Article pubs.acs.org/JPCC
Low-Temperature Adsorption of Carbon Monoxide on Gold Surfaces: IR Spectroscopy Uncovers Different Adsorption States on Pristine and Rough Au(111) Jan Pischel*,† and Annemarie Pucci†,‡ †
Kirchhoff Institute for Physics, Im Neuenheimer Feld 227, D-69120 Heidelberg, Germany Centre for Advanced Materials, Kirchhoff Institute for Physics, Im Neuenheimer Feld 227, D-69120 Heidelberg, Germany
‡
S Supporting Information *
ABSTRACT: The morphology of gold surfaces plays a major role in many domains of contemporary research. Infrared (IR) spectroscopy in combination with carbon monoxide (CO) as a probe adsorbate is able to sensitively monitor differences in the morphology of gold surfaces on an atomic level if CO adsorption on the various surfaces is clarified. Our investigation comprises the first study of CO adsorption on Au(111) under well-defined ultrahigh vacuum conditions at 30 K. We find that CO adsorbs on Au(111) in atop geometry, as has been reported before for a variety of gold surfaces, but a significantly higher frequency of the internal CO stretching vibration is observed, confirming results from recent theoretical studies. Furthermore, the presence of a submonolayer amount of gold adatoms on the Au(111) surface results in the properties of gold surfaces toward CO adsorption at higher temperatures known from the literature. Step-wise annealing of these atomically rough surfaces leads to a gradual transition between the literature case and the behavior observed on the pristine Au(111) surface at 30 K.
■
INTRODUCTION
methods able to characterize morphological but also electronic and chemical properties of metal surfaces. With the advent of scanning tunneling microscopy (STM) in the early 1980s,41 a well-suited method had been found to analyze the morphology of surfaces of grown gold films.42−52 However, its application is restricted to more or less flat surfaces and, if atomic resolution is to be achieved, to ultra high vacuum (UHV) conditions. A different approach that does not suffer from these restrictions is the use of infrared vibrational spectroscopy in combination with CO as a probe adsorbate to gain information on the chemical properties of available sites at the surface. This is possible since the frequency of the internal CO stretching vibration is in general highly sensitive to the details of bonding of the molecule to the substrate. (The gasphase frequency of the internal CO stretching vibration is 2143.4 cm−1.53) This effect is traditionally described by the Blyholder model54 where the extent of weakening of the vibration is related to the amount of charge transfer from the metal into the antibonding 2π* lowest unoccupied molecular orbital (so-called back-bonding). In a more detailed picture, hybridization of the 2π* orbital with metal electronic states and
The morphology-related properties of gold surfaces are of fundamental importance in many areas of current research. One example is the use of gold nanoparticles for plasmonic applications such like surface enhanced vibrational spectroscopies,1−4 where not only the crystalline quality of the bulk5,6 but also surface morphology7,8 and roughness9−13 determine characteristic properties of the plasmonic excitation. Additionally, the degree of surface roughness can largely influence properties of an (eventually self-assembled) adsorbate layer. This is especially true for the functionalization by thiol-based self-assembled monolayers.14−20 A second example where the morphology of gold surfaces does matter are catalytically active gold nanoparticles supporting, e.g., the oxidation of carbon monoxide (CO) or hydrogen:21,22 The turnover frequency of the catalyzed reaction has been shown to crucially depend not only on the support23−29 but also on the size and morphology of the gold nanoparticles.30−37 The implication of gold surface properties does not stop here but plays an important role also in gold surface-electrodes for usage in a Paul trap for confinement of metal ions38 or in the details of charge carrier injection in organic electronic devices,39,40 to mention just two further examples. This furnishes evidence for the importance of © 2015 American Chemical Society
Received: May 27, 2015 Revised: July 17, 2015 Published: July 20, 2015 18340
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
terrace sites. While for copper the first case is observed,63−66 platinum exhibits the opposite behavior.66 A so-called “walleffect”72,73 has been proposed to explain this observation. For gold, up to now, such a relation between frequencies of CO molecules bound in the same adsorption geometry to substrate atoms with different coordination has not yet been established. This can be traced back to the apparent impossibility to discriminate between frequencies from CO adsorbed at differently coordinated sites of gold surfaces (see Table 1). In
donation of the (likewise hybridized and antibonding) 5σ* highest occupied molecular orbital to empty metal states near the Fermi edge would have to be taken into account. Although it may fail for certain systems such as c(4 × 2)CONi(111),55 the determination of the adsorption geometry is in principle possible within the framework of a generally accepted classification scheme proposed by Sheppard and Nguyen:56 The lowest frequencies in the range of 1650 to 1880 cm−1 are expected for CO adsorbed in a hollow-site; CO in a bridge site absorbs at 1880 to 2000 cm−1, and atop-adsorbed CO leads to absorption at 2000 to 2130 cm−1. In a recent DFT-study,57 this relation has been shown to be valid for adsorption on Au(111) as well, although due to the very weak interaction of CO with that surface, the frequencies are in the range of or even slightly beyond the upper limits for each geometry. In addition to this crude analysis, even more information can be gained from more subtle changes in the vibrational frequency: On real surfaces and even on many ideal single crystal surfaces, a number of inequivalent atoms at the surface must be distinguished. They can be classified according to their coordination number, i.e., the number of nearest neighbors. On an fcc(111) surface, the ideal coordination number of an atom in the topmost layer is 9. Single adatoms but also atoms at steps or kinks with a reduced number of nearest neighbors are therefore referred to as under-coordinated or low-coordinated surface atoms. Due to their special electronic properties, these sites are in general believed to exhibit enhanced chemical reactivity and to chemically activate adsorbates.32,58−62 Copper is an example for a metal where absorption of CO adsorbed on a single crystal either at a regular terrace site or at a defect-like position at a step edge or kink could be distinguished by means of vibrational spectroscopy.63−66 It must be stressed that this is possible although CO is adsorbed atop a surface atom in either case, which reflects the fact that the stretching vibration is sensitive to the coordination of the concerned substrate atom. More generally spoken, the local charge state of the substrate can be monitored, a fact that is widely exploited in IR investigations of nanocatalysts.67 Several effects complicate the straightforward interpretation of collected IR data: First, the CO molecule as a Lewis acid can modify the circumstances at the surface under investigation by adsorption,68 which in that case makes it impossible to determine properties of the clean surface. Next, the observed frequencies may strongly depend on surface coverage. The associated frequency shifts can be decomposed into two contributions that are referred to as (physical) dipole shift and (substrate-mediated) chemical shift, respectively. While increasing coupling between parallel molecular dipoles via the respectively generated electromagnetic fields will always lead to a frequency blue-shift with increasing coverage,69,70 the competing for charge transfer with the substrate can result in a frequency shift in either direction. Since the original Blyholder model always predicts a blue-shift with increasing coverage,54 a modification has been proposed to account for the experimental observations.64 Dipole coupling between different adsorbed species whose frequencies are close enough can lead to a variety of unexpected effects such as intensity transfers between modes or even virtual peaks if the adsorbate forms islands on the surface.66,71 Finally, for a given metal, it is a priori not clear whether (usually stronger) bonding to under-coordinated defect-like sites leads (at fixed coverage and for the same adsorption geometry) to a frequency increase or decrease compared to
Table 1. Reported Frequencies νCO (in cm−1) of the CO Stretching Vibration on Different Gold Substratesa substrate, experimental conditions
reference
νCO, low coverage
νCO, high coverage
gold film, RT gold film, 115 K gold film, RT gold film, 2 K Au(110), RT Au(110)-1 × 2, 110−250 K Au(111), RT Au(111), RT sputtered Au(111), 120 K sputtered Au(111), 110 K Au(210), wet-chemically Au(211), 85 K Au(332), 92−105 K
74 75 76 77 78 79 80 81 82 83 84 85 86
2120 2120 2120 2125 2110 2118 2060 2080 2125 2124 2110 2126 2124
2110 red-shifted 2115 2110 2110 2108 2060 2080 2115 2114 2110 2112 2110
a
Substrate temperatures during exposure are indicated. Roomtemperature (RT) experiments were performed at varying CO partial pressures from 0.1−100 mbar. The low-temperature experiments were performed under UHV conditions at correspondingly low partial pressures.
