ARTICLE pubs.acs.org/Langmuir
In Situ STM Imaging of Bis-3-sodiumsulfopropyl-disulfide Molecules Adsorbed on Copper Film Electrodeposited on Pt(111) Single Crystal Electrode HsinLing Tu, PoYu Yen, Sihzih Chen, and ShuehLin Yau* Department of Chemistry, National Central University, Jhongli, Taiwan 320, ROC
Wei-Ping Dow* Department of Chemical Engineering, National Chung Hsing University, Taichung 40227, Taiwan, ROC
Yuh-Lang Lee* Department of Chemical Engineering, National Cheng Kung University, Tainan 70101, Taiwan, ROC
bS Supporting Information ABSTRACT: The adsorption of bis-3-sodiumsulfopropyldisulfide (SPS) on metal electrodes in chloride-containing media has been intensively studied to unveil its accelerating effect on Cu electrodeposition. Molecular resolution scanning tunneling microscopy (STM) imaging technique was used in this study to explore the adsorption and decomposition of SPS molecules concurring with the electrodeposition of copper on an ordered Pt(111) electrode in 0.1 M HClO4 þ 1 mM Cu(ClO4)2 þ 1 mM KCl. Depending on the potential of Pt(111), SPS molecules could react, adsorb, and decompose at chloride-capped Cu films. A submonolayer of Cu adatoms classified as the underpotential deposition (UPD) layer at 0.4 V (vs Ag/AgCl) was completely displaced by SPS molecules, possibly occurring via RSSR (SPS) þ ClCuPt f RSPtþ þ RS (MPS) þ Cu2þ þ Cl, where MPS is 3-mercaptopropanesulfonate. By contrast, at 0.2 V, where a full monolayer of Cu was presumed to be deposited, SPS molecules were adsorbed in local (4 4) structures at the lower ends of step ledges. Bulk Cu deposition driven by a small overpotential (η < 50 mV) proceeded slowly to yield an atomically smooth Cu deposit at the very beginning ( 0.2 V, followed by overpotential (OPD) stage, resulting in monolayer to multilayer of Cu, respectively. Within the UPD regime, the Cu adatoms are likely to be partially charged and a monolayer of chloride anions is coadsorbed to compensate for the charge. UPD of Cu on Pt(111) in Cl-containing medium proceeds in √ √ two steps, producing a (4 4)- and then a ( 7 7)-Cu þ Cl bilayer structure.21 Shifting the potential to E < 0.05 V triggered bulk Cu deposition. To observe the adsorption of SPS, a Cu film was first deposited on Pt(111) by potential control, followed by adding SPS into the STM cell. The STM scanner was a typical A-head (Veeco, Santa Barbara, CA) with a maximal scan size of 500 nm. The tip was a tungsten tip etched by AC in 6 M KOH. Nail polish was applied to coat the whole tip for insulating purposes. The potential of the tip electrode and the feedback current were set normally at 0.4 V and 1 nA, respectively. The electrochemical cell used for voltammetry had a three-electrode configuration, including a Ag/AgCl reference electrode and a Pt counter electrode. The potentiostat was a CHI 600 (Austin, TX). Suprapure perchloric acid (HClO4) and potassium chloride (KCl) were purchased from Merck (Darmstadt, DFG). Triple-distilled water (resistivity = 18.3 MΩ 3 cm, Lotun Technology Co., Taipei) was used to prepare the electrolytes. Copper perchlorate (Cu(ClO4)2) and SPS were obtained from Aldrich
ARTICLE
Figure 1. Cyclic voltammograms recorded at 50 mV/s with Pt(111) electrode immersed in 0.1 M HClO4 þ 1 mM KCl þ 0.1 μM SPS. The largely featureless CV profile shown in (a) indicates that the Pt(111) electrode was passivated by a layer of SPS-derived molecules. Panel (b) reveals Cu deposition on Pt(111) in the same electrolyte þ 1 mM Cu(ClO4)2. The red trace was obtained in the absence of SPS. (St. Louis, MO) and Rashing (GmbH, FRG), respectively. The use of STM to explore electrified interface is reviewed.23
’ RESULTS AND DISCUSSION Cyclic Voltammetry. Shown in Figure 1a are cyclic voltammetry (CV) profiles recorded at a potential scan rate of 50 mV/s with Pt(111) electrode in 0.1 M HClO4 þ 1 mM KCl þ 1 μM SPS. It is mostly featureless between 0 and 0.7 V. Processes such as adsorption of hydrogen normally occurring between 0.1 and 0.1 V were largely eliminated, although H2 evolution was substantial as indicated by the precipitous increase of current at 0.1 V. Thus, SPS or its derivatives such as MPS produced upon SPS adsorption on Pt(111) effectively passivated the electrode, denying access of water molecules and the production of OH-like species. This CV