pubs.acs.org/Langmuir © 2010 American Chemical Society
Thiol with an Unusual Adsorption-Desorption Behavior: 6-Mercaptopurine on Au(111) E. Pensa,† P. Carro,‡ A. A. Rubert,† G. Benı´ tez,† C. Vericat,*,† and R. C. Salvarezza† †
Instituto de Investigaciones Fisicoquı´micas Te oricas y Aplicadas (INIFTA) Facultad de Ciencias Exactas, Universidad Nacional de La Plata, CONICET Sucursal 4 Casilla de Correo 16, (1900) La Plata, Argentina, and ‡ Departamento de Quı´mica Fı´sica, Universidad de La Laguna, Avda. Astrofı´sico Francisco S anchez S/N, La Laguna 38071, Tenerife, Spain Received June 15, 2010. Revised Manuscript Received August 23, 2010 A multitechnique study of 6-mercaptopurine (6MP) adsorption on Au(111) is presented. The molecule adsorbs on Au(111), originating short-range ordered domains and irregular nanosized aggregates with a total surface coverage by chemisorbed species smaller than those found for alkanethiol SAMs, as derived from scanning tunneling microscopy (STM) and electrochemical results. X-ray photoelectron spectroscopy (XPS) results show the presence of a thiolate bond, whereas density functional theory (DFT) data indicate strong chemisorption via a S-Au bond and additional binding to the surface via a N-Au bond. From DFT data, the positive charge on the Au topmost surface atoms is markedly smaller than that found for Au atoms in alkanethiolate SAMs. The adsorption of 6MP originates Au atom removal from step edges but no vacancy island formation at (111) terraces. The small coverage of Au islands after 6MP desorption strongly suggests the presence of only a small population of Au adatom-thiolate complexes. We propose that the absence of the Au-S interface reconstruction results from the lack of significant repulsive forces acting at the Au surface atoms.
Introduction In the past 25 years, a great interest has arisen in self-assembled monolayers (SAMs) of thiols on gold because of their multiple applications in nanoscience and nanotechnology.1-5 Thiol SAMs have a key role as building blocks to develop sensors and moleculebased electronic devices in nanofabrication and to protect surfaces against corrosion and friction, among others.1,4 From a basic point of view, thiol SAMs on gold and in particular on Au(111) are considered to be a model system for molecular self-assembly. This selfassembly process, which has been extensively studied, first involves the formation of dilute, lying-down phases, followed by the forma1 Among them, the most importion of dense, standing-up phases. √ √ tant surface structures are the 3 3 R30 lattice and its c(4 2) superlattices, both with surface coverage θ = 0.33.2 The existence of defects in SAMs and the thermal, environmental, and electrochemical stabilities have also been described in the literature.3 The S-Au (thiolate) bond is a strong covalent bond, with adsorption energies in the order of 30-40 kcal mol-1.6 A partial charge transfer (∼0.1 eV) takes place from the Au surface to the S atoms, originating a small dipole on the surface and a lowering in the work function (from 5.25 to 3.75 eV).3 However, the structure of the Au-S interface is still under strong debate.3 There is some evidence that during thiol adsorption the formation of the thiolate *To whom correspondence should be addressed. E-mail: cvericat@ inifta.unlp.edu.ar. Tel: þ54 221 4257430. Fax: þ54 221 4254642. Website: http//nano.quimica.unlp.edu.ar. (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.; Salvarezza, R. C. Phys. Chem. Chem. Phys. 2005, 7, 3258–3268. (3) Vericat, C.; Vela, M. E.; Benı´ tez, G.; Carro, P.; Salvarezza, R. C. Chem. Soc. Rev. 2010, 39, 1805–1834. (4) Gates, B. D.; Xu, Q.; Stewart, M.; Ryan, D.; Willson, C. G.; Whitesides, G. M. Chem. Rev. 2005, 105, 1171–1196. (5) Kind, M.; W€oll, C. Prog. Surf. Sci. 2009, 84, 230–278. (6) Schreiber, F.; Eberhardt, A.; Leung, T. Y. B.; Schwartz, P.; Wetterer, S. M.; Lavrich, D. J.; Berman, L.; Fenter, P.; Eisenberger, P.; Scoles, G. Phys. Rev. B 1998, 57, 12476–12481.