fact, for all investigated gold surfaces, irrespective of the substrate and the experimental conditions, only one single absorption peak around 2120 cm−1 at low coverage shifting to 2110 cm−1 with increasing coverage is observed in the IR spectra. This similarity translating also into the thermal desorption behavior has already been pointed out by Yim et al.82 who could explain it by showing theoretically that the degree of under-coordination does not have a big effect on the frequency of the CO stretching vibration. (In contrast to the IR results, two desorption peaks are well distinguishable in TPD spectra. However, their associated temperatures again barely vary from substrate to substrate.) The slight scattering of the experimental data can easily be attributed to the differing definitions of “low” and “high” coverage in the literature. An exception must be made for nanocatalysts on oxide supports where coordination of CO to (partially) charged Au-atoms or clusters is possible.87 The only two available IR investigations of successful CO adsorption on pristine Au(111) were performed at room temperature and at elevated pressures.80,81 In these studies, restructuring of the surface was observed. This could explain the deviation from the otherwise uniform and consistent image. Contamination by nickel carbonyls could be an issue as well.88 However, the authors exclude nickel contamination on the basis of XPS data. In this paper, we show that the pristine Au(111) surface allows for adsorption of a CO species exhibiting a significantly different stretching frequency than was reported for all the other gold surfaces. Our results are in qualitative (refs 80 and 89) and even in quantitative (ref 57) agreement with frequencies obtained in previous studies based on density 18341
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C functional theory (DFT).57,80,89 To the best of our knowledge, our paper is the first to report on an IR investigation of the lowtemperature adsorption of CO on Au(111) under UHV conditions. Thereby we close one of the remaining blank areas of the IR-map of CO adsorbed on low-index surfaces of transition metals. Under similar conditions, only the atomic strucure of the system has been investigated before.90 To relate our result obtained for the smooth surface to the literature data, we additionally investigated adsorption of CO on the Au(111) surface roughened on an atomic scale by cold-deposition of small amounts of gold adatoms in the submonolayer range. As far as we know, the morphology of such nanostructured gold surfaces at cryogenic temperatures has not been investigated yet. Some collected information on nanostructured Au(111) at higher temperatures may be found in the Supporting Information. To demonstrate that the roughening procedure really leads to the predominant presence of under-coordinated surface atoms, we have performed the analogous experiment with a Cu(110) substrate. IR spectra of CO adsorbed on the pristine Cu(110) surface are in agreement with previous IR results from the literature.64,65 Roughening of the surface by cold-deposition of additional copper atoms establishes the properties of rough surfaces as can be observed in the associated IR spectra of adsorbed CO. The spectra obtained on the smooth and on the rough Cu(110) surface are given in the Supporting Information. It might be interesting to note that further information on some of the aspects of this paper can be found in the Ph.D. thesis of one of the authors (J.P.).91
99.999%) from a molybdenum crucible with a rate of around 5× 10−2 ML/min. It is noteworthy that the experiments on both kinds of substrates were performed within the same UHV chamber and on the same crystal ensuring the best possible comparability. The experimental results were reproducible under the same conditions. In addition to the two substrates described in the previous paragraph, we investigated CO adsorption on annealed substrates. These were prepared by allowing the Au/Au(111) sample to anneal at a certain temperature between 30 and 473 K. For annealing temperatures below room temperature, evaporation was performed at 30 K sample temperature and the sample was cooled down immediately after reaching the desired annealing temperature. Otherwise, gold was evaporated at room temperature, and the sample was held for 5 min at the respective annealing temperature. In all cases, 0.33 ML of additional gold atoms were deposited. The IR spectra R(ω) were taken in reflection geometry with p-polarized light and near grazing incidence (θ = 83°) to optimize the absorption signal,94 the device resolution was set to 2 cm−1. We present relative reflectance spectra R(ω)/R0(ω), which are normalized by a reference spectrum R0 of the COfree substrate at 30 K. In the case of the modified substrates, the reference spectrum was taken from the substrate as exposed to CO, i.e., after evaporation and after an eventual annealing step. All spectra of the CO-covered surfaces were taken in situ during exposure of the cooled surface to gaseous CO (purity of 99.997%, Messer Griesheim). The necessary partial pressure in the 10−9 mbar range was realized by backfilling of the chamber through a leak valve. The given exposures are those determined at the end of the respective measurement thereby providing an upper limit for the exposure corresponding to a given spectrum. The time to record one spectrum with 400 Scans was about 3 min corresponding to an exposure range of approximately 0.1 Lc. One Lc is defined as an exposure of 1.33 × 10−6mbar s corrected for the relative sensitivity of 1.07 of the ion gauge for CO.95 At the end of each experiment, the valve was closed and the sample held at constant temperature. No CO desorption from the sample was observed during this stage of the experiment as indicated by the constancy of the measured IR reflection− absorption (IRRA) spectra. Finally, the sample cooling was stopped, and the sample gradually heated up to above 200 K. Also during this stage, spectra were taken allowing for an estimation of the desorption temperatures.
■
EXPERIMENTAL SECTION All experiments were carried out in a UHV chamber described elsewhere92 using the commercially available Fourier transform IR spectrometer IFS 66 v/S (Bruker Optics, Germany) in combination with a liquid nitrogen cooled mercury−cadmiumtelluride detector. The base pressure was 2× 10−10 mbar or below. The gold single crystal (MaTeck) was circular with a diameter of 10 mm and a thickness of 2 mm. The nominal alignment of the surface was (111) ± 0.1°. The sample was mounted on a sample holder made out of oxygen-free high conductivity copper in a way minimizing the probability of surface contamination from the sample holder. After insertion into the UHV, the crystal was prepared by repeated cycles of Ar+ ion sputtering (ca. 1× 10−6 A/cm2, 5−10 min, angle against the surface normal θ ≈ 50°) and subsequent annealing to approximately 560 °C (30−45 min). Prior to the first experiment, the sample went through more than 30 of such cycles until a sharp LEED-pattern with low background was observed (Omicron SpectaLEED S/N 566). Directly before each experiment, the sample was prepared by a double cycle where the second sputtering was performed at approximately 200 °C. This treatment leads to a particularly smooth surface.93 Next, the sample was cooled down to low temperatures around 30 K by use of a liquid helium cryostat the sampleholder was attached to. In this study, we investigated CO adsorption on two types of surfaces: the as-prepared pristine Au(111) surface, and the same surface modified by colddeposition of additional 0.33 ML of gold where 1 ML is defined by the area density of a perfect single close-packed Au(111) layer. Due to suppressed diffusion at cryogenic temperatures, the latter substrate denoted as Au/Au(111) in the following exhibits roughness on an atomic scale. Deposition was realized using an electron beam evaporator (Omicron EFM 3T) to sublime gold (“metals basis” Premion by Alfa Aesar, purity
■
RESULTS For sufficiently high exposures, condensation of CO in the multilayer occurs. This is an essential point in that it proves that we observed complete filling of the first monolayer. Therefore, we start by summarizing the most striking features of this purely condensed species. We then proceed with description of the chemisorbed first layer, which was in the focus of the present investigation. Condensed Phase: Multilayer Adsorption. If exposure is increased beyond some 3 Lc, independently of the presence of surface roughness and an eventual annealing temperature, the CO-stretching vibration νc appropriate for the condensed phase appears in the spectra at 2142.7 cm−1 (see below). It should be noted that this is exactly the frequency that is expected to be observed in an IRRAS experiment on the basis of the dielectric function of solid CO from ref 96. It is not expected at the frequency of 2138.3 cm−1, which is observed in IR transmission 18342
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C experiments96,97 (see Supporting Information for further discussion). The close proximity of the IRRAS frequency to the gas phase frequency of the free CO molecule (2143.4 cm−1)53 is purely accidental and not exact. We were also able to observe a slight redshift for νc of about 2 cm−1 in the second monolayer (first condensed layer) with respect to the higher layers as has been observed for CO condensation on Cu(111),98 Cu(100)99 and Pt(111).100 Due to the strength of νc, the variation of its measured intensity with exposure cannot be expected to be linear, even if the thickness of the condensed layer does grow linearly with exposure. Therefore, this mode is not suitable for a quantitative treatment proving that the absorption signal really stems from the condensed multilayer and not from a weakly physisorbed species in the first layer. The inspection of the much weaker absorption of the naturally contained 12C18O and 13C16O isotopes and of a characteristic combination mode around 2200 cm−1, however, unambiguously leads to the conclusion that a film of polycrystalline CO in the α-phase97,101 has been formed on the surface (see also Supporting Information). After the end of CO exposure, no desorption of the multilayer-related features is observed while the pressure is dropping down. Very shortly after stopping the sample cooling, however, the multilayer desorbs. This proves that the temperature of the Au(111) crystal during exposure is slightly below the desorption temperature of the multilayer, which should be more or less independent of the substrate, and which has been determined to be 32 K on Au(110)-(1 × 2).102 We conclude that prior to multilayer adsorption, the Au(111) surface must be completely covered by CO. Although the possibility that CO adsorbs on Au(111) facets at low partial pressures is sometimes completely denied,88 this is of course only a reasonable statement for a given temperature range. In fact, there is a bunch of theoretical studies based on DFT indicating exothermal adsorption of CO on Au(111) without even taking dispersion forces into account.57,80,103,104 The growth of islands might additionally stabilize the (incomplete) monolayer due to attractive interaction between adjacent molecules.90 Chemisorption of the First Layer. Small Exposures ≲ 2 Lc. In Figure 1a, we present IRRA spectra of different amounts of CO chemisorbed within the first layer on the rough Au/ Au(111) surface. Note that the spectra are not shifted artificially, but that a real shift of the baseline is caused by the chemisorbing CO molecules, which act as scatterers on the gold conduction electrons near the sample surface.105−107 This is the case for all spectra shown in the present paper. This shifting of the baseline stops as multilayer adsorption sets in (see Figure 2). For all exposures up to 1.4 Lc, one single well-defined band denoted as νr is observed on the rough surface. Its frequency red-shifts with increasing exposure from the initially observed 2126 cm−1 to 2115 cm−1. The overall observation on the rough Au/Au(111) surface is consistent with available literature on CO adsorption on gold surfaces (see Table 1). This is especially true for CO adsorbed on a sputtered Au(111) surface (see refs 82 and 83). A comparison of the scaled and shifted absorption peak at 0.12 Lc (black dashed line in Figure 1a) with the line observed at 1.4 Lc reveals that part of the intensity increase with increasing exposure is transferred to the wings. For all exposures, the absorption band exhibits a distinct asymmetry with a tail at the long wavelength side extending to wavenumbers significantly below 2100 cm−1 for higher
Figure 1. IRRA spectra of varying amounts of CO adsorbed at 30 K on (a) Au/Au(111) and (b) pristine Au(111). The exposures in units of corrected Langmuirs Lc are 0, 0.03, 0.12, 0.31, 0.57, 0.94, 1.4 (a) and 0.007, 0.10, 0.19, 0.28, 0.37, 0.55, 0.84, 1.2, 1.8, 2.5 (b). Note the different ordinate scales. The frequency range where vibrations νr of CO at defect-like sites are observed is marked red; the higher frequency range where vibrations νp of CO at characteristic sites of the pristine Au(111) surface are observed is marked blue. The dashed black line in panel a is the spectrum at 0.12 Lc scaled by a factor of 9.4 and shifted by 9.3 cm−1. The vertical dashed line gives the peak position for atop adsorbed CO on Au(111) predicted by DFT calculations (kBT = 0; ref 57). The illustrations depict proposed adsorption geometries of the respective species.