profile reminisces those found with Pt(111) electrodes coated with hexanethiol, benzenethiol, and more importantly MPS.24,25 According to previous studies, each SPS molecule could decompose to give two MPS molecules upon its adsorption on metallic substrates, such as Pt, Au, Cu, and Ag, producing thiolate species on these metallic supports.26 Reductive stripping, which has been effective in giving insight into the chemical nature of thiolate adsorbed on gold electrode,2729 would not work with the Pt(111) electrode, because thiol admolecules usually cling to the Pt electrode so strongly that they would not desorb even at the onset potential of water reduction in alkaline solution.24,25 Given the featureless CV profile seen here, the SPS molecule is unlikely reduced at potential positive of 0.1 V. Figure 1b shows CV profiles revealing Cu deposition on Pt(111) electrode in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 without (red) and with (black) 0.01 mM SPS. The adsorption of SPS molecules apparently resulted in substantial differences in copper deposition, where the UPD peaks at 0.3 and 0.42 V shifted negatively to 0.1 V (A/A0 ), and bulk copper deposition (B/B0 ) occurring at E < 0 V was reduced by 50%. These features resemble those observed with MPS-modified Pt(111) electrode, which is anticipated as SPS should yield MPS upon its contact with the Pt(111) electrode. The above results were obtained with bare Pt(111), which are compared with Pt(111) precoated with sub- to monolayer of Cu adatoms. The Pt(111) electrode was first loaded with Cu at 0.2 and 0.4 V each in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2, followed by adding SPS into the cell (final [SPS] = 10 μM). The potential was then ramped positively after holding 6802
dx.doi.org/10.1021/la200250m |Langmuir 2011, 27, 6801–6807
Langmuir
Figure 2. CV profiles recorded at 50 mV/s with Pt(111) electrode immersed in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 (red traces). The CV profiles in black were recorded 5 min after SPS was added in the cell under potential control at 0.2 V (a) and at 0.4 V (b).
potential at 0.2 or 0.4 V for 5 min. The resultant CV profiles are shown in Figure 2, where the arrows mark the potential for SPS dosing. The process of Cu UPD on Pt electrode has been extensively examined, showing the formation of a Cu þ Cl bilayer structure.2022 In particular, the two well-defined peaks at 0.30 and 0.43 V correspond to partial and complete √ √ stripping of Cu, which manifest in restructuring from ( 7 7)R19.1 to (4 4)-Cu þ Cl structure, and subsequently to another undefined chloride adlayer.21 These CV results shown in Figure 2 indicate that adding SPS markedly alters the CV profiles. Peaks C and D shifted positively by þ30 and 10 mV to produce peaks C0 and D0 , respectively. Both peaks broadened noticeably. Peaks C0 and C contained roughly the same amount of charge, but the intensity of peak D0 was substantially attenuated. Peak D appeared to be affected more by SPS dosing than peak C. To substantiate this result, a Pt(111) electrode was first loaded with a submonolayer of Cu at 0.4 V, followed by exposure to SPS for 10 min and the potential was swept positively from 0.4 to 0.7 V. The result shown in Figure 2b indicates that the stripping peak D seen previously at 0.43 V was no longer there, which implies that Cu adatoms deposited at 0.4 V were completely displaced by SPS molecules. These results speak to different electrode processes occurring at Pt(111) at 0.2 and 0.4 V. It is thought that potential influenced greatly the interaction between Cu adatoms and Pt substrate. At E > 0.4 V, the CuPt interaction could be too weak to hold Cu adatoms against the competition from SPS adsorption. Although the Cu adlayer produced at 0.2 V was not displaced by SPS, it is not immediately clear if the capping chloride adlayer stayed. This issue might be resolved by studying the potentials of the UPD peaks, which are known to be sensitive to the chemical nature of the Cu-capping adlayer. As seen in Figures 1 and 2, the Cu UPD feature is a pair of broad peaks near 0.15 V for MPS- or SPS-modified Pt(111) electrode, but a sharp doublet for Clcoated Pt(111). In other words, if the capping chloride was replaced by MPS, for example, the UPD peak should change substantially. However, peaks C and C0 seen in Figure 2a appeared at approximately the same potentials, suggesting that
