 2011 American Chemical Society

the herringbone pattern,20,21 which exhibits a uniaxial 4
5% compression in comparison to a relaxed (111) plane and releases 0.7 Au/nm2. An increase in monothiol coverage implicates more adatoms that are provided by step edges and terraces, where vacancy islands then develop.13,22 It is not yet established how adatoms relate to dithiol SAMs on Au(111), even though unraveling the bonding between molecules and substrate is ultimately a key to understanding the transition from lying-down to standing-up orientation. With the success of STM in revealing adatom-mediated monothiol dimers and establishing that, in the early stage of deposition, deconstruction of the herringbone pattern is the dominant source of adatoms on terraces,13,16 we rationalize that the same technique should also enlighten our understanding of dithiol self-assembly, in particular for coverage below saturation. Therefore, we report here on our STM investigation of the deposition of dithiol molecules from dilute aqueous solutions on Au(111). 1,6-Hexanedithiol (HDT) and biphenyl-4,40 -dimethanethiol (BPDMT) were chosen as two representative molecules involving aliphatic and aromatic spacers, and water was selected as solvent to prevent coadsorption with organic solvent molecules.3 Although low-density dithiol monolayers on Au(111) have been previously analyzed with STM when prepared by annealing or electrochemical desorption of saturated SAMs,10,23 brief deposition at very low dithiol concentration and static observation were preferred since they allow us to correlate the monolayer growth to the deconstruction of the herringbone pattern. Received: July 8, 2011 Revised: September 22, 2011 Published: October 01, 2011 21800

dx.doi.org/10.1021/jp206457g | J. Phys. Chem. C 2011, 115, 21800–21803

The Journal of Physical Chemistry C

ARTICLE

Figure 2. (a, b) Constant current STM images of α-BPDMT recorded on the same sample with an uncontrolled change in the tip structure between the scans (10 pA;
0.75 V; scale bars = 2.0 nm). (c) Proposed model for pattern α. The circles and the adjacent short lines represent the Au adatoms and the dithiol molecules, respectively. The hexagonal grid in the background marks the Au(111) deconstructed surface. Figure 1. (a) Constant current STM image of a clean Au(111) surface exhibiting the herringbone pattern (50 pA; 0.5 V; scale bar = 20 nm). (b, c) Images of α-HDT recorded at different magnifications (30 pA;
0.75 V; scale bars = 2.0 and 50 nm, respectively). (d, e) Height profiles corresponding to the lines labeled P1 and P2 in the image shown in b.

2. EXPERIMENTAL SECTION All measurements were done in ambient air, at room temperature, with an AGILENT 5500 (see figure captions for tunnel current and sample bias) and tips prepared by mechanically cutting a Pt:Ir (80:20) wire. The samples were prepared by immersing freshly annealed Au/mica substrates (GEORG ALBERT PVD) in aqueous (18 MΩ deionized H2O) solutions (∼0.5 nM nominal concentration) of HDT (Aldrich) or BPDMT (see ref 24) typically for a few minutes. The samples were then rinsed with deionized water and dried with N2 gas. 3. RESULTS AND DISCUSSION Prior to immersion in the dithiol solutions, our substrates exhibit a clear herringbone pattern21 (Figure 1a) confirming that clean Au(111) terraces are indeed formed. After a brief deposition of HDT, STM reveals a new pattern (Figure 1b
e) that covers the entire surface within minutes and that is referred to as α-HDT in the following. At high resolution, rows of 0.45 Å tall protrusions are seen running along the Æ112æ axis with a periodicity of 5 Å. They are labeled A in Figure 1b and repeat along the Æ110æ direction every 3.2 and 2.4 nm in a regular alternating sequence, leading to the apparent periodicity of 5.6 nm seen at low resolution in Figure 1c. However, the variations in the angle between the molecular features and the Æ110æ axis that are visible at high resolution in Figure 1b indicate that the actual periodicity is at least twice as large. Central rows are indicated B and C, respectively. In keeping with the elongated molecular features seemingly joining these rows A and with the length of HDT estimated between 1.0 and 1.3 nm,10,25 we propose that two dithiols are accommodated in both the 3.2 and 2.4 nm spacing, giving an average area of 70 Å2/molecule.