exposures, as typical for dipole−dipole coupling in a disordered layer.69 Our simulation in the Supporting Information shows that, due to this coupling, it is nearly impossible to detect eventually present further low-frequency species at the surface by means of IR spectroscopy. The situation is differently complex on the pristine Au(111) surface as shown in Figure 1b, where the IRRA spectra of chemisorbed CO are depicted. For the lowest exposure, three absorption peaks are visible at 2111, 2119, and 2130 cm−1, respectively. While the lowest-frequency peak does not play a role for higher exposures, the middle peak whose frequency lies in the range of νr remains discernible up to the highest exposures. The spectra of the smooth Au(111) surface, however, are dominated by the absorption around νp = 2130 cm−1, which was completely absent in the spectra of Figure 1a. Therefore, we attribute this absorption feature to species adsorbed at sites that are inherently related to the pristine Au(111) surface. This will be discussed in detail in the next section. For low coverages, νp experiences a slight redshift of 2 18343
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
about 2 Lc, but instead some very weak absorption feature develops in the range of νp. Such a subsequent population of sites with different adsorption energies has already been proposed in the literature.85,86,88,108 Only after saturation of νp is the condensed species detected for exposures higher than 3 Lc, as in the case of the pristine Au(111) surface. The doublepeak nature of the absorption is considerably better resolved on the rough substrate due to less overlap with νp. On both substrates, multilayer adsorption starts at around 3 Lc. Obviously, the IR-optical properties of the condensed phase do not depend on the substrate roughness as can be inferred from inspection of the inset where the IRRA spectra of comparable exposures are shown on both substrates. The ratio of the absorption intensities at νr for the pristine and the rough surfaces around 5% is in the range reported in ref 83 ,where it has been attributed to adsorption at monatomic steps. Alternatively, a lifting of the reconstruction by release of additional adatoms out of the compressed reconstructed first gold layera mechanism that has been proposed for adsorption of other molecules or atoms on reconstructed Au(111) surfaces109−112should result in a comparable absorption intensity. Although lifting of the reconstruction upon CO adsorption has not been observed at 8 K by STM,90 this could still be possible at temperatures around 30 K. A further striking feature of the depicted spectra after complete filling of the first monolayer is that the peaks attributed to chemisorbed species are still significantly modified by CO molecules condensing in the second monolayer. This results in a further red-shift accompanied by some kind of distortion of the peak shape and a reduction of absorption intensity of the chemisorbed species. The change in frequency and the reduced absorbance can easily be explained by the increasing dielectric background influencing the measured quantity, namely, the loss function Im(ϵ−1(ω)) (see also Supporting Information). The distortion might be due to the steric interaction between the first (chemisorbed) and second (condensed) monolayers. Finally, the chemisorption features do not change anymore at exposures above 6 Lc. Since this is twice the exposure needed to fill the first layer, this exposure can be related to the completely filled second monolayer. Obviously, condensation in the third and higher monolayers does not influence the IR response of the chemisorbed species, which is a reasonable finding and supports the interpretation. CO Adsorption on Annealed Rough Substrates. The pristine and the rough Au(111) surfaces as described thus far can be interpreted as two limiting cases for rough gold surfaces annealed to a certain temperature (the annealing temperature being 30 and 833 K, respectively). Additionally, we investigated CO adsorption on surfaces with intermediate annealing temperatures of 220 K, 295 K, 373 K, and 473 K (see Experimental Section for further details on preparation). The recorded IRRA spectra are shown in Figure 3 for three distinct exposures. Obviously, the spectra exhibit a rich variety of features related to different adsorption sites. Additionally, the observed vibrational frequencies are influenced by dipole coupling between different species. The number of observed peaks depends on the annealing temperature of the substrate (which leads to a change in the portion of available adsorption sites) as well as on exposure. It varies from 1 (for the rough Au/ Au(111) surface at low exposures) to at least 4 for the intermediate annealing temperatures. We explain this by the
Figure 2. IRRA spectra of varying amounts of CO adsorbed at 30 K on (a) Au/Au(111) and (b) pristine Au(111). The exposures in units of corrected Langmuirs Lc are 1.4, 1.5, 1.6, 2.0, 2.5, 3.0, 3.5, 3.9, 4.4, 4.8, 5.3, 5.8, 6.2, 6.7 (a) and 2.5, 2.9, 3.3, 3.6, 4.0, 4.4, 4.8, 5.1, 5.5, 5.9, 6.3 (b). In addition to the colored ranges from Figure 1, the region where absorption of the condensed phase occurs is marked green. The inset compares the spectra in this region for exposures of 6.7 Lc CO/Au/ Au(111) (dotted) and 6.3 Lc CO/Au(111) (dashed).
cm−1 from its initial 2130 cm−1, which is followed by subsequent blue-shift of 4 cm−1 such that it ends up at 2132 cm−1 for the complete first monolayer. As in the case of CO adsorbed on Au/Au(111), we attribute these weak coveragedependent shifts to a combination of dipole coupling and substrate-mediated chemical interaction. Higher Exposures (≲7 Lc) near the Full Monolayer. When exposure is increased above the range for which spectra are shown in Figure 1, new absorption peaks occur in the spectra (Figure 2). In the case of the pristine Au(111) surface as substrate (Figure 2b), a peak whose frequency saturates at 2142.7 cm−1 shows up, indicating the onset of multilayer growth (see discussion above). A detailed analysis of the peak shape reveals that this peak actually contains a further contribution at about 2140.6 cm−1, which manifests itself as a shoulder on the long-wavelength side of the peak (see also inset of Figure 2). This shoulder can be ascribed to the CO molecules condensed in the second monolayer as explained before. On the rough Au/Au(111) surface, multilayer condensation of CO does not set in immediately after saturation of νr at 18344
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
Figure 3. IRRA spectra of CO adsorbed at 30 K on Au/Au(111) preannealed to temperatures of 220 K, 295 K, 373 K, and 473 K. Spectra are shown for exposures of approximately (±10%) 1, 3, and 6 Lc, respectively. The ordinate scale is the same for all spectra and for all exposures. The spectra for adsorption on the pristine Au(111) surface (bottom) and on rough Au/Au(111) (top) are shown for comparison as well. The regions of νr, νp, and νc as derived from the previous results are marked in red, blue, and green, respectively.
