Adsorption of the formate species on copper surfaces: a DFT study

doi:10.1016/S0039-6028(99)00605-6

Surface Science

Volume 432, Issue 3, 20 July 1999, Pages 279-290

https://doi.org/10.1016/S0039-6028(99)00605-6 Get rights and content

Abstract

The density functional theory and the cluster model approach have been applied to study the interaction of the formate species with the copper (100), (110) and (111) surfaces. The short-bridge, long-bridge and cross-bridge sites of the copper surface have been modelled by Cu₇ and Cu₈ metal clusters after checking the validity of the results against those obtained with much larger clusters. The results show that the formate species is stabilized strongly on the short-bridge site of the surfaces considered and this is in agreement with available experimental data. For adsorption on the short-bridge site, and for the three surfaces considered, the Cu_surfO is close to 2.02 Å, the CO bond length is 1.26 Å and the OCO angle is 128°. On this adsorption site, formate is bridge-bonded with the two oxygens almost in a top position on two copper atoms. For the long-bridge site, the oxygen atoms of the adsorbate are not located above the two copper atoms. The two OC bonds are equivalent when formate is adsorbed on these two sites. On the cross-bridge site, the formate species is bonded to the surface in a monodentate conformation. The two OC bonds are different with two clearly different bond lengths. The difference is larger for adsorption on the Cu (100) surface. In this case, the bond lengths are typical of a bond order of one and two. The bonding to the surface is essentially ionic, and the total charge in the adsorbate is 0.75 ±0.05 e, approximately.

Introduction

The decomposition of formic acid on metal surfaces has been the subject of numerous experimental studies in recent years [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], due to the simplicity of the intermediates and products involved.

Formate is observed during the methanol oxidation reaction, and it is thought to be an intermediate in the methanol synthesis reaction. The most extensively studied surfaces are those directly related to the catalysis of these two reactions, i.e. copper and silver. In industry, a silver catalyst is used to obtain formaldehyde from methanol and a Cu/ZnO catalyst to produce methanol from syn-gas.

Surface formate is easily formed from formic acid by OH bond cleavage. It can undergo dehydrogenation to produce H₂ and CO₂ and dehydration to yield H₂O and CO, depending on the metal surface. On copper surfaces, only dehydrogenation is observed [1], [8]. Formate has been studied by different experimental techniques on copper (100) [1], [2], [3], [4], [5], [6], [7], copper (110) [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], copper (111) [23] and supported copper surfaces [23], [24], [25], [26], [27]. HCO₂ has been identified by temperature-programmed reaction spectroscopy (TPRS) [8], [9], by electron energy loss spectroscopy (EELS) [1], high-resolution electron energy loss spectroscopy (HREELS) [1], [8], reflection absorption infra-red spectroscopy (RAIRS) [10], in-situ infra-red reflection absorption spectroscopy (IRAS) [10], [11], [23] and scanning tunnelling microscopy (STM) [12], [13], [14]. From the results obtained with these techniques, it was difficult to clarify how the formate species are chemisorbed on copper surfaces, if formate were bonded to the copper surfaces in a upright or tilted geometry or if it were bonded in a bidentate or monodentate geometry. This is due to the observation, or not, of the υ_as (OCO) band. More studies were performed using near-edge X-ray absorption spectroscopy (NEXAFS) [3], [4], [15], [16], and it was concluded that the formate species was oriented with its molecular plane normal to the metal surface. By means of surface extended X-ray absorption fine structure (SEXAFS) [3], [4], it was found that the HCO₂ species was adsorbed on the Cu (100) surface on the cross-bridge site and later, in a data reanalysis [5], [6], that the diagonal atop site was the preferred adsorption site for formate adsorption. On the Cu (110) surface, the aligned atop site [5], [6], [15], [16] seemed to be the most favourable for HCO₂ adsorption.

More recently and taking advantage of photoelectron diffraction (PhD) [7] and NEXAFS [17] techniques, it was found that the formate species are adsorbed on the short-bridge sites of the Cu (100) and on Cu (110) surfaces. The oxygen atoms are located at the same distance from the surface and with a CuO nearest-neighbour distance of 1.98±0.04 Å.