17068 DOI: 10.1021/la102441b
bonds induces a strong reconstruction of the Au(111) surface. The existence of either RS-Au7 or (RS)2-Au8,9 units at the interface has been proposed. The formation of these thiolate-adatom species is supported by two independent pieces of evidence. On one hand, vacancy islands (one or two gold atoms in depth) are formed during thiol adsorption, and this has been claimed to be the main adatom source to reach the required adatom coverage (θAu = 0.33 for RS-Au or 0.16 for (RS)2-Au species).10 On the other hand, after thiol desorption, both in UHV and in liquids, clusters have been imaged by STM, which are believed to be gold islands of monatomic height.3,10 6-Mercaptopurine (6MP) has interest from the point of view of its applications in leukemia treatment11 and therefore has been immobilized on gold nanoparticles with drug delivery purposes.12-15 A few groups have studied 6MP SAMs on Au(111),16-20 and also (7) 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. (8) Gr€onbeck, H.; H€akkinen, H.; Whetten, R. L. J. Phys. Chem. C 2008, 112, 15940–15942. (9) Walter, M.; Akola, J.; Lopez-Acevedo, O.; Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Whetten, R. L.; Gr€onbeck, H.; H€akkinen, H. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 9157–9162. (10) Kautz, N. A.; Kandel, S. A. J. Am. Chem. Soc. 2008, 130, 6908–6909. (11) Karran, P.; Attard, N. Nat. Rev. Cancer 2008, 8, 24–36. (12) Viudez, A. J.; Madue~no, R.; Pineda, T.; Blazquez, M. J. Phys. Chem. B 2006, 110, 17840–17847. (13) Viudez, A. J.; Madue~no, R.; Blazquez, M.; Pineda, T. J. Phys. Chem. C 2009, 113, 5186–5192. (14) Podsiadlo, P.; Sinani, V. A.; Bahng, J. H.; Kam, N. W. S.; Lee, J.; Kotov, N. A. Langmuir 2007, 24, 568–574. (15) Shen, X.-C.; Jiang, L.-F.; Liang, H.; Lu, X.; Zhang, L.-J.; Liu, X.-Y. Talanta 2006, 69, 456–462. (16) Boland, T.; Ratner, B. D. Langmuir 1994, 10, 3845–3852. (17) Madue~no, R.; Sevilla, J. M.; Pineda, T.; Roman, A. J.; Blazquez, M. J. Electroanal. Chem. 2001, 506, 92–98. (18) Madue~no, R.; Pineda, T.; Sevilla, J. M.; Blazquez, M. Langmuir 2002, 18, 3903–3909. (19) Madue~no, R.; Pineda, T.; Sevilla, J. M.; Blazquez, M. J. Electroanal. Chem. 2005, 576, 197–203. (20) Madue~no, R.; Garcı´ a-Raya, D.; Viudez, A. J.; Sevilla, J. M.; Pineda, T.; Blazquez, M. Langmuir 2007, 23, 11027–11033.
Published on Web 10/15/2010
Langmuir 2010, 26(22), 17068–17074
Pensa et al.
Article
on Ag21,22 and Hg18,19 surfaces. In all cases, it has been reported that the molecule adsorbs through the S atom. Moreover, some indirect evidence exists that 6MP is also bonded through one of its N atoms.16,22 Previous STM results have shown that the molecules arrange in ordered lattices.16 In this Article, we have studied SAMs of 6MP on Au(111) by XPS, STM, cyclic voltammetry (CV), and DFT calculations and have compared them with SAMs of hexanethiol on Au(111). 6MP chemisorbs on Au(111), forming short-range ordered domains of molecules and nanosized aggregates. Whereas some Au atom removal at step edges is evident, no vacancy island formation at (111) terraces is observed upon 6MP adsorption. After 6MP reductive desorption, only a small Au adatom island coverage is found, which is not compatible with adatom models.7,8 DFT calculations indicate that the optimized molecule-substrate configuration involves not only S-Au but also N-Au bonds, leading to a relatively high binding energy. In contrast with the case of alkanethiol SAMs on Au(111), Bader analysis indicates that a small positive charge on the topmost surface Au atoms is present. We propose that the absence of the Au-S interface reconstruction results from the lack of significant repulsive forces acting at the Au surface atoms.
