PCCP View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 3823
View Journal | View Issue
In situ IR spectroscopic studies of Ni surface segregation induced by CO adsorption on Cu–Ni/SiO2 bimetallic catalysts Yunxi Yao* and D. Wayne Goodman† It is of great importance to study the catalytic structures under real reaction conditions especially for the bimetallic catalysts, where facile surface restructure or surface segregation can be driven by adsorbate adsorption. Here, we report CO interaction with Cu–Ni/SiO2 bimetallic model catalysts studied by CO temperature programmed desorption (TPD) and in situ CO polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS) under CO pressures varying from ultrahigh vacuum (UHV) to near ambient pressure. Under UHV conditions, Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic cata-
Received 27th November 2013, Accepted 24th December 2013
lysts. CO spillover from Cu to Ni on Cu–Ni/SiO2 bimetallic catalysts has been observed at about 200 K
DOI: 10.1039/c3cp54997f
catalysts induced by CO adsorption at ambient pressure CO. The behavior of CO induced surface segre-
under UHV conditions. In situ CO PM-IRRAS shows surface segregation of Ni on the Cu–Ni/SiO2 bimetallic gation can lead to severe errors in Ni active site measurements by the selective CO chemisorption on
www.rsc.org/pccp
Cu–Ni/SiO2 bimetallic catalysts.
1. Introduction Bimetallic catalysts have attracted great research efforts in the past decades due to their chemical and physical properties different from the individual pure metals and promising applications in chemical conversion, energy technology and environmental protection.1–11 Extensive fundamental insights into the surface composition, surface atomic ensemble, electronic structure, and gas adsorption property of bimetallic catalysts have been gained.2–8 Most of these studies were conducted under ex situ or vacuum conditions. Bimetallic catalysts, mostly used as supported bimetallic nanoparticles, are very flexible to restructuring and surface segregation induced by ambient pressure gas adsorption or reaction.10,12–16 The surface structure, atomic composition, and chemical state of bimetallic catalysts obtained by ex situ techniques or ultrahigh vacuum based surface science studies are often different from those under ambient pressure adsorption or reaction conditions. Catalyst characterization under elevated pressure or reaction conditions, especially for bimetallic catalysts, becomes more demanding to understand the structure–activity relationship. Many in situ techniques have been developed for this purpose,17 and among them in situ IR is probably the mostly widely used and developed due to its low cost, easy setup, and Department of chemistry, Texas A&M University, College Station, TX 77843, USA. E-mail: [email protected] † Dr Goodman deceased on February 27, 2012.
This journal is © the Owner Societies 2014
effectiveness for catalyst characterization.18,19 Polarization modulation infrared reflection absorption spectroscopy (PMIRRAS), as one peculiar IR technique, can be operated in reflection mode with polarized IR light, which enables the surface vibrational spectra to be obtained by the subtraction of gas-phase absorption.19 This technique allows in situ measurements of catalytic surfaces from ultrahigh vacuum (UHV) to atmospheric conditions.20 Since the pioneering work of Sinfelt et al. on the hydrogenolysis of ethane over Cu–Ni alloys in 1972,21 Cu–Ni bimetallic catalysts have been considered as a representative system to study the structure–activity relationship in bimetallic catalysts.1,22–28 Many catalytic reactions, which include hydrogenation, dehydrogenation, hydrogenolysis, isomerization, and Fischer–Tropsch synthesis, have been studied using Cu–Ni bimetallic catalysts.1,22–27,29,30 Extensive and controversial studies on the surface structure, surface composition, electronic structure, and structure–activity relationship of bimetallic Cu–Ni systems have been reported, and the main findings are summarized as follows: (1) it is widely reported that Cu is enriched on the surface of Cu–Ni alloys,22,25,26,31–33 but there are two exceptions which report that the surface of Cu–Ni bimetallic catalysts has a bulk composition.24,34 (2) Ponec reviewed the reported results and found that there are no essential changes in the electronic structure of Ni atoms caused by alloying with Cu atoms,35 whereas Abbati et al. found that d band photoemission from Ni is modified in the presence of Cu on the surface and there is a significant interaction between the
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3823
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
d states of Cu and Ni atoms.36 (3) The sharp decrease in the catalytic activity of Cu–Ni bimetallic catalysts upon the addition of Cu is usually explained by the ensemble effect,1,21,22,24,25,29 while Alstrup et al. found the advanced ensemble model to be invalid in the explanation of the sulfur modification effect on nickel catalysts in propane hydrogenolysis, and they proposed that special sites are needed for breaking of the C–C bonds in the hydrogenolysis reactions on Cu–Ni bimetallic catalysts. (4) The Cu effect on the activity of Ni remains controversial. There is a three-to-five order decrease in the specific activity of ethane or propane hydrogenolysis on Cu–Ni bimetallic catalysts with increasing Cu amount.1,21,22,24,25,29 Campbell and Emmett reported that the rate of ethylene hydrogenation on Ni was increased by 7 to 15 times upon addition of Cu to Ni catalysts.37 (5) H2 and CO adsorption on Cu–Ni bimetallic catalysts was extensively studied. The selective hydrogen chemisorption was used to measure the surface active Ni sites on Cu–Ni catalysts.1,21,22,24,29,38 But Crucq and co-workers found hydrogen spillover from Ni to Cu on Cu–Ni bimetallic surfaces,39 which can lead to over counting of surface Ni sites in the hydrogen selective chemisorption.32 CO has been used as a probe molecule to study the surface structure and composition of Cu–Ni bimetallic catalysts.35,40–44 In spite of extensive studies on Cu–Ni bimetallic catalysts, in situ studies on the interaction of CO with Cu–Ni bimetallic catalysts are still missing. Based on the large difference in bond strength between CO on Ni and Cu sites, possible surface segregation induced by CO adsorption can be expected, which is the key to understand CO involved reactions on Cu–Ni bimetallic catalysts, such as methanation of CO30 and the water-gas-shift reaction on Cu–Ni bimetallic catalysts.45 In the present work, we studied CO adsorption on Cu–Ni bimetallic model catalysts supported on SiO2 thin films from UHV to ambient pressures by TPD and in situ PM-IRRAS. By taking the advantage of surface science techniques and the UHV coupled high pressure in situ PM-IRRAS, the surface composition and structure of Cu–Ni bimetallic catalysts under vacuum conditions, and also under ambient pressure conditions, were thoroughly investigated. Cu is found to be enriched on the surface of Cu–Ni bimetallic catalysts under UHV conditions, but CO can induce Ni surface segregation at elevated pressure.
PCCP
Hinds PEM-90 photoelastic modulator) spectra were recorded using 600 scans at a resolution of 4 cm 1. A linear ramp of 3 K s 1 was used for TPD measurements. The Mo(110) sample was mounted onto the end of the sample manipulator. The sample temperature can be measured using a pair of W-5%Re–W-26%Re C type thermocouple spotwelded on the back of the sample. Thin SiO2 films (B10 nm) were prepared by evaporating Si in 1 10 5 Torr O2 at room temperature and subsequently annealed to 1200 K.46 Cu and Ni were deposited by heating tantalum wires wrapped with highly pure Ni or Cu wires. The amount of metal deposited was controlled by the depositing time and calibrated with AES spectra, and was given in the unit of monolayer equivalent (MLE, 1 MLE = 1.43 1015 atom per cm2).46 C. P. grade CO (Matheson Tri-Gas) was used and further purified by passing it through a LN2 cooling trap and stored in a glass bulb maintained at LN2 temperature throughout the CO adsorption measurements.
3. Results and discussion 3.1 Room temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts Cu–Ni bimetallic catalysts were grown on SiO2 by depositing different amounts of Cu on 5 MLE Ni/SiO2, which was subsequently annealed at 700 K for 5 min. Fig. 1a shows the room temperature CO TPD after saturated CO exposure on Cu–Ni/SiO2 bimetallic catalysts as a function of Cu coverage. Under UHV conditions, CO desorbs on Cu surfaces below 250 K,47,48 so CO exclusively adsorbs on Ni sites on Cu–Ni bimetallic catalysts at room temperature. CO TPD shows two CO desorption peaks at 350 K and 410 K on clean 5 MLE Ni/SiO2, which is similar to saturated CO desorption from Ni(111).49 Upon the addition of Cu, the low temperature desorption peak keeps its position at 350 K, while the high temperature peak gradually shifts from 410 K on 5 MLE Ni/SiO2 to 387 K on 5 MLE Cu-5 MLE Ni/SiO2. The observed behavior is consistent with CO desorption from Cu(110) covered with Ni over layers.50 However, the weakened
2. Experimental All the experiments were conducted in a ultrahigh vacuum chamber using Auger electron spectroscopy (AES), polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS), and a UTI-100 mass spectrometer for temperature programmed desorption (TPD) measurements, which has been described in detail previously.46 A high pressure/IR cell equipped with CaF2 windows was attached to the bottom of the UHV chamber, which can be isolated from the main chamber by differentially pumped sliding seals for high pressure studies. The base pressure for both the main chamber and the high pressure/IR cell is less than 2 10 10 Torr. PM-IRRAS (Bruker Equinox 55,
3824 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
Fig. 1 (a) CO TPD of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at room temperature with indicated Cu coverage. (b) Normalized CO TPD area as a function of Cu coverage.