For the Cu (111) surface, as far as we know, a similar PhD study does not exist. In a recent study [23] using IRAS for HCO₂/Cu (111), a bidentate structure was found with both oxygen atoms located at the same distance from the surface, thus possibly indicating adsorption of formate on the short-bridge site of the Cu (111) surface.

For low temperatures, it seems that this type of adsorption is preferred.

In the literature, only a few theoretical studies can be found on the adsorption of HCO₂ on copper surfaces [28], [29], [30], [31], [32], [33]. These theoretical studies are 10 years old and use very poor methods (by today's standards) to evaluate chemisorption properties, with the exception of the recent studies of Casarin et al. [32] and Kakumoto et al. [33]. Casarin et al. [32] use a linear combination of atomic orbitals method within the local density approximation (LCAO–LDA) to study the chemisorption of formate on Cu (100). They studied the adsorption on the short-bridge and on the cross-bridge sites and concluded that the adsorption on the short-bridge site is the most favourable. Ab-initio MO calculations using the DFT method were presented by Kakumoto et al. [33] for the formate intermediate adsorbed on a [Cu₂]⁺ bridge site. In this work, only the energy and geometry variation of the HCO₂ species with the CuCu distance is calculated.

One of the goals of this paper is to compile theoretical results for the adsorption of the formate species on the Cu (100), (110) and (111) surfaces under the same conditions in terms of methodology and basis sets in order to compare the trends in the adsorption of this species.

This paper is organized as follows. In Section 2, the theoretical details are given for the computational method used and for the metal clusters used to describe the metal surface. In Section 3, the calculated results for HCO₂ adsorption on Cu (100), (110) and on Cu (111) surfaces are presented and in Section 4, a summary of the main conclusions obtained with this work is given.

Section snippets

Method

In the present work, the interaction of the formate species with copper (100), (110) and (111) surfaces is studied. For that purpose, the density functional theory (DFT) approach was used. The Cu (100) and the Cu (110) metal surfaces are modelled by a Cu₈ cluster and the Cu (111) surface by a Cu₇ cluster as shown in Fig. 1. These two-layer clusters are a small section of the ideal Cu (100), Cu (110) and Cu (111) surfaces with a CuCu nearest-neighbour distance taken from the bulk and equal to

Free HCO₂ and HCO⁻₂ species

Table 1 lists the computed geometry, energy and vibrational frequencies for the three different σ states of the HCOO radical, ²B₂, ²A₁ and ²A′ optimized at the DFT/6-31G^∗∗ level. The calculations were made using as a starting point the geometry values obtained in previous work [45], [46], [47].

Depending on whether the oxygen atoms in HCO₂ are identical or not, there are two possible symmetric types for the HCO₂ species, C_2v and C_s. The ²B₂ and the ²A₁ states belong to the C_2v symmetric type,

Conclusions

The ab-initio cluster model approach has been applied to study the adsorption of the HCO₂ species on the copper (100), (110) and (111) surfaces. Small metal clusters of seven and eight metal atoms have been chosen to model the short-bridge, long-bridge and cross-bridge sites in the three metallic surfaces as these have been shown to be trustworthy for studies of this type. In order to justify the use of these small clusters in this work, we have also studied HCO₂ adsorption on the short-bridge

Acknowledgements

Financial support from the Fundação para a Ciência e Tecnologia (Lisbon) and project PRAXIS/PCEX/C/QUI/61/96 is acknowledged. J.R.B.G. thanks PRAXIS for a doctoral scholarship (BD/5522/95). We thank Professor Francesc Illas for helpful discussions.