Experimental Section Materials. 6MP (Aldrich, 98%), HT (Fluka, 97%), NaOH (Sigma, 98.6%), and absolute ethanol (Carlo Erba, 99.5%) were used as received. Gold Substrates. Evaporated Au films on glass with (111) preferred orientation (AF 45 Berliner Glass KG, Germany) were used as substrates. After soft annealing for 5 min with a hydrogen flame (followed by gentle cooling under nitrogen), these Au substrates exhibit atomically smooth (111) terraces (usually 100-500 nm wide) separated by monatomic or diatomic high steps, as revealed by scanning tunneling microscopy (STM). (See the Supporting Information.) 6MP and HT SAM Preparation. 6MP (HT) SAMs were formed on the Au(111) substrates by immersion of the clean substrates in deaerated 100 μM 6MP (50 μM HT) solutions in ethanol for 24 h. After incubation, the samples were rinsed with the solvent and dried with N2. Then, they were immediately (i) placed in the electrochemical cell for voltammetry, (ii) imaged in air by STM, or (iii) placed in the UHV chamber for X-ray photoelectron spectroscopy (XPS) analysis. Electrochemical Measurements. Cyclic voltammetry (CV) was performed with a potentiostat with digital data acquisition (TEQ, Argentina). The Au(111) substrates (working electrodes) were mounted in a conventional three-electrode glass cell (volume ≈ 100 mL). A large area Pt wire and a saturated calomel electrode (SCE) with a Luggin capillary were used as counter and reference electrodes, respectively. NaOH aqueous solutions (0.1 M) were prepared by the use of deionized H2O (18 MΩ cm) from a Milli-Q purification system (Millipore Products, Bedford) and were degassed with purified nitrogen prior to the experiments. During the measurements, a gentle nitrogen flux was kept above the solution level so that the solutions were always oxygen-free. All measurements were carried out at room temperature (∼22 C). We performed thiol reductive electrodesorption by scanning the potential from -0.2 to -1.4 at 0.1 V s-1 in the 0.1 M NaOH solution. In each case, the charge density involved in the reductive desorption (calculated by integration of the peak area and taking into account the electrode real area from the gold oxide reduction peak) was taken as an indication of the surface coverage by the SAM. Results are an average of >10 measurements. (21) Vivoni, A.; Chen, S.-P.; Ejeh, D.; Hosten, C. M. Langmuir 2000, 16, 3310– 3316. (22) Chu, H.; Yang, H.; Huan, S.; Shen, G.; Yu, R. J. Phys. Chem. B 2006, 110, 5490–5497.
Langmuir 2010, 26(22), 17068–17074
Scanning Tunneling Microscopy. STM imaging was done in air in the constant current mode with a Nanoscope IIIa microscope from Veeco Instruments (Santa Barbara, CA) with commercial Pt-Ir tips. Typical tunneling currents, bias voltages, and scan rates were 0.3 to 0.5 nA, 200-500 mV, and 1-5 Hz, respectively. Surface coverage analysis of islands and vacancy islands was made with the Veeco Nanoscope 5.30 version software using the Bearing tool. X-Ray Photoelectron Spectroscopy. XPS measurements were performed using a Mg KR source (XR50, Specs GmbH) and a hemispherical electron energy analyzer (PHOIBOS 100, Specs GmbH) operating at 10 eV pass energy. A two-point calibration of the energy scale was performed using sputtered cleaned gold (Au 4f7/2, binding energy (BE) = 84.00 eV) and copper (Cu 2p3/2, BE = 932.67 eV) samples. C 1s at 285 eV was used as charging reference. High-resolution spectra were fitted with XPS Peak 4.0 software packages. A Shirley-type background was subtracted from each spectrum. The fitting of the S 2p (Au 4f) peaks was carried out using a spin-orbit splitting of 1.19 eV (3.65 eV) and a branching ratio of 0.5 (0.75). The XPS quantification analysis was done with the CasaXPS v2.3.14 software. A detailed account of the quantification procedure is given in the Supporting Information. Density Functional Theory Calculations. All calculations were performed using plane-wave pseudopotential periodic DFT. The exchange-correlation potential was described by means of the generalized gradient approach (GGA) with the Perdew-Wang (PW91)23 implementation. The one-electron wave functions have been expanded on a plane wave basis set with a cutoff of 420 eV for the kinetic energy. The Brillouin zone sampling was carried out according to the Monkhorst-Pack24 scheme with (9 5 1) dense k-points meshes. The projector augmented wave (PAW) method,25,26 as implemented by Kresse and Joubert,27 has been employed to describe the effect of the inner cores of the atoms on the valence electrons. The energy minimization (electronic density relaxation) for a given nuclear configuration was carried out using a Davidson block iteration scheme. The dipole correction was applied to minimize polarization effects