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
CO adsorption on Ni upon alloying with Cu, as observed here on Cu–Ni/SiO2 bimetallic catalysts, was not observed on the Ni(111) surface upon the addition of Cu atoms,47 which indicates that the chemical properties of supported Ni nanoparticles are more prone to be modified upon alloying with Cu than bulk Ni surfaces. The low temperature desorption feature has been attributed to CO desorption from low coordinated step Ni sites on Ni nanoparticles, while the high temperature desorption feature is due to CO desorption on terrace sites.46 The TPD results hint that addition of Cu has a more profound effect on CO adsorption on terrace sites than on step sites on supported Ni nanoparticles. Fig. 1b shows the normalized CO desorption intensity as a function of Cu coverage. CO desorption from Ni sharply decreases with increasing Cu coverage up to 1.5 MLE. We have measured the surface active Ni sites on 5 MLE Ni/SiO2 by H2 TPD to be 1.7 MLE (1.93 1015 Ni atoms). CO TPD results show that the Ni surface sites are blocked by Cu atoms almost on a one on one basis. This behavior is consistent with CO TPD from Cu layers on Ni(111) where Ni surface sites are totally covered with 1 ML Cu.47 At low Cu coverages (up to 1.5 MLE), Cu atoms prefer to populate on the surface, which agrees with the surface enrichment of Cu on the Cu–Ni bimetallic catalysts.22,25,26,31–33 After deposition of 2 MLE Cu, the CO desorption intensity remains constant, which is about 10% of CO desorption from the clean 5 MLE Ni/SiO2 surface. Fig. 2 shows the room temperature PM-IRRAS spectra after saturated exposure of CO on Cu–Ni/SiO2 with increasing Cu coverage. The PM-IRRAS of the monometallic Ni/SiO2 catalyst shows two stretching frequencies at 1950 cm 1 and 2058 cm 1, which are attributed to CO adsorbed on atop and bridging Ni sites, respectively.41,51,52 After deposition of 0.5 MLE Cu on the 5 MLE Ni/SiO2, the bridge site CO adsorption peak completely disappears, indicating a sharp decrease in the adjacent Ni sites by the addition of Cu atoms. The intensity of the atop CO adsorption peak increases and its stretching frequency shifts from 2058 cm 1 on the monometallic Ni/SiO2 to 2024 cm 1 on the 0.5 MLE Cu–Ni/SiO2. The CO stretching frequency continues to decrease to 1999 cm 1 at 1.5 MLE Cu–Ni/SiO2, and
Fig. 2 Room temperature PM-IRRAS of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces with indicated Cu coverage. The Cu–Ni/SiO2 surfaces were exposed to 1 10 6 Torr CO for 10 min.
This journal is © the Owner Societies 2014
Paper
afterwards it remains constant at 1999 cm 1, which resembles the stretching frequency of CO at 2000 cm 1 from CO adsorption on 1 ML Cu/Ni(111).41 After the addition of 0.5 MLE Cu, the intensity of the atop CO peak starts to decrease with increasing Cu coverage up to 2 MLE, and after that the atop CO peak intensity remains constant. The decrease in CO adsorption intensity is due to the surface Ni sites being blocked by Cu atoms. The large red shifts of atop CO adsorption and the absence of the bridging CO adsorption indicate that the surface Ni sites are isolated with the added Cu atoms and there is a weak dipole coupling of adsorbed CO molecules. Room temperature CO adsorption results show that Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic catalysts, which was further confirmed by our AES results. Fig. 3a shows AES spectra of Cu–Ni/SiO2 with increasing Cu coverage. With increasing Cu coverage, the Cu AES peaks increase while the Ni AES peaks decrease in intensity due to the Ni AES signal attenuated by Cu over layers. Fig. 3b shows the intensity changes of Ni and Cu as a function of Cu coverage. The Ni AES peak at about 740 eV was used instead of the main Ni AES peak at about 870 eV because it is overlapped with Cu AES signals. The Ni and Cu AES intensities are normalized to their maximum intensities, respectively. From the plots in Fig. 3b we can see that the Cu AES intensity linearly increases while the Ni AES intensity linearly decreases with increasing Cu coverages up to 1.5 MLE, which indicates Cu growth on Ni follows a layerby-layer mechanism, similar to Cu grown on Ru(0001).53 A mixed Cu–Ni layer can be formed at 700 K as observed in Cu–Ni(111) systems.41,54 But here the Cu atoms deposited at coverages less than 1.5 MLE mainly populate on the surface blocking CO adsorption as observed by CO TPD, which agrees with the Cu atoms segregating on the Cu–Ni bimetallic surface and form Cu over layers due to its lower surface free energy.26,41,47,55,56 As we mentioned before, the active Ni sites on the monometallic 5 MLE Ni/SiO2 are measured to be 1.7 MLE, which coincidently equals to the observed breaking point in the
Fig. 3 (a) AES of Cu-5 MLE Ni/SiO2 surfaces with indicated Cu coverage. (b) The evolution of Ni (740 eV) and Cu (950 eV) AES peak intensity as a function of Cu coverage. The Ni and Cu AES peak intensities were normalized to their maximum value, respectively. The Ni AES peak at about 740 eV was used instead of the main Ni AES peak at about 870 eV because it is overlapped with Cu AES signals.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3825
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
PCCP
AES spectra. After deposition of 1.5 MLE Cu, the Ni AES intensity decreases by about 18%, which agrees so well with the calculated value of 18% decrease in Ni AES signal due to the attenuation by one atomic Cu over layer by using the electron inelastic mean free paths at 740 eV in Cu of 1.3 nm and the thickness of one atomic Cu layer of 0.255 nm. AES results combined with room temperature CO adsorption show Cu atoms mainly populate on the Ni surface and block the Ni active sites at low Cu coverages, which may account for the sharp decrease in catalytic activity.21,22,24 3.2 Low temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts In order to study CO adsorption on Cu sites and the temperature dependent CO adsorption dynamics on Cu–Ni/SiO2 bimetallic catalysts, CO adsorption on Cu–Ni/SiO2 at low temperatures (around 130 K) was investigated. Fig. 4 shows CO TPD on Cu-5 MLE Ni/SiO2 with indicated Cu coverage saturated with CO at 130 K. On the monometallic 5 MLE Ni/SiO2, the main CO desorption peak is observed at 415 K, which is consistent with the room temperature CO adsorption results (Fig. 1). Below room temperature, there are two CO desorption peaks at about 200 K and 266 K. CO TPD on Ni(111) with CO dosed at low temperatures also shows a broad desorption peak below 300 K due to the increased CO coverage at low temperature.49,56 After the addition of 0.2 MLE Cu, the main CO desorption peak at 415 K on 5 MLE Ni/SiO2 shifts to 384 K, which agrees with room temperature observations and indicates that CO adsorption on Ni sites is weakened upon addition of Cu. Meanwhile, a new desorption peak appears at about 160 K attributed to CO desorption from Cu surfaces.48 Upon further increasing the Cu coverage to 0.5 MLE, besides the CO desorption from Ni sites which continues to decrease in intensity, there are two well-resolved CO desorption peaks at about 160 K and 210 K, which are similar to CO desorption from Cu/SiO2.48 From the CO TPD results, we cannot tell whether CO adsorption on Cu sites is strengthened or not. After deposition of 2 MLE Cu on Ni/SiO2, the CO desorption from Ni sites (integration of CO desorption
Fig. 4 CO TPD of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at about 130 K with indicated Cu coverage.
3826 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
Fig. 5 PM-IRRAS of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at 130 K with indicated Cu coverage.
above 300 K) is only about 4% of the original value from clean monometallic 5 MLE Ni/SiO2. This result is contrary to room temperature observation, where 10% CO desorption intensity remains on 2 MLE Cu–Ni/SiO2 surfaces. The larger CO adsorption capacity at room temperature than that at low temperature hints that possible CO induced Ni surface segregation at room temperature takes place. Low temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts was also studied by CO PM-IRRAS. Fig. 5 shows CO PM-IRRAS after saturated exposure of CO at about 130 K on Cu-5 MLE Ni/SiO2 with increasing Cu coverage. After the addition of 0.2 MLE Cu, the bridging CO adsorption on Ni sites totally disappears, and the atop CO stretching frequency shifts from 2070 cm 1 on 5 MLE Ni/SiO2 to 2028 cm 1. A new CO stretching frequency at 2126 cm 1 appears, which is attributed to atop CO on isolated Cu atoms or small Cu clusters.41,57 At low Cu coverages, the existence of single Cu atoms on Ni clusters can be expected. Upon further increasing Cu coverage to 0.5 MLE, the CO adsorption peaks on Ni sites completely disappear. The atop CO adsorption frequency on Cu sites shifts from 2126 cm 1 at 0.2 MLE Cu coverage to 2114 cm 1 at 0.5 MLE Cu coverage. The 2114 cm 1 CO stretching peak is attributed to CO adsorption on Cu over layers on Ni clusters.57–59 A new CO stretching peak appears at 2076 cm 1, which is similar to the CO stretching frequency from CO adsorption on Cu(111),55,60 indicating that multilayer Cu or Cu clusters form at this Cu coverage. From the evolution trend of CO adsorption spectra shown in Fig. 5, the possibility of this 2076 cm 1 peak due to CO adsorbed on Ni sites is excluded. At the Cu coverages above 1 MLE, there are two CO stretching peaks at 2105 cm 1 and 2076 cm 1, which maintain their positions with increasing Cu coverage. Compared with room temperature CO adsorption results, CO adsorption on Ni sites is totally blocked by 0.5 MLE Cu 130 K, while there is a clear CO adsorption peak on Ni detected at 1999 cm 1 even at a Cu coverage of 5 MLE. Another difference is that at room temperature the CO adsorption on Ni sites is blocked by Cu on a one on one basis, but at low temperature, it seems one Cu atom can block more than one Ni sites.41
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
Fig. 6 PM-IRRAS of saturated CO adsorption on (a) 0.2 MLE Cu-5 MLE Cu/SiO2 and (b) 2.0 MLE Cu-5 MLE Cu/SiO2 at about 150 K and subsequently annealed at indicated temperatures.