ARTICLE
the chloride adlayer stayed after it contacted with SPS at E < 0.3 V. The þ40 mV potential shift from C to C0 (Figure 2a) could mean that SPS molecules were adsorbed on the chloride adlayer. This is supported by STM results presented below. In Situ STM Imaging of SPS Adsorption on Pt(111) Precoated with UPD of Cu. Shown in Figure 3a are STM images obtained at 0.4 V in 0.1 M HClO4 þ 1 mM Cu(ClO4)2 þ 1 mM KCl. The deposited Cu adlayer hardly affected the atomically smooth surface morphology of Pt(111), as indicated by a (4 4) structure seen in the inset in Figure 3a. The images in Figure 3b and c were obtained 1 and 10 min after SPS was added into the STM cell. The most distinct features in these STM images are attributed to pits amid the (4 4) superlattice. These features grew in number and in size as time elapsed. The (4 4) structure shrank from 90% to 10% in 10 min and eventually was eliminated within 15 min, and a decidedly disordered adlayer emerged afterward. The rate of this reaction could be faster elsewhere, because the imaged area could be blocked by the tip. These STM observation showing the elimination of the (4 4)-Cu þ Cl bilayer could be explained by two possible reactions. First, the partially charged Cu adatoms could interact with the Pt electrode so weakly that they were simply displaced by SPS molecules, which decomposed to MPS subsequently upon its contact with the Pt electrode. Alternatively, there could be a redox reaction outlined as RSSR þ ClCuPt f RSPtþ þ RS þ Cu2þ þ Cl, where RSSR and RS represent SPS and MPS molecules, respectively. Because this reaction ended with a disordered adlayer, it is not possible to determine if it was indeed MPS molecules adsorbed on Pt(111). Nevertheless, it is worthwhile noting that SPS would react with CuCl to produce Cuthiolate and Cu2þ species in acidic aqueous solution, proposed as 4Cuþ þ SPS = 2Cu2þ þ 2Cuþ-thiolate.30 This reaction scheme agrees with our view in that the Cu deposit on Pt(111) formed at 0.4 V was similar to CuCl species. The same STM imaging method was used to examine √ the adsorption of SPS at 0.2 V, at which a Pt(111)( 7 √ 7)R19.1Cu þ Cl adlayer was preinstalled.21 Shown in Figure 4 are STM images obtained shortly after SPS was added (final [SPS ] = 1 μM). The inset of Figure 4a√is a high-resolution √ STM scan of the most prominent Pt(111)( 7 7)R19.1 Cu þ Cl structure. Although pits were also seen, they never grew larger with prolonged STM imaging, which contrasts sharply with the results shown in Figure 3. Instead, some locally ordered structures were found at the lower end of the step ledge running vertically in the middle of the STM image (Figure 4a). Those protruded islands barely grew with prolonged STM imaging. The internal structure of one of these ordered patches was discerned by higher resolution scans shown in Figure 4b. It is largely hexagonal marked by white dots. Each dot represents one SPS molecule, and rows of molecules were aligned in the main axis of the Pt(111) substrate indicated by arrows. Each SPS molecule was imaged as a V-shaped protrusion as indicated, and two neighboring SPS molecules were separated by 1.2 nm, or four times that of the lattice spacing of the Pt(111) substrate (dnn = 0.278 nm). These STM results yield a (4 4) structure marked by the rhombus in Figure 4b. The V-shaped molecular appearance implies a folded conformation of SPS admolecules, as also noted by others.14 Judging from the higher corrugation height of those ordered patches (Δz = 0.1 nm) with respect to the neighboring terrace, those SPS patches likely resided on the chloride adlattice. 6803
dx.doi.org/10.1021/la200250m |Langmuir 2011, 27, 6801–6807
Langmuir
ARTICLE
Figure 3. In situ STM images acquired with Pt(111) electrode precoated with a submonolayer Cu at 0.4 V in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2. The first image was obtained before the addition of SPS, followed by (b) and (c) acquired 1 and 10 min after adding 0.1 μM (final concentration) SPS. The inset in panel (a) is an atomic-resolution STM image (4 4 nm) of the (4 4)-Cu þ Cl bilayer. Nearly all the ordered structure due to Cu þ chloride was displaced by SPS in 10 min. The tunneling resistance was 90 MΩ.