Figure 4. Integrated absorption areas as defined in the text for all investigated annealed surfaces from Figure 3. Integration has been restricted to the indicated wavenumber ranges related to νr, νp and νc, respectively. The vertical dashed lines in the left panel at 0.75, 1.15, and 1.80 Lc illustrate the shifting saturation exposure of νr on the different surfaces. Note that the ordinate of the middle panel has been scaled by a factor of 4.
interplay of the sequential filling of different adsorption sites and the dipole coupling between these species. The spectra in Figure 3 reveal the gradual transition from the adsorption behavior of the rough Au/Au(111) surface to that of the pristine Au(111) surface with increasing annealing temperature. This transition is accompanied by a gradual decrease of absorption intensity in the region of νr which is partially compensated by an increase of mode νp. It is noteworthy that evaporation of only 0.33 ML of additional gold on the pristine Au(111) surface (thus on the perfect substrate for epitaxial growth) at room temperature and at extremely low deposition rate leads to strong deviations from the smooth surface. This explains the difficulties to prepare atomically smooth gold films at room temperature.42−44 Even after annealing to 473 K for 5 min, significantly more defect-like
places are observed as monitored by the enhanced intensity of νr as compared to the pristine Au(111) surface. We also have evaluated the absorbances on all surfaces in the ranges of νr (2020−2127 cm−1), νp (2127−2139.3 cm−1), and νc (2139.3-2160 cm−1) by integrating ∫ dω [y0(ω) −R(ω)/ R0(ω)] over the indicated wavenumber ranges, respectively. y0(ω) is the baseline assumed to vary linearly between R(2020 cm−1)/R0(2020 cm−1) and R(2160 cm−1)/R0(2160 cm−1) . The ranges are (except for the very upper and the very lower limit) the regions that are marked by colors in Figures 1, 2, and 3. This evaluation procedure only slightly suffers from peaks extending into a neighboring interval. The result is plotted in Figure 4. The gradual transition from the rough Au/Au(111) surface to the pristine Au(111) surface with increasing annealing temperature is clearly reflected by the amount of absorption in the range of νr. A similar (but inverted) trend is 18345
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C observed in the region of νp. Only the surface with an annealing temperature of 220 K does not follow this trend. This can be explained by the especially strong overlap between bands in the range of νr and νp leading to intensity transfer to νp by dipole coupling (see Figure 3). The nonmonotonic behavior in the region of νp at low exposures on the roughest surfaces can be explained by the contribution of νr in that region: When νr shifts to lower wavenumbers, the integrated absorption decreases in the range of νp until the population of species νp increases at higher coverages giving rise to an increasing absorption.
the DFT calculations from ref 82: Actually such high desorption temperatures have been predicted already for adsorption at single adatoms attached to a step. If coordination of the gold surface atoms is decreased further as is the case for free adatoms generated in the stochastic growth regime, the desorption temperature should lie above 200 K as observed. Our hypothesis is further supported by the series of adsorption experiments on preannealed roughened Au(111) surfaces: the clear trend that the intensity of modes in the region of νr is decreasing with increasing annealing temperature, whereas the intensity of νp is increasing, is expected if the defect-like adsorption sites anneal successively by enhanced gold diffusion with increasing temperature to minimize surface energy. In the case of intermediate annealing temperatures, it is not clear whether each observed absorption band can be directly related to one adsorbed species or if dipole coupling leads to a splitting and thereby to a multiplication of observed modes. The intriguing interplay between subsequent decoration of available adsorption sites and dipole interactions as deduced from the behavior in Figure 3 is most obvious if the in situ IRRA spectra of the adsorption systems are plotted as a function of exposure. Such contour plots can be found for the various annealing temperatures in the Supporting Information. The similarity of the spectra of CO on pristine Au(111) and the surface annealed to 473 K together with the marked differences to the surface annealed only to 373 K shows that in the temperature range between 373 and 473 K, a transition in the self-diffusion behavior of gold atoms occurs. This temperature range is in agreement with previous studies.49,114−116 Interestingly, the peak area of νr in the spectrum of the full first monolayer on Au/Au(111) is larger by a factor of 3 to 4 than that of νp in the spectrum of the full first monolayer on Au(111) (see Figure 4). Since the number of adsorption sites isif at allonly weakly increased by our method, we have to attribute the stronger absorption on the rough surface to properties of the adsorbed species itself. (This is a clear advantage of our method over introducing surface roughness by sputtering. Since we deposited only 0.33 ML of gold, an upper limit for the increase of the number of adsorption sites is 33%. However, since addition of an adatom buries original adsorption sites, the number of adsorption sites may even be reduced: If only atop adsorption occurs as our spectra suggest, then three original terrace sites will be replaced by one adatom site. The independence of the onset of multilayer adsorption on surface roughness can be explained consistently by assuming an invariant sticking factor and an invariant number of available sites per unit area.) These properties must then be related to special adsorption sites that are generated by the evaporation step and that differ from regular adsorption sites in the way they interact with the adsorbate. Such sites are known to play a crucial role in the so-called chemical enhancement of surfaceenhanced Raman spectroscopy (SERS),117,118 where they are made responsible for an increase of the Raman signal mediated by the creation of electron−hole pairs and transient chargetransfer. The same effect on IR absorption has been reported for ultrathin iron films in the presence of atomical roughness (see Supporting Information for further details). It is noteworthy that the maximum absorbance of νr in the case of the rough Au/Au(111) is reached before complete filling of the first monolayer takes place at 3 Lc (see Figure 4). Such a behavior is expected for the case of strong dipole coupling (see, e.g., p. 322 of ref 70) emphasizing once more the
■
DISCUSSION For simplicity, we will henceforth use the symbols νc, νr, and νp to designate the species corresponding to the observed frequencies as discussed above. We start the discussion with some general conclusions that can be derived from the above observations: On both the smooth and the roughened Au(111) surfaces, we were able to observe absorption in a frequency range where the internal stretching vibration is expected for a very weakly adsorbed CO molecule. The so-called metal surface selection rule in IRRAS82,94 states that this is only possible if the adsorbate is oriented in a way that a significant component of the related dynamic dipole moment is parallel to the surface normal. In the case of CO, the dynamic dipole moment of the stretching vibration is parallel to the molecular axis. Furthermore, the observed frequencies clearly lie within the range for atopadsorbed species according to the mentioned classification scheme.56 For symmetry reasons, it is therefore very likely that CO adsorbs on both the smooth and the rough Au(111) surface with its molecular axis (nearly) parallel to the surface normal. Our findings suggest that the higher of the observed frequencies, νp around 2130 cm−1, is related to special sites of the pristine surface, while the lower frequency, νr at about 2115 cm−1, which appears when additional adatoms are deposited on the surface and which is also observed on other more open or polycrystalline samples (cf. Table 1) seems to be typical for defect-like under-coordinated adsorption sites. This interpretation is supported by the order of CO decoration of the adsorption sites. Due to stronger bonding to undercoordinated sites, such a behavior is expected since CO molecules are mobile on the surface at temperatures as low as 30 K. There are two additional observations corroborating our interpretation: First, species νp desorbs at significantly lower temperatures when cooling is stopped and the CO-covered samples rewarm gradually. This can be monitored by in situ IRRA spectroscopy. We estimate the desorption temperature for νp to lie between 60 and 70 K, whereas species νr remains at the surface up to temperatures above 200 K. In accordance with that desorption temperature, Kim et al. observed on Au(211) in a TPD experiment85 a CO desorption peak at 190 K and another one at 140 K. Similar findings have been reported by Ruggiero and Hollins for CO adsorption on Au(332)86 and by Yim at al. for a variety of gold surfaces.82 It should be noted that even the desorption peak at lower temperature must be assigned to species adsorbed at under-coordinated sites as has been shown by Yim at al. 82 Since it is generally accepted26,61,82,83,85,102,113 that bonding to defect-like sites is much stronger, the observation that νp desorbs at much lower temperatures is in line with our interpretation. The very high desorption temperature of νr can be understood on the basis of 18346
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
the signal is very weak and only visible if a very good signal-tonoise ratio is achieved (see Figure 2). At this point we do not want to conceal that there is a very thorough investigation published by Dumas, Tobin, and Richards that might seem to contradict our results. In that study, the authors examined CO adsorption at 2 K on gold films deposited at various temperatures TD on a sapphire substrate.77 For TD < 290 K, the infrared response was in very good agreement with our results on Au/Au(111). For TD > 290 K however, only the ice peak at 2143 cm−1 was observed without any indication for a band in the range of νp. This apparent contradiction can be explained as follows: Obviously, the effective surface area of the cold-deposited film is much larger than that of the room temperature-deposited film (the authors estimate a factor between 20−50 on the basis of analog experiments with silver films). Therefore, the response of CO adsorbed at smooth Au(111) facets would correspondingly be suppressed to the same extent if the absorbance was the same for the two species. Additionally, as discussed before, the signal from CO at under-coordinated sites is stronger by a factor of up to 4 than that from terrace sites. This makes an eventually adsorbed species at terrace sites undetectable within the noise of the spectra presented in ref 77. The early onset of multilayer growth that was observed (and which was partially attributed to physisorption in the first monolayer) can be explained by the drastically reduced number of available sites as well. One final remark concerns the quality of the gold films grown on sapphire substrate without adhesion layer: From today’s perspective, it seems clear that even the gold films annealed to 150 °C exhibited a considerable degree of roughness and that the presence of reconstructed Au(111) facets can be excluded.51,52,122,123 In summary, we do not believe that the results of our study are in contradiction with the results presented in ref 77, but rather that they complete them.