Another information from low temperature CO PM-IRRAS is the blue shift of CO bound to Cu over layers with respect to CO adsorbed on multilayer Cu, which indicates that, between Ni clusters and Cu overalyers, there is interface charge transfer or electronic modification.61–65 Goodman et al. have proposed that the blue shift of CO adsorption on Cu over layers on Rh(100) is due to CO bound to positively charged or selfpolarized Cu.57 However, the d-electron back donation becomes less dominant for CO adsorption on Cu than for transition metals.59 The blue shift may arise due to the electronic structure of Cu modified via the Ni–Cu interface interaction.36 Temperature effects on CO adsorption on Cu–Ni/SiO2 were studied by variable temperature CO PM-IRRAS. Fig. 6a shows CO PM-IRRAS after saturated CO exposure on 0.2 MLE Cu–Ni/SiO2 at 150 K and subsequently increasing the sample temperature to the indicated temperatures. At 150 K, there are two adsorption peaks at 2116 cm 1 and 2016 cm 1 attributed respectively to atop CO adsorption on Cu and Ni sites on 0.2 MLE Cu–Ni/SiO2. Upon increasing the sample temperature to 200 K, the intensity of CO adsorbed on Cu decreases, but its peak position remains unchanged. The CO adsorption on Ni sites increases in intensity and its peak position blue shifts from 2016 cm 1 to 2032 cm 1. The blue shift can be explained by the increased CO coverage at Ni sites, which is evidenced by the increased intensity of the corresponding CO stretching peak. Upon further increasing the sample temperature to 250 K, CO bound to Cu sites completely disappears, which is consistent with our CO TPD results (Fig. 4). CO adsorbed on atop Ni sites further gains in intensity and its peak position is maintained at 2032 cm 1. Upon further increasing the sample temperature, the CO stretching frequency shifts to 2005 cm 1 due to the decreased intermolecular interaction by partial desorption of CO on Ni sites. The CO PM-IRRAS results on 0.2 MLE Cu–Ni/SiO2 show CO spillover from Cu to Ni at about 200 K accompanied with CO partial desorption from Cu surfaces. This result indicates that CO adsorption at 200 K can induce Ni surface segregation on Cu–Ni/SiO2 bimetallic catalysts. Fig. 6b shows another variable temperature CO PM-IRRAS on 2 MLE Cu–Ni/SiO2 saturated with CO at 150 K. No CO adsorption on Ni sites is detected at 150 K showing that all
This journal is © the Owner Societies 2014
Paper
the surface Ni sites are blocked by 2 MLE Cu. However, upon increasing the sample temperature to 200 K, a CO adsorption peak appears at 1997 cm 1 attributed to atop CO on Ni sites, while a sharp decrease of CO adsorption on Cu sites is observed. The evolution of the CO adsorption peaks indicates CO spillover from Cu to Ni while CO partially desorbs from Cu surfaces at about 200 K. It should be noted that at 200 K on 2 MLE Cu–Ni/SiO2, the 2078 cm 1 stretching peak due to CO adsorbed on bulk or multilayer Cu completely disappears while the 2112 cm 1 stretching peak due to CO bonded on Cu over layers remains as a clear adsorption peak. This observation shows that CO adsorption on Cu over layers is strengthened by interaction of Cu with the underlayer Ni atoms compared to CO adsorption on bulk Cu surfaces. The newly appeared CO adsorption peak on Ni sites, which arises from CO spillover from Cu to Ni at about 200 K, completely disappears after annealing at 350 K, which is consistent with CO TPD results that CO mainly desorbs at 350 K at high Cu coverages (>2 MLE Cu) on Cu–Ni/SiO2. The variable temperature CO PM-IRRAS results clearly show CO spillover from Cu to Ni sites at about 200 K, even the Ni surface sites are totally blocked or covered by Cu at lower temperatures. The CO spillover from Cu to Ni can be rationalized by the larger CO adsorption energy on Ni than that on Cu.50 The possible CO induced Ni surface segregation above 200 K can be expected. Even after CO exposure at low temperature, CO still can induce Ni surface segregation by CO spillover from Cu to Ni sites at about 200 K. This behavior makes CO not an appropriate probe molecule to titrate the surface composition of Cu–Ni bimetallic catalysts. 3.3 CO adsorption on Cu–Ni/SiO2 bimetallic catalysts at elevated pressure Catalytic reactions normally take place at ambient pressure, where adsorption induced surface segregation can become more profound. A 5 MLE Cu-5 MLE Ni/SiO2 bimetallic catalyst was chosen to study the CO interaction with Cu–Ni bimetallic
Fig. 7 PM-IRRAS of CO adsorption on 5.0 MLE Cu-5 MLE Cu/SiO2 surfaces at the indicated CO pressure varying from 1 10 6 Torr to 5.0 Torr, and subsequently pumped down to UHV, and pumped in UHV overnight.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3827
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
catalysts at ambient pressure. Fig. 7 shows in situ CO PM-IRRAS on 5 MLE Cu-5 MLE Ni/SiO2 with the CO pressure varying from 1 10 6 Torr to 5 Torr. Under UHV conditions, there is only one CO adsorption peak at about 2000 cm 1 attributed to atop CO adsorption on Ni sites. Upon increasing CO pressure to 0.05 Torr, there are two new CO adsorption features observed at 2117 cm 1 and 2078 cm 1, which are attributed to CO adsorption on Cu over layers directly attached to Ni surfaces and CO adsorption on multilayer Cu or Cu clusters, respectively.57 Upon further increasing CO pressure to 0.5 Torr, the CO adsorption features on Cu surfaces completely disappear. CO adsorption on Ni sites gain intensity significantly, and its stretching frequency shifts to 2043 cm 1. The blue shift of CO adsorption frequency on Ni sites is caused by the increased CO coverage on Ni sites at elevated pressures evidenced by the increased peak intensity. The extinction of CO adsorption features on Cu surfaces indicates that the Cu–Ni/SiO2 bimetallic nanoparticles are totally encapsulated by the surface layer of Ni induced by high pressure CO. It also indicates that the numbers of totally isolated monometallic Cu clusters on our Cu–Ni/SiO2 samples are limited. The Cu–Ni/SiO2 bimetallic catalysts switch from the structure with Cu enriched on the surface under vacuum conditions to that with Ni surface-enriched under ambient CO pressure conditions. The surface segregation induced by CO is saturated at 0.5 Torr CO. There is no obvious change in the IR spectra with a further increase in the CO pressure up to 5 Torr. After pumping down to UHV, the atop CO stretching peak shifts to 2031 cm 1, and meanwhile there is a broad peak at about 1900 cm 1 attributed to bridging site CO adsorption on Ni sites, which indicates that adjacent Ni sites formed by Ni surface segregation. This bridging site CO adsorption is unstable and disappears after overnight pumping in UHV. However, the atop CO adsorption is very stable. After overnight pumping in UHV, the atop CO adsorption stretching frequency shifts to 2022 cm 1 and maintains a strong adsorption peak. On this sample, CO TPD was conducted to measure the amount of CO adsorption. Fig. 8 shows a comparison between CO TPD on Cu–Ni/bimetallic catalysts saturated with
PCCP
CO under UHV conditions (1 10 6 Torr CO) and saturated with CO under ambient pressure conditions (5 Torr). As shown above, the intensity of CO desorption on 5 MLE Cu-5 MLE Ni/SiO2 is about 10% of that on monometallic 5 MLE Ni/SiO2 when the Cu–Ni/SiO2 bimetallic catalysts were saturated with 1 10 6 Torr CO at room temperature. While the CO TPD from 5 MLE Cu-5 MLE Ni/SiO2 saturated with 5 Torr CO at room temperature, which is subsequently pumped down in UHV overnight, shows about 50% CO desorption of that on monometallic 5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO at room temperature. This TPD result shows that the active sites measured by CO selective adsorption on Cu–Ni/SiO2 bimetallic catalysts under ambient CO pressure conditions can be over estimated by a factor of 4 times higher than the real value even without considering the Ni active sites already over estimated by CO TPD measurements under UHV conditions.
4. Conclusions We have studied CO adsorption on Cu–Ni/SiO2 bimetallic model catalysts supported on SiO2 thin films from UHV to ambient pressure conditions by using CO TPD and in situ PM-IRRAS techniques. The results show that Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic catalysts under vacuum conditions. Under ambient CO pressure conditions (>0.5 Torr), CO adsorption induces Ni surface segregation on the Cu–Ni/ SiO2 bimetallic catalysts. In situ CO PM-IRRAS shows CO spillover from Cu to Ni taking place at about 200 K. It has been demonstrated that large errors can occur in the Ni active site measurements by CO selective chemisorption under ambient pressure conditions because of the Ni surface segregation induced by CO adsorption. Based on the results obtained from the present study, useful information on surface composition and atomic ensemble on bimetallic catalysts could be obtained from the adsorption of CO as a probe molecule; however, caution is always required to quantitatively analyze the adsorption data due to the possible surface segregation induced by CO.