In Situ STM Imaging of Bulk Copper Deposition on Pt(111). Shifting the potential of the Pt(111) electrode negatively
Figure 4. In situ STM images acquired under conditions the same as those of Figure 3, except the potential of Pt(111) was held at 0.2 V. Protruded islands at the lower end of the step ledge emerged 2 min after the addition√of 0.1 μM √ SPS into the STM cell. The inset in (a) shows the Pt(111)( 7 7)R19.1Cu þ Cl structure. A set of arrows indicates the main axis of Pt(111). (b) Internal molecular structure of one of the islands seen in (a). Each SPS molecule exhibits as a V-shaped protrusion, suggesting folded molecular conformation. This ordered array is characterized as (4 4) marked as a rhombus. The tunneling resistance was 90 MΩ.
Unfortunately, we have not found an appropriate imaging condition that could be used to image the upper SPS√structures √ and the underlying chloride lattice simultaneously. The ( 7 7)R19.1 chloride structure is presumed to stay. These ordered arrays were sensitive to the imaging condition used to achieve high-resolution imaging, as they could be removed by the tip if it was positioned too close to the substrate, as noted also in previous studies with a Cu(100) single-crystal electrode.14,15 Clear imaging of the SPS islands was possible at 90 MΩ tunneling resistance, but lowering it to 60 MΩ caused damage to the adlayer, as illustrated by an STM image shown in the Supporting Information. Nonetheless, proper imaging conditions allowed STM imaging of these SPS arrays for hours. This STM result contrasts sharply with those reported earlier showing sparsely adsorbed or even nonadsorbed SPS molecules on Cl-covered Cu(100).14,15 Because the Cu film produced here on Pt(111) most likely adopted a fcc (111) orientation, as compared with Cu(100) used in previous studies, it is not impossible that the orientation of the copper electrode could affect the adsorption of chloride and the subsequent nucleation of SPS. In addition, the copper monolayer supported by Pt(111) could be electronically different from a bulk Cu crystal used by others. The concentration of SPS is anticipated to be another factor to its adsorption. This issue awaits further investigation.
from 0.2 to 0.08 V in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 þ 0.1 μM SPS initiated bulk Cu deposition. A series of STM images were acquired consecutively to reveal the process of copper deposition (Figure 5). The same morphology seen with the imaged area guarded against thermal drift in the process of imaging. Figure 5a acquired at 0.2 V (before the switch of potential) reveals a typical Pt(111) surface with smooth terraces spanning hundreds of nanometers and steps aggregated at the upper portion of the image. At this stage, the Pt(111) electrode was decorated with a Cu þ Cl bilayer having tiny vacancy defects on terraces. Figure 5b was obtained 30 s after the switch of potential, where islands emerged at the lower ends of steps and on terraces. These features grew laterally, along with the appearance of new seeds on terraces, as revealed by the subsequent STM images shown in Figure 5ce recorded in a period of 10 min. All islands appear to be 0.2 nm or monatomic higher than their neighboring terraces. It is worthwhile noting that the second layer of copper grew solely with minimal nucleation of the third layer of copper. But this layered growth finally gave way to simultaneous nucleation of multiple layers as the copper film thickened. The surface became rougher with time. A typical topography STM scan shown in Figure 5f was acquired with roughly 10 Cu layers, estimated from the depth (2.2 nm) of the crater seen in the middle of the image. This layer-then-island growth mode is anticipated for Cu on Pt, because of lattice strain occurring at the interface. All nuclei assumed flakelike morphology with atomically smooth terrains but evidently rugged perimeters. As Cu nuclei expanded laterally with time, two neighboring domains eventually coalesced to produce a larger grain of Cu. However, this coalescence of two Cu nuclei was usually not seamless, but it left pits unfilled in each Cu layer. These features could result from strains due to unlike lattice constants between the copper deposit and the Pt(111) substrate. The grain size of Cu crystallite tended to decrease with the thickness of Cu deposit or the surface roughness increased with the thickness of Cu deposit. Switching to high-resolution STM scans on bilayer or multilayer Cu deposit on the Pt(111) electrode (counting the UPD layer as the first layer) resulted in their structural details (Figures 6 and 7). The image shown in Figure 6a was obtained at the beginning of second layer Cu deposition with image conditions of 130 mV in bias voltage and 1.7 nA in tunneling current (or a tunneling resistance of 76 MΩ). Protruded islands (Δz = 0.2 nm) with well-defined striped patterns formed near a step site. The 6804
dx.doi.org/10.1021/la200250m |Langmuir 2011, 27, 6801–6807
Langmuir
ARTICLE
Figure 5. Time-dependent in situ STM images acquired with a Pt(111) electrode after its potential was stepped from 0.2 to 0.08 V in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 þ 0.1 μM SPS. The first image was acquired before the potential was switched, followed by image (b) obtained 1 (b), 2 (c), 3 (d), 5 (e), and 35 min (f) after bulk Cu deposition at 0.08 V. The Cu film seen in (f) is 10 layers thick, as estimated from the depth of the valley seen in the middle of (f). All images are 300 300 nm2.