important role of this mechanism for layers of CO adsorbed on gold surfaces. Finally, the behavior of mode νc displays in a very nice fashion that onset at 3 Lc and evolution of condensation in the multilayer are independent of the morphology of the substrate. This proves that the sticking coefficient for CO on gold at 30 K does not depend on the coordination of available substrate sites and that it is very close to unity. We return now to the prototypal cases of the pristine and rough Au(111) surfaces: It is important to clarify that the peak that we attribute to adsorption at the smooth surface cannot stem from multilayer adsorption since it is observed already at the lowest coverages investigated, which are far below one complete monolayer. It might be possible, however, that the long-range reconstruction of the Au(111) surface plays a role for adsorption of this species. It is also important to note that the extraordinarily high frequency of species νp cannot be attributed to any contamination effect: If this peak was due to CO adsorbed at substrate impurities, it would be impossible to quench it completely by only 0.33 ML of gold adatoms on the surface. As this quenching is observed, we must assign the species νp with the higher CO stretching frequency between 2128 and 2132 cm−1 to sites that are inherently related to the pristine Au(111) surface while CO adsorbed at undercoordinated sites gives rise to a lower frequency between 2110 and 2125 cm−1. This behavior is a key result from our study and is actually exactly opposed to the case of CO adsorbed on single crystalline copper surfaces where CO species adsorbed stronger at under-coordinated atop sites give rise to a higher stretching frequency.66 While the trend observed on gold has up to now been unreported and is inverted as compared to copper, it is in agreement with theoretical predictions stating that the CO stretching frequency should be highest for CO adsorbed at Au(111) terrace sites.80,89 Furthermore, our observed frequency νp≈ 2130 cm−1 on the pristine Au(111) surface is in good quantitative agreement with recent DFT calculations57 of the CO stretching vibration for atop adsorption on Au(111) indicating a frequency of 2142 cm−1. With this respect, it is important to mention that the authors of ref 57 regret the poor agreement of their calculated frequency for bridge-site adsorption on Au(111) with the experimental result obtained on a sputtered Au(111) surface82 and a gold film electrode.119 From an experimental point of view, it is undoubted that CO adsorbs in an atop geometry on gold surfaces under UHV conditions as indicated unambiguously by the CO stretching vibration above 2100 cm−1 (see Table 1). Therefore, their obtained frequency for atop adsorption should be compared to our results and this comparison is quite convincing as mentioned above. Since the Au(111) surface is the thermodynamically favored one120,121 and therefore usually makes a major contribution to the facets of gold films and nanoparticles, it might seem surprising that the higher CO stretching frequency νp has not been reported before. Actually, due to dipole−dipole interaction, this mode should even be enhanced by intensity transfer from lower-frequency vibrations at defect sites as observed for copper66,70 (where high-frequency signals from defect sites dominate already at low defect concentrations). However, the complete absence of the band can be expected in the majority of published studies on CO adsorption on gold due to the temperatures significantly below 77 K needed for adsorption of the species. Additionally, on the rough substrate,
■
SUMMARY AND CONCLUSIONS We presented the first low-temperature IR investigation of CO adsorption on the pristine Au(111) surface under UHV conditions. Our results indicate that CO chemisorbs on the surface in an atop geometry with its molecular axis oriented preferentially perpendicular to the surface. Although interaction with the surface is very weak (we estimate desorption to occur between 60 and 70 K), it leads to a significant redshift (as compared to the condensed CO phase) of the CO stretching frequency down to about 2130 cm−1. This frequency depends only weakly on coverage (±2 cm−1) and is in good agreement with a recent theoretical prediction for atop adsorbed CO.57 Interestingly, this frequency could be well separated from CO adsorbed at under-coordinated substrate sites (with a coverage dependent CO stretching frequency of 2125 to 2110 cm−1). This is in contrast to other gold surfaces investigated thus far, where it was impossible to discriminate between different adsorbed species on the basis of infrared spectra. The pristine Au(111) surface modified by small amounts of gold adatoms in the submonolayer range exhibits the same properties for adsorption of CO as observed in other IR investigations at higher temperatures on a variety of single crystalline and polycrystalline gold surfaces. In accordance with the interpretation of previous TPD experiments,82 our investigations lead to the conclusion that all species observed on gold surfaces up to now under similar conditions but at higher temperatures must be interpreted as CO adsorbed at under-coordinated sites like steps or kinks. 18347
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
■
Properties of Plasmonic Nanoparticles. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 081412. (11) Rodríguez-Fernández, J.; Funston, A. M.; Pérez-Juste, J.; Á lvarez Puebla, R. A.; Liz-Marzán, L. M.; Mulvaney, P. The Effect of Surface Roughness on the Plasmonic Response of Individual Sub-Micron Gold Spheres. Phys. Chem. Chem. Phys. 2009, 11, 5909−5914. (12) Barchiesi, D.; Kessentini, S. Roughness Effect on the Efficiency of Dimer Antenna Based Biosensor. Adv. Electromagnetics 2012, 1, 41− 47. (13) Trügler, A.; Tinguely, J.-C.; Jakopic, G.; Hohenester, U.; Krenn, J. R.; Hohenau, A. Near-Field and SERS Enhancement from Rough Plasmonic Nanoparticles. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 165409. (14) Creager, S. E.; Hockett, L. A.; Rowe, G. K. Consequences of Microscopic Surface Roughness for Molecular Self-Assembly. Langmuir 1992, 8, 854−861. (15) Guo, L.-H.; Facci, J. S.; McLendon, G.; Mosher, R. Effect of Gold Topography and Surface Pretreatment on the Self-Assembly of Alkanethiol Monolayers. Langmuir 1994, 10, 4588−4593. (16) Lee, M.-T.; Hsueh, C.-C.; Freund, M. S.; Ferguson, G. S. Air Oxidation of Self-Assembled Monolayers on Polycrystalline Gold: The Role of the Gold Substrate. Langmuir 1998, 14, 6419−6423. (17) Tsuneda, S.; Ishida, T.; Nishida, N.; Hara, M.; Sasabe, H.; Knoll, W. Tailoring of a Smooth Polycrystalline Gold Surface as a Suitable Anchoring Site for a Self-Assembled Monolayer. Thin Solid Films 1999, 339, 142−147. (18) Hoogvliet, J. C.; Dijksma, M.; Kamp, B.; van Bennekom, W. P. Electrochemical Pretreatment of Polycrystalline Gold Electrodes To Produce a Reproducible Surface Roughness for Self-Assembly: A Study in Phosphate Buffer pH 7.4. Anal. Chem. 2000, 72, 2016−2021. (19) Tkac, J.; Davis, J. J. An Optimised Electrode Pre-Treatment for SAM Formation on Polycrystalline Gold. J. Electroanal. Chemi. 2008, 621, 117−120. (20) Feng, G.; Niu, T.; You, X.; Wan, Z.; Kong, Q.; Bi, S. Studies on the Effect of Electrode Pretreatment on the Coverage of SelfAssembled Monolayers of Dodecanethiol on Gold by Electrochemical Reductive Desorption Determination. Analyst 2011, 136, 5058−5063. (21) Haruta, M.; Yamada, N.; Kobayashi, T.; Iijima, S. Gold Catalysts Prepared by Coprecipitation for Low-Temperature Oxidation of Hydrogen and of Carbon Monoxide. J. Catal. 1989, 115, 301−309. (22) Haruta, M. Gold as a Novel Catalyst in the 21st Century: Preparation, Working Mechanism and Applications. Gold Bulletin 2004, 37, 27−36. (23) Liu, H.; Kozlov, A. I.; Kozlova, A. P.; Shido, T.; Asakura, K.; Iwasawa, Y. Active Oxygen Species and Mechanism for LowTemperature CO Oxidation Reaction on a TiO2-Supported Au Catalyst Prepared from Au(PPh3)(NO3) and As-Precipitated Titanium Hydroxide. J. Catal. 1999, 185, 252−264. (24) Schubert, M. M.; Hackenberg, S.; van Veen, A. C.; Muhler, M.; Plzak, V.; Behm, R. J. CO Oxidation over Supported Gold Catalysts−“Inert” and “Active” Support Materials and Their Role for the Oxygen Supply during Reaction. J. Catal. 2001, 197, 113−122. (25) Liu, Z.-P.; Hu, P.; Alavi, A. Catalytic Role of Gold in Gold-Based Catalysts: A Density Functional Theory Study on the CO Oxidation on Gold. J. Am. Chem. Soc. 2002, 124, 14770−14779. (26) Molina, L. M.; Hammer, B. Some Recent Theoretical Advances in the Understanding of the Catalytic Activity of Au. Appl. Catal., A 2005, 291, 21−31. (27) Wang, J. G.; Hammer, B. Oxidation State of Oxide Supported Nanometric Gold. Top. Catal. 2007, 44, 49−56. (28) Risse, T.; Shaikhutdinov, S.; Nilius, N.; Sterrer, M.; Freund, H.-J. Gold Supported on Thin Oxide Films: From Single Atoms to Nanoparticles. Acc. Chem. Res. 2008, 41, 949−956. (29) Liu, X. Y.; Wang, A.; Zhang, T.; Mou, C.-Y. Catalysis by Gold: New Insights into the Support Effect. Nano Today 2013, 8, 403−416. (30) Bamwenda, G. R.; Tsubota, S.; Nakamura, T.; Haruta, M. The Influence of the Preparation Methods on the Catalytic Activity of Platinum and Gold Supported on TiO2 for CO Oxidation. Catal. Lett. 1997, 44, 83−87.