Acknowledgements The authors acknowledge the support from the US Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Bio-sciences (DE-FG02-95ER-14511).
References
Fig. 8 Comparison of CO TPD of (top) 5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO, (middle) 5 MLE Cu-5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO, and (bottom) 5 MLE Cu-5 MLE Ni/SiO2 saturated with 5 Torr CO (pumped in UHV overnight as shown in Fig. 7).
3828 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
1 J. H. Sinfelt, Bimetallic catalysts:discoveries, concepts, and applications, Wiley, New York, 1983. 2 C. T. Campbell, Annu. Rev. Phys. Chem., 1990, 41, 775–837. 3 J. A. Rodriguez, Surf. Sci. Rep., 1996, 24, 223–287. 4 W. Yu, M. D. Porosoff and J. G. Chen, Chem. Rev., 2012, 112, 5780–5817. 5 J. A. Rodriguez and D. W. Goodman, Science, 1992, 257, 897–903. 6 F. Gao and D. W. Goodman, Chem. Soc. Rev., 2012, 41, 8009–8020.
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
7 F. Maroun, F. Ozanam, O. M. Magnussen and R. J. Behm, Science, 2001, 293, 1811–1814. 8 F. Besenbacher, I. Chorkendorff, B. S. Clausen, B. Hammer, A. M. Molenbroek, J. K. Norskov and I. Stensgaard, Science, 1998, 279, 1913–1915. 9 Z. Wei, J. Sun, A. K. Datye and Y. Wang, Chem. Soc. Rev., 2012, 41, 7994–8008. 10 F. Tao, M. E. Grass, Y. Zhang, D. R. Butcher, J. R. Renzas, Z. Liu, J. Y. Chung, B. S. Mun, M. Salmeron and G. A. Somorjai, Science, 2008, 322, 932–934. 11 X. Liu, C. Meng and Y. Han, J. Phys. Chem. C, 2013, 117, 1350–1357. 12 Q. Fu, Y. X. Yao, X. Guo, M. Wei, Y. Ning, H. Liu, F. Yang, Z. Liu and X. H. Bao, Phys. Chem. Chem. Phys., 2013, 15, 14708–14714. 13 R. Mu, Q. Fu, H. Liu, D. Tan, R. Zhai and X. H. Bao, Appl. Surf. Sci., 2009, 255, 7296–7301. 14 F. Tao, M. E. Grass, Y. Zhang, D. R. Butcher, F. Aksoy, S. Aloni, V. Altoe, S. Alayoglu, J. R. Renzas, C. K. Tsung, Z. Zhu, Z. Liu, M. Salmeron and G. A. Somorjai, J. Am. Chem. Soc., 2010, 132, 8697–8703. 15 F. Tao, S. Zhang, L. Nguyen and X. Zhang, Chem. Soc. Rev., 2012, 41, 7980–7993. 16 C. A. Menning and J. G. Chen, J. Chem. Phys., 2009, 130, 174709. 17 F. Tao and M. Salmeron, Science, 2011, 331, 171–174. 18 C. Lamberti, A. Zecchina, E. Groppo and S. Bordiga, Chem. Soc. Rev., 2010, 39, 4951–5001. 19 G. Rupprechter, Adv. Catal., 2007, 51, 133–263. 20 F. Gao, Y. Wang and D. W. Goodman, J. Phys. Chem. C, 2010, 4036–4043. 21 J. H. Sinfelt, J. L. Carter and D. J. C. Yates, J. Catal., 1972, 24, 283–296. 22 J. H. Sinfelt, Acc. Chem. Res., 1977, 10, 15–20. 23 J. H. Sinfelt, G. H. Via and F. W. Lytle, Catal. Rev., 1984, 26, 81–140. 24 J. A. Dalmon and G. A. Martin, J. Catal., 1980, 66, 214–221. 25 I. Alstrup, U. E. Petersen and J. R. Rostrup-Nielsen, J. Catal., 2000, 191, 401–408. 26 H. H. Brongersma and M. J. Sparnany, Surf. Sci., 1978, 71, 657–678. 27 K. C. Khulbe and R. S. Mann, Catal. Rev., 1982, 24, 311–328. 28 V. Ponec, Appl. Catal., A, 2001, 222, 31–45. 29 J. H. Sinfelt, J. Catal., 1973, 29, 308–315. 30 M. Araki and V. Ponec, J. Catal., 1976, 44, 439–448. 31 A. R. Naghash, T. H. Etsell and S. Xu, Chem. Mater., 2006, 18, 2480–2488. ¨ttinger, Catal. Lett., 2013, 32 A. Kitla, O. V. Safonova and K. Fo 143, 517–530. 33 Y. S. Ng, S. B. Mclane Jr. and T. T. Tsong, J. Vac. Sci. Technol., 1980, 17, 154–158. ¨ppers, Surf. Sci., 1971, 71, 657–678. 34 G. Ertl and J. Ku 35 V. Ponec, Mater. Sci. Eng., 1980, 42, 135–140. 36 I. Abbati, L. Braicovich and A. Fasana, Phys. Rev. B: Condens. Matter Mater. Phys., 1981, 23, 6448–6455.
This journal is © the Owner Societies 2014
Paper
37 C. T. Campbell and P. H. Emmett, J. Catal., 1967, 7, 252–262. 38 E. M. Silverman, R. J. Madix and P. Delrue, Surf. Sci., 1981, 109, 127–139. 39 A. Crucq, L. Degols, G. Lienard and A. Frennet, Stud. Surf. Sci. Catal., 1983, 17, 137–147. 40 D. T. Ling and W. E. Spicer, Surf. Sci., 1980, 94, 403–423. 41 X. H. Feng, M. R. Yu, S. Yang, G. Meigs and E. Garfunkel, J. Chem. Phys., 1989, 90, 7516–7523. 42 J. C. M. Harberts, A. F. Bourgonje, J. J. Stephan and V. Ponec, J. Catal., 1977, 47, 92–99. 43 S. Andersson and J. B. Pendry, J. Phys. C: Solid State Phys., 1980, 13, 3547–3561. 44 M. S. Kharson, A. A. Slinkin, E. A. Fedorovskaya, V. G. Pimenov and S. L. Kiperman, React. Kinet. Catal. Lett., 1984, 24, 389–393. 45 J. H. Lin and V. V. Guliants, Appl. Catal., A, 2012, 445–446, 187–194. 46 Y. X. Yao, Z. Yan, L. Chen, Z. Zhou and D. W. Goodman, Catal. Lett., 2012, 142, 1437–1444. ¨ck, Surf. Sci., 2000, 47 H. Koschel, G. Held and H. P. Steinru 453, 201–213. 48 X. Xu and D. W. Goodman, J. Phys. Chem., 1993, 97, 683–689. 49 F. P. Netzer and T. E. Madey, J. Chem. Phys., 1982, 76, 710–715. 50 E. Demirci, C. Carbogno, A. Groß and A. Winkler, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 085421. 51 Z. Yan and D. W. Goodman, Catal. Lett., 2012, 142, 517–520. 52 M. Trenary, K. J. Uram and J. T. Yates Jr., Surf. Sci., 1985, 157, 512–538. 53 C. H. F. Peden and D. W. Goodman, Ing. Eng. Chem. Fundam., 1986, 25, 58–62. 54 E. Garfunkel, M. R. Yu, S. Yang and X. H. Feng, J. Vac. Sci. Technol., A, 1989, 7, 1579–1584. 55 S. Yang, M. R. Yu, G. Meigs, X. H. Feng and E. Garfunkel, Surf. Sci., 1988, 205, L777–L785. ¨ck, Surf. 56 G. Held, W. Sklarek, M. Mayan and H. P. Steinru Sci., 1998, 402–404, 322–326. 57 J. W. He, W. K. Kuhn and D. W. Goodman, Mater. Res. Soc. Symp. Proc., 1991, 202, 325–331. 58 F. M. Hoffmann and L. Pan, J. Chem. Phys., 1987, 87, 1857–1865. 59 F. M. Hoffmann and J. Paul, J. Chem. Phys., 1987, 86, 2990–2996. 60 X. Xu, S. M. Vesecky and D. W. Goodman, Science, 1992, 258, 788–790. 61 K. Yao, X. Liu, Z. Li, C. Li, H. Zeng and Y. Han, ChemCatChem, 2012, 4, 1938–1942. 62 X. Liu, K. Yao, C. Meng and Y. Han, Dalton Trans., 2012, 41, 1289–1296. 63 X. Liu, C. Meng and Y. Han, Nanoscale, 2012, 4, 2288–2295. 64 Y. X. Yao, X. Liu, Q. Fu, W. X. Li, D. L. Tan and X. H. Bao, ChemPhysChem, 2008, 9, 975–979. 65 Q. Fu, W. X. Li, Y. X. Yao, H. Liu, H. Su, D. Ma, X. Gu, L. Chen, Z. Wang, B. Wang and X. H. Bao, Science, 2010, 328, 1141–1144.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3829
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 3823
View Journal | View Issue
In situ IR spectroscopic studies of Ni surface segregation induced by CO adsorption on Cu–Ni/SiO2 bimetallic catalysts Yunxi Yao* and D. Wayne Goodman† It is of great importance to study the catalytic structures under real reaction conditions especially for the bimetallic catalysts, where facile surface restructure or surface segregation can be driven by adsorbate adsorption. Here, we report CO interaction with Cu–Ni/SiO2 bimetallic model catalysts studied by CO temperature programmed desorption (TPD) and in situ CO polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS) under CO pressures varying from ultrahigh vacuum (UHV) to near ambient pressure. Under UHV conditions, Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic cata-
Received 27th November 2013, Accepted 24th December 2013
lysts. CO spillover from Cu to Ni on Cu–Ni/SiO2 bimetallic catalysts has been observed at about 200 K
DOI: 10.1039/c3cp54997f
catalysts induced by CO adsorption at ambient pressure CO. The behavior of CO induced surface segre-
under UHV conditions. In situ CO PM-IRRAS shows surface segregation of Ni on the Cu–Ni/SiO2 bimetallic gation can lead to severe errors in Ni active site measurements by the selective CO chemisorption on
www.rsc.org/pccp
Cu–Ni/SiO2 bimetallic catalysts.