As will be detailed below, when left in ambient air for several days, α-HDT evolves into another ordered pattern closely resembling one characterized previously by others10 that involves lyingdown molecules also with 70 Å2/HDT, strengthening thus our proposition. Substitution of the aliphatic backbone by an aromatic one, using BPDMT instead of HDT, has no major effect on the observed pattern and on its dimensions (Figure 2a), within our experimental accuracy ((5%). Although images such as the one in Figure 2a were typically measured, the STM also revealed other contrasts such as the one shown in Figure 2b. Since both were recorded on the same BPDMT samples, which are homogeneous in large-scale images, and since which pattern is seen depends on the tip history, it is proposed that both correspond to the same molecular arrangement but highlight a variation in the tip structure. Rows A
C are marked in Figure 2b to emphasize that correspondence. A appears now as a double row of protrusions, and C consists of a single row of protrusions, which evidence that C is seen as a dark row in the other images because of a reduced conductivity and not because of a topographic depression. With the herringbone pattern not visible after adsorption of HDT or BPDMT, and with the density of protrusions in the single rows A in α-HDT and α-BPDMT (Figure 1b and Figure 2a) matching perfectly with the amount of adatoms expected from the herringbone deconstruction (0.7 Au/nm2),13,16 we propose to attribute the protrusions in A to Au adatoms. The occasional doubling of these rows (Figure 2b) is then interpreted as evidence of the terminal sulfur atoms in the molecules adjacent to the Au adatoms. A qualitative model of the interface is summarized in Figure 2c. Although small variations in the dimensions are expected between α-HDT and α-BPDMT, these were not detected in the experiments and no attempt was made to account for them in the model. We choose arbitrarily to position the Au adatoms on bridge sites on a deconstructed Au(111) plane to fit with the calculations made for Au adatoms in methanethiol SAMs.13,16 As drawn, the model accounts for a molecular length of 1.1 nm, and 21801

dx.doi.org/10.1021/jp206457g |J. Phys. Chem. C 2011, 115, 21800–21803

The Journal of Physical Chemistry C

Figure 3. (a) Constant current STM image of HDT after 5 days showing both α and β phases; enlargement of β-HDT in inset (10 pA;
0.75 V; scale bars = 50 and 5.0 nm in inset). (b, c) Height profiles P3 and P4 extracted from the images shown in a. (d) STM images of HDT after 19 days highlighting large vacancy islands (10 pA;
0.75 V; scale bars = 20 and 5.0 nm in inset). (e, f) Height profiles P5 and P6 extracted from d. (g) STM image of BPDMT after 4 days (30 pA;
0.80 V; scale bar = 20 nm).

for the dimensions between rows A (i.e., 3.2 and 2.4 nm) determined in the STM images. Because of the strong similarity of our images to those of adatom-mediated methanethiol dimers,13,16 we attribute the structure to dithiol dimers mediated in the same way by one Au atom, with adatoms and molecules virtually in the same plane above the deconstructed Au(111) surface. Notably, there is no similar protrusion attributable to an adatom at the other end of the dithiols; either the molecules terminate by a minor protrusion in B that can be related to the second thiol moiety or the molecules end in the dark row C, where the proximity of both S moieties raises the possibility of intermolecular bonding. Regardless of our interpretations, the experimental data imply that neither HDT nor BPDMT ever terminate by the same feature on both sides and that all dithiols are thus bonded asymmetrically on the substrate. Such observations highlight a complex balance involving the dithiols, the Au(111) substrate, and the Au adatoms. With the cumulated length ABA and ACA nearly equivalent to the periodicity of the herringbone pattern, we propose that the nucleation of the α-HDT and α-BPDMT domains also plays a determining role in the equilibrium that is eventually reached. STM images recorded after 5 days on a sample which showed a uniform α-HDT coverage immediately after deposition still display wide areas of α-HDT (Figure 3a). Interestingly, these do not show any vacancy islands that are known to progressively expand on SAMs following the coalescence of atomic vacancies often created upon chemisorption of thiols.22,26 The observation is also made for BPDMT, and it is then confirmed that no adatoms