Figure 6. In situ STM images acquired with a Pt(111) electrode at 0.08 V in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 þ 0.1 μM SPS. These images were obtained at the beginning of bulk Cu deposition, featuring the production of Cu islands at a step site. (b) Higher resolution STM scan revealing a striped pattern on the island. This striped pattern was not stable at high resolution scan and could change to a different structure shown in (c) and (d). This structure could be adsorbed chloride lying below SPS admolecules.
stripes seen here were aligned in three directions (rotated by 60) marked by arrows, which happened to be the Æ110æ directions of the Pt(111) substrate. Further higher resolution scans acquired with a similar tunneling resistance of ∼80 MΩ revealed details of the atomic ar-
rangement (Figure 6b), which features paired lines of protrusions (Δz = 0.06 nm). Two paired lines are separated by 0.71 nm from each other, and their spacing to the next lines is either 1.1 or 1.5 nm. These locally ordered arrays could be identified as SPS molecules oriented parallel to the substrate. The two sulfonate groups of a SPS molecule are responsible for the paired protrusions seen by STM. Some singular spots were also imaged, and they could be MPS molecules. Two sulfonate groups within a SPS molecule could measure up from 0.55 to 1.1 nm, depending on its molecular conformation. The 0.71 nm measured here by the STM is much shorter than the value of the most stable molecular structure in the unsolvated state, meaning that SPS admolecules could be folded here. In addition to solvation, the SPS conformation could be affected by its interaction with the Cu þ Cl bilayer. This striped pattern was subject to variations in high resolution STM scans, as another ordered array without the striped pattern was imaged. This new structure is shown in Figure 6c, an unfiltered STM image of 8 8 nm2. A portion of the filtered image is shown in Figure 6d. This ordered array is mostly square with an edge length of 0.36 ( 0.2 nm. This square array could be reverted back to the striped pattern. These STM results suggest that either SPS admolecules drifted to different sites or selective imaging determined by the scanning probe. It is worthwhile noting that adsorbed thiol molecules such as dimethyl disulfide and 1-octanethiol on Cu(111) could force reconstruction of the uppermost Cu layer from a hexagonal to a square array, or the pseudo-Cu(100) structure.3134 The present square seen in Figure 6d could be the pseudo-Cu(100) lattice, but the dimension of the present square pattern of 0.36 nm does not match with that (0.417 nm) observed for the pseudo-Cu(100). The√spatial structure seen in Figure 6d was determined to be (2 3)rect represented by the rectangle marked there. The average edge lengths of the unit cell are 0.52 and 0.45 nm, as compared with the expected values of 0.56 (0.50) and 0.48 (0.43) nm for Pt (Cu). Thus, the second copper layer could have a lattice constant lying between those of Pt (0.278 nm) and Cu 6805
dx.doi.org/10.1021/la200250m |Langmuir 2011, 27, 6801–6807
Langmuir
ARTICLE
Figure 7. Ball model in side (a) and top (b) view showing stacked layers of SPS, chloride, and bilayer copper on Pt(111), which is proposed to account for the STM results shown in Figure 6. To simplify the top view in (b), only SPS admolecules, chloride layer (deep blue), and copper (light blue) are shown. SPS admolecules could lie parallel to the chloride plane.
Figure 8. In situ STM images (40 nm2 (a) and 8 nm2 (b)) acquired with a Pt(111) electrode at 0.1 V in 0.1 M HClO4 þ 1 mM KCl þ 1 mM Cu(ClO4)2 þ 0.1 μM SPS. These images reveal the smooth morphology of the deposited copper film and the ordered (2 2) SPS adlattice marked by the rhombus in (b).