References (55)

B.A Sexton
Surf. Sci.
(1979)
L.H Dubois et al.
Surf. Sci.
(1986)
D Outka et al.
Surf. Sci.
(1985)
M.D Crapper et al.
Surf. Sci.
(1987)
D.P Woodruff et al.
Surf. Sci.
(1988)
D.H.S Ying et al.
J. Catal.
(1980)
M Bowker et al.
Surf. Sci.
(1981)
B.E Hayden et al.
Surf. Sci.
(1983)
M Bowker et al.
Surf. Sci.
(1996)
A.H Jones et al.
Surf. Sci.
(1997)

A Ignaczak et al.

Chem. Phys. Lett.

(1996)

A Ignaczak et al.

J. Electroanal. Chem.

(1997)

Cited by (55)

Understanding of Au-CeO<inf>2</inf> interface and its role in catalytic oxidation of formaldehyde
2020, Applied Catalysis B: Environmental
To design an efficient catalyst for HCHO oxidation, it requires good understanding of the correlation between structural properties of catalyst and the reaction mechanism. We investigated performance of commercial CeO₂ (CeO₂-C) and nanorod CeO₂ (CeO₂-R) and the corresponding Au/CeO₂ catalysts in HCHO oxidation. Reactivity tests showed that Au/CeO₂-R is highly efficient and stable in HCHO oxidation at room temperature, which is attributed to the highly active Au-CeO₂-R interface. The strong interaction between Au and CeO₂-R greatly promotes the creation of oxygen vacancies and Au³⁺ near the interface, which promotes the formation of formate. The conversion of formate can follow both a decomposition and a direct oxidation pathway, which is remarkably enhanced by the coexistence of Au³⁺ and oxygen vacancies. In addition, formate accumulation and Au particles aggregation led to deactivation. The presence of humidity promotes the oxidation of formate, which improves the reactivity and stability of Au/CeO₂ in HCHO oxidation.
Adsorption of formate species on Cu(h,k,l) low index surfaces
2016, Surface Science
Citation Excerpt :
Later, Poulston et al. used INS to arrive at similar conclusions [11]. Several theoretical studies have been also undertaken to understand the electronic structure and the interaction between formate species and a metallic surface [12–14]. For these studies density functional theory (DFT) and semiempirical methods were employed on both periodic and cluster model approaches [15,16].
We report a density functional theory study on the relative stability of formate species on Cu(h,k,l) low index surfaces using a range of exchange-correlation functionals. We find that these functionals predict similar geometries for the formate molecule adsorbed on the Cu surface. A comparison of the calculated vibrational transition energies of a perpendicular configuration of formate on Cu surface shows an excellent agreement with the experimental spectrum obtained from inelastic neutron spectroscopy. From the calculations on adsorption energy we find that formate is most stable on the Cu(110) surface as compared to Cu(111) and Cu(100) surfaces. Bader analysis shows that this feature could be related to the higher charge transfer from the Cu(110) surface and optimum charge density at the interfacial region due to bidirectional electron transfer between the formate and the Cu surface. Analysis of the partial density of states finds that in the –5.5 eV to –4.0 eV region, hybridization between O p and the non-axial Cu d_yz and d_xz orbitals takes place on the Cu(110) surface, which is energetically more favourable than on the other surfaces.
Adsorption and thermal chemistry of formic acid on clean and oxygen-predosed Cu(110) single-crystal surfaces revisited
2016, Surface Science
The thermal chemistry of formic acid on clean and oxygen-predosed Cu(110) single-crystal surfaces was studied under ultrahigh-vacuum (UHV) conditions by temperature programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS). Key results reported in the past were confirmed, including the partial switchover from H₂ to H₂O desorption upon oxygen addition on the surface and the development of a second decomposition regime at 420 K, in addition to the one observed at 460 K on the clean substrate. In addition, new observations were added, including the previously missed desorption of H₂ at 420 K and the existence of a normal kinetic isotope effect in both TPD peaks. Peak fitting of the XPS data afforded the identification of an asymmetric geometry for the formate intermediate, which was established to form by 200 K, and the presence of coadsorbed molecular formic acid up to the temperatures of decomposition, probably in a second layer and held by hydrogen bonding. Quantitative analysis of the TPD data indicated a one-to-one correspondence between the increase in oxygen coverage beyond θ_O = 0.5 ML and a decrease in formic acid uptake that mainly manifests itself in a decrease in the decomposition seen in the 460 K TPD peak. All these observations were interpreted in terms of a simple decomposition mechanism involving hydrogen abstraction from adsorbed formate species, possibly aided by coadsorbed oxygen, and a change in reaction activation energy as a function of the structure of the oxygen overlayer, which reverts from a O-c(6 × 2) structure at high oxygen coverages to the O-(2 × 1) order seen at θ_O = 0.5 ML.