ARTICLE

are extracted from the Au(111) terraces to self-assemble α-HDT and α-BPDMT. The herringbone pattern is thus the sole source of adatoms, and because no Au islands are detected,22 all adatoms must be embedded in the monolayers, in line with our model. After 5 days, the STM also reveals a different pattern labeled β-HDT (Figure 3a
c). The new domains remain characterized by rows aligned with the Æ112æ axis but exhibit a periodicity of 1.4 nm in the Æ110æ direction. This β-HDT pattern matches closely with HDT SAMs prepared by others by vacuum deposition, where the rows were identified as lying-down molecules stacked every 5 Å with their molecular axis aligned with the Æ110æ direction,10 and the same arrangement is proposed here. Even though both α and β structures involve lying-down molecules, the profile P4 reveals that the top height in β-HDT is 0.9 Å above that of α-HDT (Figure 3c). With the STM also showing no asymmetry in the molecular contrast, unlike for α-HDT, we deduce that a drastic change in interfacial bonding occurred upon formation of β-HDT. Contrasting the model of in-plane Au adatom-mediated monothiol dimers mentioned earlier,13,16 other authors suggested that, for saturated alkanethiol SAMs, the molecules are bonded atop one Au adatom, forming a complex that is adsorbed on the deconstructed Au(111) surface.17 The substantial height increase in β-HDT compared to α-HDT may thus mark a rearrangement of the adatoms to an out-of-plane geometry. However, since β-HDT domains are bordered by unresolved areas, the amount of adatoms involved is not revealed in the experiments and thus their true involvement cannot be further modeled. After 19 days (Figure 3d
f), large vacancy islands are unambiguously detected because of their characteristic depth of 2.4 Å and highlight that eventually more adatoms are accommodated in the monolayer compared to α-HDT. At that stage, the periodicity of the row is measured at 1.1 nm, suggesting further rearrangement of the SAM and possibly formation of a phase similar to one previously reported for octanedithiol SAMs prepared by electrochemical desorption that showed a periodicity of 1.2 nm.23 Left in ambient for 4 days (Figure 3g), originally uniform α-BPDMT samples evolve into disordered structures, in contrast to HDT, thus revealing an eventual influence of the aromatic moieties.

4. CONCLUSIONS In summary, by self-assembling dithiols on Au(111) from diluted aqueous solutions, we have observed with STM a lowdensity phase α common to HDT and BPDMT. The images reveal that the molecules are lying down on the substrate and that their ends are not symmetric, implying a substantial asymmetry in bonding. This observation has been rationalized by proposing that the 0.7 adatom/nm2 released upon deconstruction of the herringbone pattern are accommodated in the monolayer with a molecule to adatom ratio of 2:1. It is tempting to relate the bonding asymmetry seen here with the typically observed, but nonetheless puzzling, evolution of dithiol SAMs, which involve lying-down molecules at low coverage and standing-up ones at saturation,1
3 by proposing that phases of asymmetrically bonded molecules such as α are precursors. Moreover, our α-HDT SAMs evolve after several days into β-HDT where the STM images show the molecules to be symmetric. With the reasonable assumption that β-HDT is more stable since it involves a lengthy rearrangement of the interface, we relate the occasional trapping of HDT in SAMs, where the molecules remain lying down,10 to the now symmetric involvement of the thiol moieties with 21802

dx.doi.org/10.1021/jp206457g |J. Phys. Chem. C 2011, 115, 21800–21803

The Journal of Physical Chemistry C respect to the bonding. Novel features regarding the complex self-assembly of dithiol molecules on Au(111) were thus unraveled, emphasizing the involvement of Au adatoms and a remarkable asymmetry in bonding at the initial stage. With the ubiquity of the thiol chemistry on Au, we believe that our observations will fuel further and much needed understanding of this important interface.

ARTICLE

(22) Poirier, G. E. Langmuir 1997, 13, 2019–2026. (23) Esplandiu, M. J.; Carot, M. L.; Cometto, F. P.; Macagno, V. A.; Patrito, E. M. Surf. Sci. 2006, 600, 155–172. (24) Tai, Y.; Shaporenko, A.; Rong, H.-T.; Buck, M.; Eck, W; Grunze, M.; Zharnikov, M. J. Phys. Chem. B 2004, 108, 16806–16810. (25) Ishizuka, K.; Suzuki, M.; Fujii, S.; Akiba, U.; Takayama, Y.; Sato, F.; Fujihira, M. Jpn. J. Appl. Phys. 2005, 44, 5382–5385. (26) Cavalleri, O.; Hirstein, A.; Kern, K. Surf. Sci. 1995, 340, 960–964.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Tel.: 353 61234177. Fax: 353 61202423.