1. Introduction Bimetallic catalysts have attracted great research efforts in the past decades due to their chemical and physical properties different from the individual pure metals and promising applications in chemical conversion, energy technology and environmental protection.1–11 Extensive fundamental insights into the surface composition, surface atomic ensemble, electronic structure, and gas adsorption property of bimetallic catalysts have been gained.2–8 Most of these studies were conducted under ex situ or vacuum conditions. Bimetallic catalysts, mostly used as supported bimetallic nanoparticles, are very flexible to restructuring and surface segregation induced by ambient pressure gas adsorption or reaction.10,12–16 The surface structure, atomic composition, and chemical state of bimetallic catalysts obtained by ex situ techniques or ultrahigh vacuum based surface science studies are often different from those under ambient pressure adsorption or reaction conditions. Catalyst characterization under elevated pressure or reaction conditions, especially for bimetallic catalysts, becomes more demanding to understand the structure–activity relationship. Many in situ techniques have been developed for this purpose,17 and among them in situ IR is probably the mostly widely used and developed due to its low cost, easy setup, and Department of chemistry, Texas A&M University, College Station, TX 77843, USA. E-mail: [email protected] † Dr Goodman deceased on February 27, 2012.
This journal is © the Owner Societies 2014
effectiveness for catalyst characterization.18,19 Polarization modulation infrared reflection absorption spectroscopy (PMIRRAS), as one peculiar IR technique, can be operated in reflection mode with polarized IR light, which enables the surface vibrational spectra to be obtained by the subtraction of gas-phase absorption.19 This technique allows in situ measurements of catalytic surfaces from ultrahigh vacuum (UHV) to atmospheric conditions.20 Since the pioneering work of Sinfelt et al. on the hydrogenolysis of ethane over Cu–Ni alloys in 1972,21 Cu–Ni bimetallic catalysts have been considered as a representative system to study the structure–activity relationship in bimetallic catalysts.1,22–28 Many catalytic reactions, which include hydrogenation, dehydrogenation, hydrogenolysis, isomerization, and Fischer–Tropsch synthesis, have been studied using Cu–Ni bimetallic catalysts.1,22–27,29,30 Extensive and controversial studies on the surface structure, surface composition, electronic structure, and structure–activity relationship of bimetallic Cu–Ni systems have been reported, and the main findings are summarized as follows: (1) it is widely reported that Cu is enriched on the surface of Cu–Ni alloys,22,25,26,31–33 but there are two exceptions which report that the surface of Cu–Ni bimetallic catalysts has a bulk composition.24,34 (2) Ponec reviewed the reported results and found that there are no essential changes in the electronic structure of Ni atoms caused by alloying with Cu atoms,35 whereas Abbati et al. found that d band photoemission from Ni is modified in the presence of Cu on the surface and there is a significant interaction between the
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3823
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
d states of Cu and Ni atoms.36 (3) The sharp decrease in the catalytic activity of Cu–Ni bimetallic catalysts upon the addition of Cu is usually explained by the ensemble effect,1,21,22,24,25,29 while Alstrup et al. found the advanced ensemble model to be invalid in the explanation of the sulfur modification effect on nickel catalysts in propane hydrogenolysis, and they proposed that special sites are needed for breaking of the C–C bonds in the hydrogenolysis reactions on Cu–Ni bimetallic catalysts. (4) The Cu effect on the activity of Ni remains controversial. There is a three-to-five order decrease in the specific activity of ethane or propane hydrogenolysis on Cu–Ni bimetallic catalysts with increasing Cu amount.1,21,22,24,25,29 Campbell and Emmett reported that the rate of ethylene hydrogenation on Ni was increased by 7 to 15 times upon addition of Cu to Ni catalysts.37 (5) H2 and CO adsorption on Cu–Ni bimetallic catalysts was extensively studied. The selective hydrogen chemisorption was used to measure the surface active Ni sites on Cu–Ni catalysts.1,21,22,24,29,38 But Crucq and co-workers found hydrogen spillover from Ni to Cu on Cu–Ni bimetallic surfaces,39 which can lead to over counting of surface Ni sites in the hydrogen selective chemisorption.32 CO has been used as a probe molecule to study the surface structure and composition of Cu–Ni bimetallic catalysts.35,40–44 In spite of extensive studies on Cu–Ni bimetallic catalysts, in situ studies on the interaction of CO with Cu–Ni bimetallic catalysts are still missing. Based on the large difference in bond strength between CO on Ni and Cu sites, possible surface segregation induced by CO adsorption can be expected, which is the key to understand CO involved reactions on Cu–Ni bimetallic catalysts, such as methanation of CO30 and the water-gas-shift reaction on Cu–Ni bimetallic catalysts.45 In the present work, we studied CO adsorption on Cu–Ni bimetallic model catalysts supported on SiO2 thin films from UHV to ambient pressures by TPD and in situ PM-IRRAS. By taking the advantage of surface science techniques and the UHV coupled high pressure in situ PM-IRRAS, the surface composition and structure of Cu–Ni bimetallic catalysts under vacuum conditions, and also under ambient pressure conditions, were thoroughly investigated. Cu is found to be enriched on the surface of Cu–Ni bimetallic catalysts under UHV conditions, but CO can induce Ni surface segregation at elevated pressure.
PCCP
Hinds PEM-90 photoelastic modulator) spectra were recorded using 600 scans at a resolution of 4 cm 1. A linear ramp of 3 K s 1 was used for TPD measurements. The Mo(110) sample was mounted onto the end of the sample manipulator. The sample temperature can be measured using a pair of W-5%Re–W-26%Re C type thermocouple spotwelded on the back of the sample. Thin SiO2 films (B10 nm) were prepared by evaporating Si in 1 10 5 Torr O2 at room temperature and subsequently annealed to 1200 K.46 Cu and Ni were deposited by heating tantalum wires wrapped with highly pure Ni or Cu wires. The amount of metal deposited was controlled by the depositing time and calibrated with AES spectra, and was given in the unit of monolayer equivalent (MLE, 1 MLE = 1.43 1015 atom per cm2).46 C. P. grade CO (Matheson Tri-Gas) was used and further purified by passing it through a LN2 cooling trap and stored in a glass bulb maintained at LN2 temperature throughout the CO adsorption measurements.
3. Results and discussion 3.1 Room temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts Cu–Ni bimetallic catalysts were grown on SiO2 by depositing different amounts of Cu on 5 MLE Ni/SiO2, which was subsequently annealed at 700 K for 5 min. Fig. 1a shows the room temperature CO TPD after saturated CO exposure on Cu–Ni/SiO2 bimetallic catalysts as a function of Cu coverage. Under UHV conditions, CO desorbs on Cu surfaces below 250 K,47,48 so CO exclusively adsorbs on Ni sites on Cu–Ni bimetallic catalysts at room temperature. CO TPD shows two CO desorption peaks at 350 K and 410 K on clean 5 MLE Ni/SiO2, which is similar to saturated CO desorption from Ni(111).49 Upon the addition of Cu, the low temperature desorption peak keeps its position at 350 K, while the high temperature peak gradually shifts from 410 K on 5 MLE Ni/SiO2 to 387 K on 5 MLE Cu-5 MLE Ni/SiO2. The observed behavior is consistent with CO desorption from Cu(110) covered with Ni over layers.50 However, the weakened
2. Experimental All the experiments were conducted in a ultrahigh vacuum chamber using Auger electron spectroscopy (AES), polarization modulation infrared reflection absorption spectroscopy (PM-IRRAS), and a UTI-100 mass spectrometer for temperature programmed desorption (TPD) measurements, which has been described in detail previously.46 A high pressure/IR cell equipped with CaF2 windows was attached to the bottom of the UHV chamber, which can be isolated from the main chamber by differentially pumped sliding seals for high pressure studies. The base pressure for both the main chamber and the high pressure/IR cell is less than 2 10 10 Torr. PM-IRRAS (Bruker Equinox 55,
3824 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
Fig. 1 (a) CO TPD of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at room temperature with indicated Cu coverage. (b) Normalized CO TPD area as a function of Cu coverage.