caused by asymmetry of the slabs. All calculations have been carried out using the VASP 4.6 package.27,28 The surface was modeled by a periodic slab composed of five metal layers and a vacuum of ∼12 A˚. Adsorption occurs only on one side of the slab. During the geometry optimization, the two bottom layers were kept fixed at their optimized bulk truncated geometry for the Au(111) surface. The three outermost atomic metal layers as well as the atomic coordinates of the adsorbed species were allowed to relax without further constraints. The atomic positions were relaxed until the force on the unconstrained ˚ atoms was 166 eV; that is, no oxidized S species are present on the Au surface. (See the Supporting Information.) From the S/ Au peak area ratio (corrected for the sensitivity factors), the thiolate coverage results in θ=0.25 ( 0.04 and 0.33 ( 0.04 for 6MP and HT, respectively. In the case of 6MP, we have also recorded the N 1s signal (Supporting Information) and checked that the N/S peak area ratio is ∼4, in good agreement with the molecular formula. In Figure 2, CV profiles recorded at 0.1 V s-1 in 0.1 M NaOH solution are shown for (a) 6 MP and (b) HT on Au(111). Blank voltammograms recorded for clean Au(111) in the same electrolyte are also included as a reference (dotted lines in Figure 2). For 6MP, a sharp reductive desorption peak can be found at -0.68 ( 0.01 V (vs SCE), as previously reported.18,20 The peak charge is 56 ( 4 μC cm-2, which corresponds to a surface coverage θ=0.25 ( 0.02, in good agreement with that obtained from XPS measurements. As is well known, HT yields a desorption peak at -1.07 ( 0.03 V and a charge of 75 ( 7 μC cm-2, in accordance with θ = 0.33. An in-air STM image of a HT-covered Au(111) substrate is shown in Figure 3. The typical vacancy islands (pits) usually found upon alkanethiol SAM formation can be clearly observed 17070 DOI: 10.1021/la102441b
Figure 3. In-air STM images and related data for HT SAMs on Au(111). (a) 200 nm 200 nm image showing vacancy islands (pits). (b) 5.28 nm 5.28 nm high-resolution image of the c(4 2) superlattice. (c) Histogram of pit coverage (from the analysis of >15 images).
(dark regions in Figure 3a). As already mentioned, these features have been regarded as the source of Au adatoms needed to form RS-Au or (RS)2-Au units.3,10 In Figure 3b, a high-resolution image √ of the √ same region shows the “zig-zag” c(4 2) superlattice of the 3 3 R30 lattice.2 The vacancy island coverage (θvac), estimated from large scale STM images, is ∼0.12, as shown in the histogram in Figure 3c. In-air STM images of 6MP-covered Au(111) substrates are shown in Figure 4a,b. In this case, there is no evidence of vacancy island formation, as reported for 2-amino purinethiol on Au(111).29 Instead, terraces are decorated with bright spots that tend to align (29) Sawaguchi, T.; Sato, Y.; Mizutani, F. Electrochemistry 2001, 69, 962–965.
Langmuir 2010, 26(22), 17068–17074
Pensa et al.
Figure 4. (a-e) In-air STM images and related data for 6MP
SAMs on Au(111). (a) 200 nm 200 nm image showing aligned islands. Small and large islands are indicated with green and gray arrows, respectively. (b) 300 nm 300 nm image showing aligned islands and serrated steps (yellow arrows). (c) Cross section of the red line in part a, where two different island heights can be observed. (d) Island height bimodal distribution (from the analysis of >150 islands). (e) Small island coverage (from the analysis of more than 40 images).
along certain directions, sometimes with 120 angles (Figure 4b). The fact that 6MP is chemisorbed on Au(111) as a thiolate, as observed √ from XPS and CV data, is consistent with the lifting of the 22 3 Au(111) “herringbone” reconstruction, as seen with many other adsorbates.30-33 The observed distances between aligned rows do not seem to directly correlate with the herringbone reconstruction. However, because the islands are aligned following preferred directions, we cannot discard the fact that the herringbone reconstruction could act as some kind of template. Note that these features are not observed in clean Au(111) surfaces. (See the Supporting Information.) A detailed analysis of the island height (cross section in Figure 4c) indicates the presence of two kinds of structures, as reflected in the height histogram (Figure 4d), which shows a bimodal distribution function. The brighter islands (gray arrow in Figure 4a) are better defined and have a more irregular shape. These features, whose coverage is ∼0.02, are 0.24 nm high (Figure 4c), that is, the interplanar distance in Au(111), and thus can be regarded as Au islands. The majority of the islands, though, are smaller and more diffuse (green arrow in Figure 4a) and have a smaller height (