(0.25 nm). The quality of our present STM results however could not yield a definitive measurement here. This assignment is based on the assumption that the two layers of copper sitting underneath the chloride adlayer adopted hexagonal close packed structures with copper adatoms aligned in the main axis of Pt(111). The surface structure seen in Figure 6 can be reconciled by a ball model shown in Figure 7a (side view) and b (top view). The former outlines the stacking sequence of SPS, Cl layer (deep blue), and bilayer Cu (light blue) on the Pt(111) substrate. In the top view, √SPS admolecules are adsorbed in 1D striped pattern on a (2 3)rect-chloride structure sitting on a Cu(111)-like network. Since this structure differs greatly from those reported for chloride adsorbed on Cu(111),3537 it is likely then the adsorbed SPS molecules were responsible for this hitherto unreported structure of chloride adsorbed on Cu(111). This view is supported by the fact that the 0.71 nm spacing measured for the paired lines (Figure 6b) √ happens to match the length of the diagonal axis of the (2 3)rect unit cell. This result contrasts markedly with those found with SPS adsorbed on Cl-coated Cu(100), where chloride adsorbates are arranged in c(2 2), unchanged by the adsorption of SPS molecules. This effect of substrate orientation on SPS adsorption would have to be resolved by conducting the same experiment with Pt(100), where a Cu(100)-like thin film can be established for studying the adsorption of SPS. Also it is mentioned here that a number of STM studies have addressed the adsorption of porphyrin molecules on iodine-modified Au(111) electrode.38,39 Presumably, the adsorption of SPS on Cl-capped Cu thin film is analogous to these systems, where SPS or porphyrin molecules interact with halide adlayers via the van der Waals interaction.
From the third layer up, the paired lines of protrusions were no longer seen. Rather, STM revealed different structures sitting atop the Cu deposit. One of the typical structures is illustrated by the STM results presented in Figure 8. The thickness of the Cu film in this case was six layers, estimated from its height with respect to a neighboring depression. Given a time resolution of seconds for our STM instrument, it would be difficult to see more action in the course of Cu reduction. Clear STM imaging was possible only when Cu deposition ceased and the system reached a steady state. The scan shown in Figure 8a reveals an adlayer sitting atop the atomically smooth Cu deposit. Although this adlayer was not highly ordered, its internal structure was discerned at some local areas (Figure 8b). This lattice is mostly hexagonal with protrusions aligned in the main axis of the Pt(111) substrate and a nearest neighbor spacing of 0.5 nm, twice the tentative Cu spacing of 0.25 nm. These results suggest a (2 2) adlattice. Coincidently, a (2 2) structure was also observed in our previous study with MPS molecules adsorbed on multilayer Cu film supported by Pt(111).19 Subsequently, it is thought that SPS molecule was more prone to decomposition, yielding MPS at more negative potential, as this reaction could be outlined as (RSSR þ 2e f 2RS). At 0.2 V, this reaction might not be significant enough, so that the adsorption of SPS molecules predominated to yield a (4 4) structure (Figure 4). The structure seen in Figure 8b can be described by the ball model shown in Figure 8c, where MPS molecules adsorb to a chloride layer which caps the Cu deposit on Pt(111). This view is consistent with a report showing that MPS could be the species responsible for the acceleration of copper electrodeposition.11 6806
dx.doi.org/10.1021/la200250m |Langmuir 2011, 27, 6801–6807
Langmuir
’ CONCLUSIONS In situ STM has yielded insight into the electrodeposition of copper on a Pt(111) electrode in acidic solution containing SPS and chloride. Partially charged Cu adatoms deposited on Pt(111) in the UPD region could not coexist with SPS molecules, as they could react via a reaction of RSSR þ ClCuPt f RSPtþ þ RS þ Cu2þ þ Cl, where RSSR and RS represent SPS and MPS, respectively. By contrast, the Cu þ Cl bilayer structure installed at E < 0.2 V (the end of UPD) was stable against SPS adsorption, affording local (4 4)-SPS structure preferentially at the lower end of step defects. Proper imaging conditions were needed to image these SPS patches; they could be destroyed by the STM tip, otherwise. These results contrast markedly with those reported for SPS adsorbed on chloride-covered Cu(100) electrode, where SPS admolecules dispersed randomly and drifted on the substrate. The orientation of the copper substrate, (100) or (111), could influence the strength and the spatial structure of SPS admolecules. Also, as SPS molecules are adsorbed in disarray on Cu(111) in 0.1 M H2SO4, it seems then that anions, chloride or bisulfate, could influence the adsorption of SPS molecules. Layered type Cu deposition was possible at a small overpotential (