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
CO adsorption on Ni upon alloying with Cu, as observed here on Cu–Ni/SiO2 bimetallic catalysts, was not observed on the Ni(111) surface upon the addition of Cu atoms,47 which indicates that the chemical properties of supported Ni nanoparticles are more prone to be modified upon alloying with Cu than bulk Ni surfaces. The low temperature desorption feature has been attributed to CO desorption from low coordinated step Ni sites on Ni nanoparticles, while the high temperature desorption feature is due to CO desorption on terrace sites.46 The TPD results hint that addition of Cu has a more profound effect on CO adsorption on terrace sites than on step sites on supported Ni nanoparticles. Fig. 1b shows the normalized CO desorption intensity as a function of Cu coverage. CO desorption from Ni sharply decreases with increasing Cu coverage up to 1.5 MLE. We have measured the surface active Ni sites on 5 MLE Ni/SiO2 by H2 TPD to be 1.7 MLE (1.93 1015 Ni atoms). CO TPD results show that the Ni surface sites are blocked by Cu atoms almost on a one on one basis. This behavior is consistent with CO TPD from Cu layers on Ni(111) where Ni surface sites are totally covered with 1 ML Cu.47 At low Cu coverages (up to 1.5 MLE), Cu atoms prefer to populate on the surface, which agrees with the surface enrichment of Cu on the Cu–Ni bimetallic catalysts.22,25,26,31–33 After deposition of 2 MLE Cu, the CO desorption intensity remains constant, which is about 10% of CO desorption from the clean 5 MLE Ni/SiO2 surface. Fig. 2 shows the room temperature PM-IRRAS spectra after saturated exposure of CO on Cu–Ni/SiO2 with increasing Cu coverage. The PM-IRRAS of the monometallic Ni/SiO2 catalyst shows two stretching frequencies at 1950 cm 1 and 2058 cm 1, which are attributed to CO adsorbed on atop and bridging Ni sites, respectively.41,51,52 After deposition of 0.5 MLE Cu on the 5 MLE Ni/SiO2, the bridge site CO adsorption peak completely disappears, indicating a sharp decrease in the adjacent Ni sites by the addition of Cu atoms. The intensity of the atop CO adsorption peak increases and its stretching frequency shifts from 2058 cm 1 on the monometallic Ni/SiO2 to 2024 cm 1 on the 0.5 MLE Cu–Ni/SiO2. The CO stretching frequency continues to decrease to 1999 cm 1 at 1.5 MLE Cu–Ni/SiO2, and
Fig. 2 Room temperature PM-IRRAS of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces with indicated Cu coverage. The Cu–Ni/SiO2 surfaces were exposed to 1 10 6 Torr CO for 10 min.
This journal is © the Owner Societies 2014
Paper
afterwards it remains constant at 1999 cm 1, which resembles the stretching frequency of CO at 2000 cm 1 from CO adsorption on 1 ML Cu/Ni(111).41 After the addition of 0.5 MLE Cu, the intensity of the atop CO peak starts to decrease with increasing Cu coverage up to 2 MLE, and after that the atop CO peak intensity remains constant. The decrease in CO adsorption intensity is due to the surface Ni sites being blocked by Cu atoms. The large red shifts of atop CO adsorption and the absence of the bridging CO adsorption indicate that the surface Ni sites are isolated with the added Cu atoms and there is a weak dipole coupling of adsorbed CO molecules. Room temperature CO adsorption results show that Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic catalysts, which was further confirmed by our AES results. Fig. 3a shows AES spectra of Cu–Ni/SiO2 with increasing Cu coverage. With increasing Cu coverage, the Cu AES peaks increase while the Ni AES peaks decrease in intensity due to the Ni AES signal attenuated by Cu over layers. Fig. 3b shows the intensity changes of Ni and Cu as a function of Cu coverage. The Ni AES peak at about 740 eV was used instead of the main Ni AES peak at about 870 eV because it is overlapped with Cu AES signals. The Ni and Cu AES intensities are normalized to their maximum intensities, respectively. From the plots in Fig. 3b we can see that the Cu AES intensity linearly increases while the Ni AES intensity linearly decreases with increasing Cu coverages up to 1.5 MLE, which indicates Cu growth on Ni follows a layerby-layer mechanism, similar to Cu grown on Ru(0001).53 A mixed Cu–Ni layer can be formed at 700 K as observed in Cu–Ni(111) systems.41,54 But here the Cu atoms deposited at coverages less than 1.5 MLE mainly populate on the surface blocking CO adsorption as observed by CO TPD, which agrees with the Cu atoms segregating on the Cu–Ni bimetallic surface and form Cu over layers due to its lower surface free energy.26,41,47,55,56 As we mentioned before, the active Ni sites on the monometallic 5 MLE Ni/SiO2 are measured to be 1.7 MLE, which coincidently equals to the observed breaking point in the
Fig. 3 (a) AES of Cu-5 MLE Ni/SiO2 surfaces with indicated Cu coverage. (b) The evolution of Ni (740 eV) and Cu (950 eV) AES peak intensity as a function of Cu coverage. The Ni and Cu AES peak intensities were normalized to their maximum value, respectively. The Ni AES peak at about 740 eV was used instead of the main Ni AES peak at about 870 eV because it is overlapped with Cu AES signals.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3825
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
PCCP
AES spectra. After deposition of 1.5 MLE Cu, the Ni AES intensity decreases by about 18%, which agrees so well with the calculated value of 18% decrease in Ni AES signal due to the attenuation by one atomic Cu over layer by using the electron inelastic mean free paths at 740 eV in Cu of 1.3 nm and the thickness of one atomic Cu layer of 0.255 nm. AES results combined with room temperature CO adsorption show Cu atoms mainly populate on the Ni surface and block the Ni active sites at low Cu coverages, which may account for the sharp decrease in catalytic activity.21,22,24 3.2 Low temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts In order to study CO adsorption on Cu sites and the temperature dependent CO adsorption dynamics on Cu–Ni/SiO2 bimetallic catalysts, CO adsorption on Cu–Ni/SiO2 at low temperatures (around 130 K) was investigated. Fig. 4 shows CO TPD on Cu-5 MLE Ni/SiO2 with indicated Cu coverage saturated with CO at 130 K. On the monometallic 5 MLE Ni/SiO2, the main CO desorption peak is observed at 415 K, which is consistent with the room temperature CO adsorption results (Fig. 1). Below room temperature, there are two CO desorption peaks at about 200 K and 266 K. CO TPD on Ni(111) with CO dosed at low temperatures also shows a broad desorption peak below 300 K due to the increased CO coverage at low temperature.49,56 After the addition of 0.2 MLE Cu, the main CO desorption peak at 415 K on 5 MLE Ni/SiO2 shifts to 384 K, which agrees with room temperature observations and indicates that CO adsorption on Ni sites is weakened upon addition of Cu. Meanwhile, a new desorption peak appears at about 160 K attributed to CO desorption from Cu surfaces.48 Upon further increasing the Cu coverage to 0.5 MLE, besides the CO desorption from Ni sites which continues to decrease in intensity, there are two well-resolved CO desorption peaks at about 160 K and 210 K, which are similar to CO desorption from Cu/SiO2.48 From the CO TPD results, we cannot tell whether CO adsorption on Cu sites is strengthened or not. After deposition of 2 MLE Cu on Ni/SiO2, the CO desorption from Ni sites (integration of CO desorption
Fig. 4 CO TPD of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at about 130 K with indicated Cu coverage.
3826 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
Fig. 5 PM-IRRAS of saturated CO adsorption on Cu-5 MLE Ni/SiO2 surfaces at 130 K with indicated Cu coverage.
above 300 K) is only about 4% of the original value from clean monometallic 5 MLE Ni/SiO2. This result is contrary to room temperature observation, where 10% CO desorption intensity remains on 2 MLE Cu–Ni/SiO2 surfaces. The larger CO adsorption capacity at room temperature than that at low temperature hints that possible CO induced Ni surface segregation at room temperature takes place. Low temperature CO adsorption on Cu–Ni/SiO2 bimetallic catalysts was also studied by CO PM-IRRAS. Fig. 5 shows CO PM-IRRAS after saturated exposure of CO at about 130 K on Cu-5 MLE Ni/SiO2 with increasing Cu coverage. After the addition of 0.2 MLE Cu, the bridging CO adsorption on Ni sites totally disappears, and the atop CO stretching frequency shifts from 2070 cm 1 on 5 MLE Ni/SiO2 to 2028 cm 1. A new CO stretching frequency at 2126 cm 1 appears, which is attributed to atop CO on isolated Cu atoms or small Cu clusters.41,57 At low Cu coverages, the existence of single Cu atoms on Ni clusters can be expected. Upon further increasing Cu coverage to 0.5 MLE, the CO adsorption peaks on Ni sites completely disappear. The atop CO adsorption frequency on Cu sites shifts from 2126 cm 1 at 0.2 MLE Cu coverage to 2114 cm 1 at 0.5 MLE Cu coverage. The 2114 cm 1 CO stretching peak is attributed to CO adsorption on Cu over layers on Ni clusters.57–59 A new CO stretching peak appears at 2076 cm 1, which is similar to the CO stretching frequency from CO adsorption on Cu(111),55,60 indicating that multilayer Cu or Cu clusters form at this Cu coverage. From the evolution trend of CO adsorption spectra shown in Fig. 5, the possibility of this 2076 cm 1 peak due to CO adsorbed on Ni sites is excluded. At the Cu coverages above 1 MLE, there are two CO stretching peaks at 2105 cm 1 and 2076 cm 1, which maintain their positions with increasing Cu coverage. Compared with room temperature CO adsorption results, CO adsorption on Ni sites is totally blocked by 0.5 MLE Cu 130 K, while there is a clear CO adsorption peak on Ni detected at 1999 cm 1 even at a Cu coverage of 5 MLE. Another difference is that at room temperature the CO adsorption on Ni sites is blocked by Cu on a one on one basis, but at low temperature, it seems one Cu atom can block more than one Ni sites.41
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
Fig. 6 PM-IRRAS of saturated CO adsorption on (a) 0.2 MLE Cu-5 MLE Cu/SiO2 and (b) 2.0 MLE Cu-5 MLE Cu/SiO2 at about 150 K and subsequently annealed at indicated temperatures.