’ ACKNOWLEDGMENT A.M.S. acknowledges a grant from the Ministry of Higher Education (MOHE) in Malaysia and C.S. funding from the Integrated Nanoscience Platform for Ireland (INSPIRE), initiated by the Higher Education Authority in Ireland within the PRTLI4 framework. ’ REFERENCES (1) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. Rev. 2005, 105, 1103–1170. (2) Vericat, C.; Vela, M. E.; Benitez, G.; Carro, P.; Salvarezza, R. C. Chem. Soc. Rev. 2010, 39, 1805–1834. (3) Schreiber, F. Prog. Surf. Sci. 2000, 65, 151–256. (4) McCreery, R. L.; Bergren, A. J. Adv. Mater. 2009, 21, 1–20. (5) Hamoudi, H.; Prato, M.; Dablemont, C.; Cavalleri, O.; Canepa, M.; Esaulov, V. A. Langmuir 2010, 26, 7242–7247. (6) Qu, D; Kim, B.-C.; Lee, C.-W. J.; Ito, M.; Noguchi, H.; Uosaki, K. J. Phys. Chem. C 2010, 114, 497–505. (7) Silien, C.; Dreesen, L.; Cecchet, F.; Thiry, P. A.; Peremans, A. J. Phys. Chem. C. 2007, 111, 6357–6364. (8) Pasquali, L.; Terzi, F.; Seeber, R.; Doyle, B. P.; Nannarone, S. J. Chem. Phys. 2008, 128, 134711. (9) García-Raya, D; Madue~no, R.; Blazquez, M.; Pineda, T. J. Phys. Chem. C 2010, 114, 3568–3574. (10) Leung, T. Y. B.; Gerstenberg, M. C.; Lavrich, D. J.; Scoles, G.; Schreiber, F.; Poirier, G. E. Langmuir 2000, 16, 549–561. (11) Kim, Y.-H.; Gorman, C. B. Langmuir 2011, 27, 6069–6075. (12) Akkerman, H. B.; Kronemeijer, A. J.; van Hal, P. A.; de Leeuw, D. M.; Blom, P. W. M.; de Boer, B. Small 2008, 4, 100. (13) Maksymovych, P.; Voznyy, O.; Dougherty, D. B.; Sorescu, D. C.; Yates, J. T., Jr. Prog. Surf. Sci. 2010, 85, 206–240. (14) Cossaro, A.; Mazzarello, R.; Rousseau, R.; Casalis, L.; Verdini, A.; Kohlmeyer, A.; Floreano, L.; Scandolo, S.; Morgante, A.; Klein, M. L.; Scoles, G. Science 2008, 321, 943–946. (15) Chesneau, F.; Zhao, J.; Shen, C.; Buck, M.; Zharnikov, M. J. Phys. Chem. C 2010, 114, 7122–7119. (16) Maksymovych, P.; Sorescu, D. C.; Yates, J. T., Jr. Phys. Rev. Lett. 2006, 97, 146103. (17) Yu, M.; Bovet, N.; Satterley, C. J.; Bengio, S.; Lovelock, K. R. J.; Milligan, P. K.; Jones, R. G.; Woodruff, D. P.; Dhanak, V. Phys. Rev. Lett. 2006, 97, 166102. (18) Chaudhuri, A.; Jackson, D. C.; Lerotholi, T. J.; Jones, R. G.; Lee, T. L.; Detlefs, B.; Woodruff, D. P. Phys. Chem. Chem. Phys. 2010, 12, 3229–3238. (19) Kautz, N. A.; Kandel, S. A. J. Am. Chem. Soc. 2008, 130, 6908–6909. (20) Harten, U.; Lahee, A. M.; Toennies, J. P.; W€oll, C. Phys. Rev. Lett. 1985, 54, 2619–2622. (21) Barth, J. V.; Brune, H.; Ertl, G.; Behm, R. J. Phys. Rev. B 1990, 42, 9307–9318. 21803

dx.doi.org/10.1021/jp206457g |J. Phys. Chem. C 2011, 115, 21800–21803