ASSOCIATED CONTENT
S Supporting Information *
Further information on experimental details, the morphology of nanostructured Au(111), CO adsorption on smooth and nanostructured Cu(110), CO adsorption on smooth and nanostructured iron films and chemical enhancement, remarks concerning the frequency of the CO stretching vibration in the solid phase as observed in IRRAS experiments, intensity of isotope bands with increasing exposure, influence of the dielectric background on intensity and position of an IRreflectance-absorption band, combination modes in the condensed phase, asymmetry of CO/Au/Au(111) and 3DIRRA spectra of CO adsorbed on the various substrates as a function of exposure. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.5b05051.
■
AUTHOR INFORMATION
Corresponding Author
*Phone: +49 6221 54-9891; E-mail: jpischel@kip.uniheidelberg.de. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS J.P. would like to thank the State of Baden-Württemberg for a Scholarship within the graduate college “Connecting Molecular π-Systems into Advanced Functional Materials”. During the time of the study, he was a member of the Heidelberg Graduate Academy and of the Heidelberg Graduate School of Fundamental Physics.
■
REFERENCES
(1) Aroca, R. F.; Alvarez-Puebla, R. A.; Pieczonka, N.; SanchezCortez, S.; García-Ramos, J. V. Surface-Enhanced Raman Scattering on Colloidal Nanostructures. Adv. Colloid Interface Sci. 2005, 116, 45−61. (2) Neubrech, F.; Pucci, A. Plasmonic Enhancement of Vibrational Excitations in the Infrared. IEEE J. Sel. Top. Quantum Electron. 2013, 19, 4600809. (3) Stiles, P. L.; Dieringer, J. A.; Shah, N. C.; Van Duyne, R. P. Surface-Enhanced Raman Spectroscopy. Annu. Rev. Anal. Chem. 2008, 1, 601−626. (4) Suzuki, M.; Niidome, Y.; Kuwahara, Y.; Terasaki, N.; Inoue, K.; Yamada, S. Surface-Enhanced Nonresonance Raman Scattering from Size- and Morphology-Controlled Gold Nanoparticle Films. J. Phys. Chem. B 2004, 108, 11660−11665. (5) Laroche, T.; Vial, A.; Roussey, M. Crystalline Structure’s Influence on the Near-Field Optical Properties of Single Plasmonic Nanowires. Appl. Phys. Lett. 2007, 91, 123101. (6) Kusar, P.; Gruber, C.; Hohenau, A.; Krenn, J. R. Measurement and Reduction of Damping in Plasmonic Nanowires. Nano Lett. 2012, 12, 661−665. (7) Nauert, S.; Paul, A.; Zhen, Y.-R.; Solis, D.; Vigderman, L.; Chang, W.-S.; Zubarev, E. R.; Nordlander, P.; Link, S. Influence of Cross Sectional Geometry on Surface Plasmon Polariton Propagation in Gold Nanowires. ACS Nano 2014, 8, 572−580. (8) Liz-Marzán, L. M. Tailoring Surface Plasmons through the Morphology and Assembly of Metal Nanoparticles. Langmuir 2006, 22, 32−41. (9) Wang, H.; Fu, K.; Drezek, R. A.; Halas, N. J. Light Scattering from Spherical Plasmonic Nanoantennas: Effects of Nanoscale Roughness. Appl. Phys. B: Lasers Opt. 2006, 84, 191−195. (10) Trügler, A.; Tinguely, J.-C.; Krenn, J. R.; Hohenau, A.; Hohenester, U. Influence of Surface Roughness on the Optical 18348
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C
(52) Kamiko, M.; Yamamoto, R. Seeded Epitaxy of Co/Au(1 1 1) Multilayers on α-Al2O3(0 0 0 1): Influence of Co Seed Layer. Mater. Sci. Eng., B 2007, 141, 16−22. (53) Rank, D. H.; Eastman, D. P.; Rao, B. S.; Wiggins, T. A. Highly Precise Wavelengths in the Infrared. II. HCN, N2O, and CO. J. Opt. Soc. Am. 1961, 51, 929−936. (54) Blyholder, G. Molecular Orbital View of Chemisorbed Carbon Monoxide. J. Phys. Chem. 1964, 68, 2772−2777. (55) Schindler, K. M.; Hofmann, P.; Weiß, K. U.; Dippel, R.; Gardner, P.; Fritzsche, V.; Bradshaw, A. M.; Woodruff, D. P.; Davila, M. E.; Asensio, M. C.; et al. Is the Frequency of the Internal Mode of an Adsorbed Diatomic Molecule a Reliable Guide to Its Local Adsorption Site? J. Electron Spectrosc. Relat. Phenom. 1993, 64−65, 75− 83. (56) Sheppard, N.; Nguyen, T. The Vibrational Spectra of Carbon Monoxide Chemisorbed on the Surfaces of Metal Catalysts - a Suggested Scheme of Interpretation. Adv. IR and Raman Spectrosc. 1978, 5, 67−148. (57) Santiago-Rodríguez, Y.; Herron, J. A.; Curet-Arana, M. C.; Mavrikakis, M. Atomic and Molecular Adsorption on Au(111). Surf. Sci. 2014, 627, 57−69. (58) Hvolbæk, B.; Janssens, T. V. W.; Clausen, B. S.; Falsig, H.; Christensen, C. H.; Nørskov, J. K. Catalytic Activity of Au Nanoparticles. Nano Today 2007, 2, 14−18. (59) Mills, G.; Gordon, M. S.; Metiu, H. Oxygen Adsorption on Au Clusters and a Rough Au(111) Surface: The Role of Surface Flatness, Electron Confinement, Excess Electrons, and Band Gap. J. Chem. Phys. 2003, 118, 4198−4205. (60) Lemire, C.; Meyer, R.; Shaikhutdinov, S. K.; Freund, H. J. CO Adsorption on Oxide Supported Gold: from Small Clusters to Monolayer Islands and Three-Dimensional Nanoparticles. Surf. Sci. 2004, 552, 27−34. (61) Lopez, N.; Janssens, T. V. W.; Clausen, B. S.; Xu, Y.; Mavrikakis, M.; Bligaard, T.; Nørskov, J. K. On the Origin of the Catalytic Activity of Gold Nanoparticles for Low-Temperature CO Oxidation. J. Catal. 2004, 223, 232−235. (62) Biener, M. M.; Biener, J.; Friend, C. M. Enhanced Transient Reactivity of an O-Sputtered Au(1 1 1) Surface. Surf. Sci. 2005, 590, L259−L265. (63) Horn, K.; Hussain, M.; Pritchard, J. The Adsorption of CO on Cu(110). Surf. Sci. 1977, 63, 244−253. (64) Woodruff, D. P.; Hayden, B. E.; Prince, K.; Bradshaw, A. M. Dipole Coupling and Chemical Shifts in IRAS of CO adsorbed on Cu(110). Surf. Sci. 1982, 123, 397−412. (65) Hollins, P.; Davies, K. J.; Pritchard, J. Infrared Spectra of CO Chemisorbed on a Surface Vicinal to Cu(110): The Influence of Defect Sites. Surf. Sci. 1984, 75, 75−83. (66) Hollins, P. The Influence of Surface Defects on the Infrared Spectra of Adsorbed Species. Surf. Sci. Rep. 1992, 16, 51−94. (67) Gong, J. Structure and Surface Chemistry of Gold-Based Model Catalysts. Chem. Rev. 2012, 112, 2987−3054. (68) Sterrer, M.; Yulikov, M.; Risse, T.; Freund, H.-J.; Carrasco, J.; Illas, F.; Di Valentin, C.; Giordano, L.; Pacchioni, G. When the Reporter Induces the Effect: Unusual IR spectra of CO on Au1/ MgO(001)/Mo(001). Angew. Chem., Int. Ed. 2006, 45, 2633−2635. (69) Persson, B. N. J.; Ryberg, R. Vibrational Interaction Between Molecules Adsorbed on a Metal Surface: the Dipole-Dipole Interaction. Phys. Rev. B: Condens. Matter Mater. Phys. 1981, 24, 6954−6970. (70) Hollins, P.; Pritchard, J. Infrared Studies of Chemisorbed Layers on Single Crystals. Prog. Surf. Sci. 1985, 19, 275−349. (71) Browne, V. M.; Fox, S. G.; Hollins, P. Infrared Spectroscopy as an In Situ Probe of Morphology. Catal. Today 1991, 9, 1−14. (72) Bagus, P. S.; Müller, W. The Origin of the Shift in the C-O Vibration of Chemisorbed CO: Cluster Model Studies for CO/ Cu(100). Chem. Phys. Lett. 1985, 115, 540−544. (73) Bagus, P. S.; Pacchioni, G. Electric Field Effects on the Surface− Adsorbate Interaction: Cluster Model Studies. Electrochim. Acta 1991, 36, 1669−1675.