Another information from low temperature CO PM-IRRAS is the blue shift of CO bound to Cu over layers with respect to CO adsorbed on multilayer Cu, which indicates that, between Ni clusters and Cu overalyers, there is interface charge transfer or electronic modification.61–65 Goodman et al. have proposed that the blue shift of CO adsorption on Cu over layers on Rh(100) is due to CO bound to positively charged or selfpolarized Cu.57 However, the d-electron back donation becomes less dominant for CO adsorption on Cu than for transition metals.59 The blue shift may arise due to the electronic structure of Cu modified via the Ni–Cu interface interaction.36 Temperature effects on CO adsorption on Cu–Ni/SiO2 were studied by variable temperature CO PM-IRRAS. Fig. 6a shows CO PM-IRRAS after saturated CO exposure on 0.2 MLE Cu–Ni/SiO2 at 150 K and subsequently increasing the sample temperature to the indicated temperatures. At 150 K, there are two adsorption peaks at 2116 cm 1 and 2016 cm 1 attributed respectively to atop CO adsorption on Cu and Ni sites on 0.2 MLE Cu–Ni/SiO2. Upon increasing the sample temperature to 200 K, the intensity of CO adsorbed on Cu decreases, but its peak position remains unchanged. The CO adsorption on Ni sites increases in intensity and its peak position blue shifts from 2016 cm 1 to 2032 cm 1. The blue shift can be explained by the increased CO coverage at Ni sites, which is evidenced by the increased intensity of the corresponding CO stretching peak. Upon further increasing the sample temperature to 250 K, CO bound to Cu sites completely disappears, which is consistent with our CO TPD results (Fig. 4). CO adsorbed on atop Ni sites further gains in intensity and its peak position is maintained at 2032 cm 1. Upon further increasing the sample temperature, the CO stretching frequency shifts to 2005 cm 1 due to the decreased intermolecular interaction by partial desorption of CO on Ni sites. The CO PM-IRRAS results on 0.2 MLE Cu–Ni/SiO2 show CO spillover from Cu to Ni at about 200 K accompanied with CO partial desorption from Cu surfaces. This result indicates that CO adsorption at 200 K can induce Ni surface segregation on Cu–Ni/SiO2 bimetallic catalysts. Fig. 6b shows another variable temperature CO PM-IRRAS on 2 MLE Cu–Ni/SiO2 saturated with CO at 150 K. No CO adsorption on Ni sites is detected at 150 K showing that all
This journal is © the Owner Societies 2014
Paper
the surface Ni sites are blocked by 2 MLE Cu. However, upon increasing the sample temperature to 200 K, a CO adsorption peak appears at 1997 cm 1 attributed to atop CO on Ni sites, while a sharp decrease of CO adsorption on Cu sites is observed. The evolution of the CO adsorption peaks indicates CO spillover from Cu to Ni while CO partially desorbs from Cu surfaces at about 200 K. It should be noted that at 200 K on 2 MLE Cu–Ni/SiO2, the 2078 cm 1 stretching peak due to CO adsorbed on bulk or multilayer Cu completely disappears while the 2112 cm 1 stretching peak due to CO bonded on Cu over layers remains as a clear adsorption peak. This observation shows that CO adsorption on Cu over layers is strengthened by interaction of Cu with the underlayer Ni atoms compared to CO adsorption on bulk Cu surfaces. The newly appeared CO adsorption peak on Ni sites, which arises from CO spillover from Cu to Ni at about 200 K, completely disappears after annealing at 350 K, which is consistent with CO TPD results that CO mainly desorbs at 350 K at high Cu coverages (>2 MLE Cu) on Cu–Ni/SiO2. The variable temperature CO PM-IRRAS results clearly show CO spillover from Cu to Ni sites at about 200 K, even the Ni surface sites are totally blocked or covered by Cu at lower temperatures. The CO spillover from Cu to Ni can be rationalized by the larger CO adsorption energy on Ni than that on Cu.50 The possible CO induced Ni surface segregation above 200 K can be expected. Even after CO exposure at low temperature, CO still can induce Ni surface segregation by CO spillover from Cu to Ni sites at about 200 K. This behavior makes CO not an appropriate probe molecule to titrate the surface composition of Cu–Ni bimetallic catalysts. 3.3 CO adsorption on Cu–Ni/SiO2 bimetallic catalysts at elevated pressure Catalytic reactions normally take place at ambient pressure, where adsorption induced surface segregation can become more profound. A 5 MLE Cu-5 MLE Ni/SiO2 bimetallic catalyst was chosen to study the CO interaction with Cu–Ni bimetallic
Fig. 7 PM-IRRAS of CO adsorption on 5.0 MLE Cu-5 MLE Cu/SiO2 surfaces at the indicated CO pressure varying from 1 10 6 Torr to 5.0 Torr, and subsequently pumped down to UHV, and pumped in UHV overnight.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3827
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
Paper
catalysts at ambient pressure. Fig. 7 shows in situ CO PM-IRRAS on 5 MLE Cu-5 MLE Ni/SiO2 with the CO pressure varying from 1 10 6 Torr to 5 Torr. Under UHV conditions, there is only one CO adsorption peak at about 2000 cm 1 attributed to atop CO adsorption on Ni sites. Upon increasing CO pressure to 0.05 Torr, there are two new CO adsorption features observed at 2117 cm 1 and 2078 cm 1, which are attributed to CO adsorption on Cu over layers directly attached to Ni surfaces and CO adsorption on multilayer Cu or Cu clusters, respectively.57 Upon further increasing CO pressure to 0.5 Torr, the CO adsorption features on Cu surfaces completely disappear. CO adsorption on Ni sites gain intensity significantly, and its stretching frequency shifts to 2043 cm 1. The blue shift of CO adsorption frequency on Ni sites is caused by the increased CO coverage on Ni sites at elevated pressures evidenced by the increased peak intensity. The extinction of CO adsorption features on Cu surfaces indicates that the Cu–Ni/SiO2 bimetallic nanoparticles are totally encapsulated by the surface layer of Ni induced by high pressure CO. It also indicates that the numbers of totally isolated monometallic Cu clusters on our Cu–Ni/SiO2 samples are limited. The Cu–Ni/SiO2 bimetallic catalysts switch from the structure with Cu enriched on the surface under vacuum conditions to that with Ni surface-enriched under ambient CO pressure conditions. The surface segregation induced by CO is saturated at 0.5 Torr CO. There is no obvious change in the IR spectra with a further increase in the CO pressure up to 5 Torr. After pumping down to UHV, the atop CO stretching peak shifts to 2031 cm 1, and meanwhile there is a broad peak at about 1900 cm 1 attributed to bridging site CO adsorption on Ni sites, which indicates that adjacent Ni sites formed by Ni surface segregation. This bridging site CO adsorption is unstable and disappears after overnight pumping in UHV. However, the atop CO adsorption is very stable. After overnight pumping in UHV, the atop CO adsorption stretching frequency shifts to 2022 cm 1 and maintains a strong adsorption peak. On this sample, CO TPD was conducted to measure the amount of CO adsorption. Fig. 8 shows a comparison between CO TPD on Cu–Ni/bimetallic catalysts saturated with
PCCP
CO under UHV conditions (1 10 6 Torr CO) and saturated with CO under ambient pressure conditions (5 Torr). As shown above, the intensity of CO desorption on 5 MLE Cu-5 MLE Ni/SiO2 is about 10% of that on monometallic 5 MLE Ni/SiO2 when the Cu–Ni/SiO2 bimetallic catalysts were saturated with 1 10 6 Torr CO at room temperature. While the CO TPD from 5 MLE Cu-5 MLE Ni/SiO2 saturated with 5 Torr CO at room temperature, which is subsequently pumped down in UHV overnight, shows about 50% CO desorption of that on monometallic 5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO at room temperature. This TPD result shows that the active sites measured by CO selective adsorption on Cu–Ni/SiO2 bimetallic catalysts under ambient CO pressure conditions can be over estimated by a factor of 4 times higher than the real value even without considering the Ni active sites already over estimated by CO TPD measurements under UHV conditions.
4. Conclusions We have studied CO adsorption on Cu–Ni/SiO2 bimetallic model catalysts supported on SiO2 thin films from UHV to ambient pressure conditions by using CO TPD and in situ PM-IRRAS techniques. The results show that Cu is enriched on the surface of Cu–Ni/SiO2 bimetallic catalysts under vacuum conditions. Under ambient CO pressure conditions (>0.5 Torr), CO adsorption induces Ni surface segregation on the Cu–Ni/ SiO2 bimetallic catalysts. In situ CO PM-IRRAS shows CO spillover from Cu to Ni taking place at about 200 K. It has been demonstrated that large errors can occur in the Ni active site measurements by CO selective chemisorption under ambient pressure conditions because of the Ni surface segregation induced by CO adsorption. Based on the results obtained from the present study, useful information on surface composition and atomic ensemble on bimetallic catalysts could be obtained from the adsorption of CO as a probe molecule; however, caution is always required to quantitatively analyze the adsorption data due to the possible surface segregation induced by CO.
Acknowledgements The authors acknowledge the support from the US Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Bio-sciences (DE-FG02-95ER-14511).
References
Fig. 8 Comparison of CO TPD of (top) 5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO, (middle) 5 MLE Cu-5 MLE Ni/SiO2 saturated with 1 10 6 Torr CO, and (bottom) 5 MLE Cu-5 MLE Ni/SiO2 saturated with 5 Torr CO (pumped in UHV overnight as shown in Fig. 7).
3828 | Phys. Chem. Chem. Phys., 2014, 16, 3823--3829
1 J. H. Sinfelt, Bimetallic catalysts:discoveries, concepts, and applications, Wiley, New York, 1983. 2 C. T. Campbell, Annu. Rev. Phys. Chem., 1990, 41, 775–837. 3 J. A. Rodriguez, Surf. Sci. Rep., 1996, 24, 223–287. 4 W. Yu, M. D. Porosoff and J. G. Chen, Chem. Rev., 2012, 112, 5780–5817. 5 J. A. Rodriguez and D. W. Goodman, Science, 1992, 257, 897–903. 6 F. Gao and D. W. Goodman, Chem. Soc. Rev., 2012, 41, 8009–8020.
This journal is © the Owner Societies 2014
View Article Online
Published on 03 January 2014. Downloaded by University of California - San Diego on 15/09/2014 10:23:43.
PCCP
7 F. Maroun, F. Ozanam, O. M. Magnussen and R. J. Behm, Science, 2001, 293, 1811–1814. 8 F. Besenbacher, I. Chorkendorff, B. S. Clausen, B. Hammer, A. M. Molenbroek, J. K. Norskov and I. Stensgaard, Science, 1998, 279, 1913–1915. 9 Z. Wei, J. Sun, A. K. Datye and Y. Wang, Chem. Soc. Rev., 2012, 41, 7994–8008. 10 F. Tao, M. E. Grass, Y. Zhang, D. R. Butcher, J. R. Renzas, Z. Liu, J. Y. Chung, B. S. Mun, M. Salmeron and G. A. Somorjai, Science, 2008, 322, 932–934. 11 X. Liu, C. Meng and Y. Han, J. Phys. Chem. C, 2013, 117, 1350–1357. 12 Q. Fu, Y. X. Yao, X. Guo, M. Wei, Y. Ning, H. Liu, F. Yang, Z. Liu and X. H. Bao, Phys. Chem. Chem. Phys., 2013, 15, 14708–14714. 13 R. Mu, Q. Fu, H. Liu, D. Tan, R. Zhai and X. H. Bao, Appl. Surf. Sci., 2009, 255, 7296–7301. 14 F. Tao, M. E. Grass, Y. Zhang, D. R. Butcher, F. Aksoy, S. Aloni, V. Altoe, S. Alayoglu, J. R. Renzas, C. K. Tsung, Z. Zhu, Z. Liu, M. Salmeron and G. A. Somorjai, J. Am. Chem. Soc., 2010, 132, 8697–8703. 15 F. Tao, S. Zhang, L. Nguyen and X. Zhang, Chem. Soc. Rev., 2012, 41, 7980–7993. 16 C. A. Menning and J. G. Chen, J. Chem. Phys., 2009, 130, 174709. 17 F. Tao and M. Salmeron, Science, 2011, 331, 171–174. 18 C. Lamberti, A. Zecchina, E. Groppo and S. Bordiga, Chem. Soc. Rev., 2010, 39, 4951–5001. 19 G. Rupprechter, Adv. Catal., 2007, 51, 133–263. 20 F. Gao, Y. Wang and D. W. Goodman, J. Phys. Chem. C, 2010, 4036–4043. 21 J. H. Sinfelt, J. L. Carter and D. J. C. Yates, J. Catal., 1972, 24, 283–296. 22 J. H. Sinfelt, Acc. Chem. Res., 1977, 10, 15–20. 23 J. H. Sinfelt, G. H. Via and F. W. Lytle, Catal. Rev., 1984, 26, 81–140. 24 J. A. Dalmon and G. A. Martin, J. Catal., 1980, 66, 214–221. 25 I. Alstrup, U. E. Petersen and J. R. Rostrup-Nielsen, J. Catal., 2000, 191, 401–408. 26 H. H. Brongersma and M. J. Sparnany, Surf. Sci., 1978, 71, 657–678. 27 K. C. Khulbe and R. S. Mann, Catal. Rev., 1982, 24, 311–328. 28 V. Ponec, Appl. Catal., A, 2001, 222, 31–45. 29 J. H. Sinfelt, J. Catal., 1973, 29, 308–315. 30 M. Araki and V. Ponec, J. Catal., 1976, 44, 439–448. 31 A. R. Naghash, T. H. Etsell and S. Xu, Chem. Mater., 2006, 18, 2480–2488. ¨ttinger, Catal. Lett., 2013, 32 A. Kitla, O. V. Safonova and K. Fo 143, 517–530. 33 Y. S. Ng, S. B. Mclane Jr. and T. T. Tsong, J. Vac. Sci. Technol., 1980, 17, 154–158. ¨ppers, Surf. Sci., 1971, 71, 657–678. 34 G. Ertl and J. Ku 35 V. Ponec, Mater. Sci. Eng., 1980, 42, 135–140. 36 I. Abbati, L. Braicovich and A. Fasana, Phys. Rev. B: Condens. Matter Mater. Phys., 1981, 23, 6448–6455.
This journal is © the Owner Societies 2014
Paper
37 C. T. Campbell and P. H. Emmett, J. Catal., 1967, 7, 252–262. 38 E. M. Silverman, R. J. Madix and P. Delrue, Surf. Sci., 1981, 109, 127–139. 39 A. Crucq, L. Degols, G. Lienard and A. Frennet, Stud. Surf. Sci. Catal., 1983, 17, 137–147. 40 D. T. Ling and W. E. Spicer, Surf. Sci., 1980, 94, 403–423. 41 X. H. Feng, M. R. Yu, S. Yang, G. Meigs and E. Garfunkel, J. Chem. Phys., 1989, 90, 7516–7523. 42 J. C. M. Harberts, A. F. Bourgonje, J. J. Stephan and V. Ponec, J. Catal., 1977, 47, 92–99. 43 S. Andersson and J. B. Pendry, J. Phys. C: Solid State Phys., 1980, 13, 3547–3561. 44 M. S. Kharson, A. A. Slinkin, E. A. Fedorovskaya, V. G. Pimenov and S. L. Kiperman, React. Kinet. Catal. Lett., 1984, 24, 389–393. 45 J. H. Lin and V. V. Guliants, Appl. Catal., A, 2012, 445–446, 187–194. 46 Y. X. Yao, Z. Yan, L. Chen, Z. Zhou and D. W. Goodman, Catal. Lett., 2012, 142, 1437–1444. ¨ck, Surf. Sci., 2000, 47 H. Koschel, G. Held and H. P. Steinru 453, 201–213. 48 X. Xu and D. W. Goodman, J. Phys. Chem., 1993, 97, 683–689. 49 F. P. Netzer and T. E. Madey, J. Chem. Phys., 1982, 76, 710–715. 50 E. Demirci, C. Carbogno, A. Groß and A. Winkler, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 085421. 51 Z. Yan and D. W. Goodman, Catal. Lett., 2012, 142, 517–520. 52 M. Trenary, K. J. Uram and J. T. Yates Jr., Surf. Sci., 1985, 157, 512–538. 53 C. H. F. Peden and D. W. Goodman, Ing. Eng. Chem. Fundam., 1986, 25, 58–62. 54 E. Garfunkel, M. R. Yu, S. Yang and X. H. Feng, J. Vac. Sci. Technol., A, 1989, 7, 1579–1584. 55 S. Yang, M. R. Yu, G. Meigs, X. H. Feng and E. Garfunkel, Surf. Sci., 1988, 205, L777–L785. ¨ck, Surf. 56 G. Held, W. Sklarek, M. Mayan and H. P. Steinru Sci., 1998, 402–404, 322–326. 57 J. W. He, W. K. Kuhn and D. W. Goodman, Mater. Res. Soc. Symp. Proc., 1991, 202, 325–331. 58 F. M. Hoffmann and L. Pan, J. Chem. Phys., 1987, 87, 1857–1865. 59 F. M. Hoffmann and J. Paul, J. Chem. Phys., 1987, 86, 2990–2996. 60 X. Xu, S. M. Vesecky and D. W. Goodman, Science, 1992, 258, 788–790. 61 K. Yao, X. Liu, Z. Li, C. Li, H. Zeng and Y. Han, ChemCatChem, 2012, 4, 1938–1942. 62 X. Liu, K. Yao, C. Meng and Y. Han, Dalton Trans., 2012, 41, 1289–1296. 63 X. Liu, C. Meng and Y. Han, Nanoscale, 2012, 4, 2288–2295. 64 Y. X. Yao, X. Liu, Q. Fu, W. X. Li, D. L. Tan and X. H. Bao, ChemPhysChem, 2008, 9, 975–979. 65 Q. Fu, W. X. Li, Y. X. Yao, H. Liu, H. Su, D. Ma, X. Gu, L. Chen, Z. Wang, B. Wang and X. H. Bao, Science, 2010, 328, 1141–1144.
Phys. Chem. Chem. Phys., 2014, 16, 3823--3829 | 3829