(31) Mavrikakis, M.; Stoltze, P.; Nørskov, J. K. Making Gold Less Noble. Catal. Lett. 2000, 64, 101−106. (32) Lopez, N.; Nørskov, J. K. Catalytic CO Oxidation by a Gold Nanoparticle: A Density Functional Study. J. Am. Chem. Soc. 2002, 124, 11262−11263. (33) Shaikhutdinov, S. K.; Meyer, R.; Naschitzki, M.; Bäumer, M.; Freund, H.-J. Size and Support Effects for CO Adsorption on Gold Model Catalysts. Catal. Lett. 2003, 86, 211−219. (34) Lemire, C.; Meyer, R.; Shaikhutdinov, S.; Freund, H.-J. Do Quantum Size Effects Control CO Adsorption on Gold Nanoparticles? Angew. Chem., Int. Ed. 2003, 43, 118−121. (35) Chen, M.; Cai, Y.; Yan, Z.; Goodman, D. W. On the Origin of the Unique Properties of Supported Au Nanoparticles. J. Am. Chem. Soc. 2006, 128, 6341−6346. (36) Laoufi, I.; Saint-Lager, M.-C.; Lazzari, R.; Jupille, J.; Robach, O.; Garaudée, S.; Cabailh, G.; Dolle, P.; Cruguel, H.; Bailly, A. Size and Catalytic Activity of Supported Gold Nanoparticles: An in Operando Study during CO Oxidation. J. Phys. Chem. C 2011, 115, 4673−4679. (37) Peterson, A. A.; Grabow, L. C.; Brennan, T. P.; Shong, B.; Ooi, C.; Wu, D. M.; Li, C. W.; Kushwaha, A.; Medford, A. J.; Mbuga, F.; et al. Finite-Size Effects in O and CO Adsorption for the Late Transition Metals. Top. Catal. 2012, 55, 1276−1282. (38) Seidelin, S.; Chiaverini, J.; Reichle, R.; Bollinger, J. J.; Leibfried, D.; Britton, J.; Wesenberg, J. H.; Blakestad, R. B.; Epstein, R. J.; Hume, D. B.; et al. Microfabricated Surface-Electrode Ion Trap for Scalable Quantum Information Processing. Phys. Rev. Lett. 2006, 96, 253003. (39) Sam, F. L. M.; Mills, C. A.; Rozanski, L. J.; Silva, S. R. P. Thin Film Hexagonal Gold Grids as Transparent Conducting Electrodes in Organic Light Emitting Diodes. Laser & Photonics Rev. 2014, 8, 172− 179. (40) Hu, Y.; Warwick, C.; Sou, A.; Jiang, L.; Sirringhaus, H. Fabrication of Ultra-Flexible, Ultra-Thin Organic Field-Effect Transistors and Circuits by a Peeling-Off Method. J. Mater. Chem. C 2014, 2, 1260−1263. (41) Binnig, G.; Rohrer, H.; Gerber, C.; Weibel, E. Surface Studies by Scanning Tunneling Microscopy. Phys. Rev. Lett. 1982, 49, 57−61. (42) Vancea, J.; Reiss, G.; Schneider, F.; Bauer, K.; Hoffmann, H. Substrate Effects on the Surface Topography of Evaporated Gold Films - a Scanning Tunnelling Microscopy Investigation. Surf. Sci. 1989, 218, 108−126. (43) Semaltianos, N. G.; Wilson, E. G. Investigation of the Surface Morphology of Thermally Evaporated Thin Gold Films on Mica, Glass, Silicon and Calcium Fluoride Substrates by Scanning Tunneling Microscopy. Thin Solid Films 2000, 366, 111−116. (44) Watanabe, M. O.; Kuroda, T.; Tanaka, K.; Sakai, A. Scanning Tunneling Microscopy Observation of Crystal Growth of Deposited Gold Films During Annealing. J. Vac. Sci. Technol., B: Microelectron. Process. Phenom. 1991, 9, 924−927. (45) Ito, Y.; Kushida, K.; Takeuchi, H. Role of Chromium Sublayers in the Growth of Highly Crystalline (111)-Oriented Gold Films on Sapphire. J. Cryst. Growth 1991, 112, 427−436. (46) Peale, D. R.; Cooper, B. H. Adsorbate-Promoted Mass Flow on the Gold (111) Surface Observed by Scanning Tunneling Microscopy. J. Vac. Sci. Technol., A 1992, 10, 2210−2215. (47) Nogues, C.; Wanunu, M. A Rapid Approach to Reproducible, Atomically Flat Gold Films on Mica. Surf. Sci. 2004, 573, L383−L389. (48) Fujita, D.; Yakabe, T.; Nejoh, H.; Sato, T.; Iwatsuki, M. Scanning Tunneling Microscopy Study on the Initial Adsorption Behavior of C60 Molecules on a Reconstructed Au(111)-(23×√3) Surface at Various Temperatures. Surf. Sci. 1996, 366, 93−98. (49) Kowalczyk, P.; Kozlowski, W.; Klusek, Z.; Olejniczak, W.; Datta, P. STM Studies of the Reconstructed Au(111) Thin-film at Elevated Temperatures. Appl. Surf. Sci. 2007, 253, 4715−4720. (50) Kästle, G.; Boyen, H.-G.; Koslowski, B.; Plettl, A.; Weigl, F.; Ziemann, P. Growth of Thin, Flat, Epitaxial (1 1 1) Oriented Gold Films on C-Cut Sapphire. Surf. Sci. 2002, 498, 168−174. (51) Kamiko, M.; Yamamoto, R. Epitaxial Growth of Au(1 1 1) on αAl2O3(0 0 0 1) by Using a Co Seed Layer. J. Cryst. Growth 2006, 293, 216−222. 18349
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C (74) Yates, D. J. C. Spectroscopic Investigations of Gold Surfaces. J. Colloid Interface Sci. 1969, 29, 194−204. (75) Bradshaw, A. M.; Pritchard, J. Infrared Spectra of Carbon Monoxide Chemisorbed on Metal Films: A Comparative Study of Copper, Silver, Gold, Iron Cobalt and Nickel. Proc. R. Soc. London, Ser. A 1970, 316, 169−183. (76) Kottke, M. L.; Greenler, R. G.; Tompkins, H. G. An Infrared Spectroscopic Study of Carbon Monoxide Adsorbed on Polycrystalline Gold Using the Reflection-Absorption Technique. Surf. Sci. 1972, 32, 231−243. (77) Dumas, P.; Tobin, R. G.; Richards, P. L. Study of Adsorption States and Interactions of CO on Evaporated Noble Metal Surfaces by Infrared Absorption Spectroscopy: II. Gold and Copper. Surf. Sci. 1986, 171, 579−599. (78) Jugnet, Y.; Cadete Santos Aires, F. J.; Deranlot, C.; Piccolo, L.; Bertolini, J. C. CO Chemisorption on Au(110) Investigated under Elevated Pressures by Polarized Reflection Absorption Infrared Spectroscopy and Scanning Tunneling Microscopy. Surf. Sci. 2002, 521, L639−L644. (79) Meier, D. C.; Bukhtiyarov, V.; Goodman, D. W. CO Adsorption on Au(110)-(1 × 2): An IRAS Investigation. J. Phys. Chem. B 2003, 107, 12668−12671. (80) Piccolo, L.; Loffreda, D.; Cadete Santos Aires, F. J.; Deranlot, C.; Jugnet, Y.; Sautet, P.; Bertolini, J. C. The Adsorption of CO on Au(111) at Elevated Pressures Studied by STM, RAIRS and DFT Calculations. Surf. Sci. 2004, 566−568, 995−1000. (81) Nakamura, I.; Takahashi, A.; Fujitani, T. Selective Dissociation of O3 and Adsorption of CO on Various Au Single Crystal Surfaces. Catal. Lett. 2009, 129, 400−403. (82) Yim, W.-L.; Nowitzki, T.; Necke, M.; Schnars, H.; Nickut, P.; Biener, J.; Biener, M. M.; Zielasek, V.; Al-Shamery, K.; Klüner, T.; et al. Universal Phenomena of CO Adsorption on Gold Surfaces with LowCoordinated Sites. J. Phys. Chem. C 2007, 111, 445−451. (83) Hrbek, J.; Hoffmann, F. M.; Park, J. B.; Liu, P.; Stacchiola, D.; Hoo, Y. S.; Ma, S.; Nambu, A.; Rodriguez, J. A.; White, M. G. Adsorbate-Driven Morphological Changes of a Gold Surface at Low Temperatures. J. Am. Chem. Soc. 2008, 130, 17272−17273. (84) Chang, S.-C.; Hamelin, A.; Weaver, M. J. Reactive and Inhibiting Adsorbates for the Catalytic Electrooxidation of Carbon Monoxide on Gold(210) as Characterized by Surface Infrared Spectroscopy. Surf. Sci. 1990, 239, L543−L547. (85) Kim, J.; Samano, E.; Koel, B. E. CO Adsorption and Reaction on Clean and Oxygen-Covered Au(211) Surfaces. J. Phys. Chem. B 2006, 110, 17512−17517. (86) Ruggiero, C.; Hollins, P. Adsorption of Carbon Monoxide on the Gold(332) Surface. J. Chem. Soc., Faraday Trans. 1996, 92, 4829− 4834. (87) Boronat, M.; Concepción, P.; Corma, A. Unravelling the Nature of Gold Surface Sites by Combining IR Spectroscopy and DFT Calculations. Implications in Catalysis. J. Phys. Chem. C 2009, 113, 16772−16784. (88) Pászti, Z.; Hakkel, O.; Keszthelyi, T.; Berkó, A.; Balázs, N.; Bakó, I.; Guczi, L. Interaction of Carbon Monoxide with Au(111) Modified by Ion Bombardment: A Surface Spectroscopy Study under Elevated Pressure. Langmuir 2010, 26, 16312−16324. (89) Mehmood, F.; Kara, A.; Rahman, T. S.; Henry, C. R. Comparative Study of CO Adsorption on Flat, Stepped, and Kinked Au Surfaces Using Density Functional Theory. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 075422. (90) Maksymovych, P.; Yates, J. Unexpected Spontaneous Formation of CO Clusters on the Au(111) Surface. Chem. Phys. Lett. 2006, 421, 473−477. (91) Pischel, J. Ü ber die schwingungsspektroskopische Untersuchung von Anregungen im mittleren und fernen Infrarot an Oberflächen metallischer Einkristalle. Ph.D. Thesis, Heidelberg University, 2014. (92) Krauth, O.; Fahsold, G.; Pucci, A. Asymmetric Line Shapes and Surface Enhanced Infrared Absorption of CO Adsorbed on Thin Iron Films on MgO(001). J. Chem. Phys. 1999, 110, 3113−3117.
(93) Rai, A.; Nayak, J.; Roy Barman, S. Temporal Evolution and Nature of Nanostructures on Au(111). Surf. Sci. 2014, 625, 97−103. (94) Greenler, R. G. Infrared Study of Adsorbed Molecules on Metal Surfaces by Reflection Techniques. J. Chem. Phys. 1966, 44, 310−315. (95) Henzler, M.; Göpel, W. Oberflächenphysik des Festkörpers; B. G. Teubner: Stuttgart, 1991. (96) Elisabetta Palumbo, M.; Baratta, G. A.; Collings, M. P.; McCoustra, M. R. S. The Profile of the 2140 cm−1 Solid CO Band on Different Substrates. Phys. Chem. Chem. Phys. 2006, 8, 279−284. (97) Ewing, G. E.; Pimentel, G. C. Infrared Spectrum of Solid Carbon Monoxide. J. Chem. Phys. 1961, 35, 925−930. (98) Eve, J. K.; McCash, E. M. Low-Temperature Adsorption of CO on Cu(111) Studied by Reflection Absorption Infrared Spectroscopy. Chem. Phys. Lett. 1999, 313, 575−581. (99) Cook, J. C.; McCash, E. M. Reversible Phase Formation for Chemisorption/Physisorption of CO on Cu(100). Surf. Sci. 1996, 356, L445−L449. (100) Nekrylova, J. V.; French, C.; Artsyukhovich, A. N.; Ukraintsev, V. A.; Harrison, I. Low Temperature Adsorption of CO on Pt(111): Disequilibrium and the Occupation of Three-Fold Hollow Sites. Surf. Sci. 1993, 295, L987−L992. (101) Vegard, I. Struktur und Leuchtfäh igkeit von festem Kohlenoxyd. Eur. Phys. J. A 1930, 61, 185−190. (102) Gottfried, J. M.; Schmidt, K. J.; Schroeder, S. L. M.; Christmann, K. Adsorption of Carbon Monoxide on Au(110)-(1 × 2). Surf. Sci. 2003, 536, 206−224. (103) Hussain, A.; Curulla Ferré, D.; Gracia, J.; Nieuwenhuys, B. E.; Niemantsverdriet, J. W. DFT Study of CO and NO Adsorption on Low Index and Stepped Surfaces of Gold. Surf. Sci. 2009, 603, 2734− 2741. (104) Okamoto, Y. Density-Functional Calculations of Atomic and Molecular Adsorptions on 55-Atom Metal Clusters: Comparison with (1 1 1) Surfaces. Chem. Phys. Lett. 2005, 405, 79−83. (105) Persson, B. N. J. Surface Resistivity and Vibrational Damping in Adsorbed Layers. Phys. Rev. B: Condens. Matter Mater. Phys. 1991, 44, 3277−3296. (106) Persson, B. N. J. Surface Resistivity: Theory and Applications. Surf. Sci. 1992, 269−270, 103−112. (107) Persson, B. N. J. Applications of Surface Resistivity to Atomic Scale Friction, to the Migration of “Hot” Adatoms, and to Electrochemistry. J. Chem. Phys. 1993, 98, 1659−1672. (108) Boccuzzi, F.; Tsubota, S.; Haruta, M. Vibrational Investigation of CO Adsorbed on Gold Deposited on TiO2. J. Electron Spectrosc. Relat. Phenom. 1993, 64−65, 241−250. (109) Maksymovych, P.; Voznyy, O.; Dougherty, D. B.; Sorescu, D. C.; Yates, J. T., Jr. Gold Adatom as a Key Structural Component in Self-Assembled Monolayers of Organosulfur Molecules on Au(1 1 1). Prog. Surf. Sci. 2010, 85, 206−240. (110) Gao, W.; Baker, T. A.; Zhou, L.; Pinnaduwage, D. S.; Kaxiras, E.; Friend, C. M. Chlorine Adsorption on Au(111): Chlorine Overlayer or Surface Chloride? J. Am. Chem. Soc. 2008, 130, 3560− 3565. (111) Shi, Z.; Lin, N. Porphyrin-Based Two-Dimensional Coordination Kagome Lattice Self-Assembled on a Au(111) Surface. J. Am. Chem. Soc. 2009, 131, 5376−5377. (112) Min, B. K.; Deng, X.; Pinnaduwage, D.; Schalek, R.; Friend, C. M. Oxygen-Induced Restructuring with Release of Gold Atoms from Au(111). Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 121410. (113) Remediakis, I. N.; Lopez, N.; Nørskov, J. K. CO Oxidation on Gold Nanoparticles: Theoretical Studies. Appl. Catal., A 2005, 291, 13−20. (114) Iski, E. V.; Jewell, A. D.; Tierney, H. L.; Kyriakou, G.; Sykes, E. C. H. Controllable Restructuring of a Metal Substrate: Tuning the Surface Morphology of Gold. Surf. Sci. 2012, 606, 536−541. (115) Michely, T.; Besocke, K. H.; Comsa, G. Observation of Sputtering Damage on Au(111). Surf. Sci. 1990, 230, L135−L139. (116) Nan, L.; Allan, D.; Gang-yu, L. In Situ STM Study of Thermal Annealing of Au Thin Films: an Investigation on Decay of Nanometer 18350
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351
Article
The Journal of Physical Chemistry C Au Clusters and 2D Islands. Acta Phys. Sin. (Overseas Ed.) 1997, 6, 531−549. (117) Pemberton, J. E.; Guy, A. L.; Sobocinski, R. L.; Tuschel, D. D.; Cross, N. A. Surface Enhanced Raman Scattering in Electrochemical Systems: The Complex Roles of Surface Roughness. Appl. Surf. Sci. 1988, 32, 33−56. (118) Otto, A. The ‘Chemical’ (Electronic) Contribution to SurfaceEnhanced Raman Scattering. J. Raman Spectrosc. 2005, 36, 497−509. (119) Sun, S.-G.; Cai, W.-B.; Wan, L.-J.; Osawa, M. Infrared Absorption Enhancement for CO Adsorbed on Au Films in Perchloric Acid Solutions and Effects of Surface Structure Studied by Cyclic Voltammetry, Scanning Tunneling Microscopy, and Surface-Enhanced IR Spectroscopy. J. Phys. Chem. B 1999, 103, 2460−2466. (120) Skriver, H. L.; Rosengaard, N. M. Surface Energy and Work Function of Elemental Metals. Phys. Rev. B: Condens. Matter Mater. Phys. 1992, 46, 7157−7168. (121) Vitos, L.; Ruban, A. V.; Skriver, H. L.; Kollár, J. The Surface Energy of Metals. Surf. Sci. 1998, 411, 186−202. (122) Carmichael, E. S.; Gruebele, M. Controlling the Smoothness of Optically Transparent Gold Films by Temperature Tuning. J. Phys. Chem. C 2009, 113, 4495−4501. (123) Amram, D.; Rabkin, E. On the Role of Fe in the Growth of Single Crystalline Heteroepitaxial Au Thin Films on Sapphire. Acta Mater. 2013, 61, 4113−4126.
18351
DOI: 10.1021/acs.jpcc.5b05051 J. Phys. Chem. C 2015, 119, 18340−18351