Generalized Brønsted-Evans-Polanyi relationships and descriptors for O-H bond cleavage of organic molecules on transition metal surfaces
2014, Journal of Catalysis
Citation Excerpt :
In particular, the interaction of formic acid with solid catalysts has been widely studied both from experimental [44–47] and theoretical [48–52] points of view. Often, formic acid dehydrogenates to formate (HCOO), which is a quite stable intermediate [2,48,53]. In the recent comparative DFT study of Luo et al. [48], it was found that formic acid decomposes to formate species on Ni(1 1 1) [48], with an activation energy barrier of about 0.4 eV, which is smaller than barriers calculated for the reaction on Pd(1 1 1), Pt(1 1 1) or MgO(0 0 1) surfaces.
Periodic density functional theory (DFT) based calculations were used to explore the relationship between the activation energy corresponding to RO–H bond cleavage of organic compounds on catalytically active transition metal surfaces and other simpler quantities which can be used as descriptors. Taking data for methanol on various surfaces, several Brønsted–Evans–Polanyi (BEP)-like relationships linking the activation energy barrier to the reaction energy, the adsorption energy of the reaction products or to the adsorption energy of an oxygen atom were explored. The general validity of these relationships has been explored by considering cases not included in the database used to extract the BEP relationships. For the more promising BEP relationship, the database for methanol was combined with results corresponding to O–H bond breaking of ethanol, formic acid and water on a sufficiently broad number of transition metal surfaces. This extended database provided a more general and statistically meaningful general BEP type relationship connecting the activation energy for the O–H bond breakage of general RO–H compounds on catalytic transition metal systems to the adsorption energy of the reaction products. Finally, a protocol is presented that allows one to determine good candidates for bond breakage of general RO–H compounds on metallic and bimetallic surfaces limiting the explicit calculation of the activation energy barriers to a few, previously detected, interesting cases only.
Insights into the reaction mechanisms of methanol decomposition, methanol oxidation and steam reforming of methanol on Cu(111): A density functional theory study
2014, International Journal of Hydrogen Energy
Cu-based catalysts have been widely used for hydrogen production from methanol decomposition, methanol oxidation and steam reforming of methanol (MSR). In this study, we have systematically identified possible reaction paths for the thermodynamics and dynamics involved in the three reactions on a Cu(111) surface at the molecular level. We find that the reaction paths of the three reactions are the same at the beginning, where methanol scission is favourable involving O–H bond scission followed by sequential dehydrogenation to formaldehyde. Formaldehyde is an important intermediate in the three reactions, where direct dehydrogenation of formaldehyde to CO is favourable for methanol decomposition; for methanol oxidation, formaldehyde tends to react with oxygen to form dioxymethylene through C–H bond breaking and finally the end products are mainly CO₂ and hydrogen; for MSR, formaldehyde tends to react with hydroxyl to form hydroxymethoxy through formic acid and formate formation, followed by dissociation to CO₂. CH₂O formation from methoxy dehydrogenation is considered to be the rate-limiting step for the three reactions. In general, the thermodynamic and kinetic preference of the three reactions shows the order methanol oxidation > MSR > methanol decomposition. Methanol oxidation and MSR are faster than methanol decomposition by about 500 and 85 times at typical catalytic conditions (e.g., 523 K), respectively. The result may be useful for computational design and optimization of Cu-based catalysts.
Catalytic deoxygenation chemistry: Upgrading of liquids derived from biomass processing
2013, Advances in Catalysis
The composition of a typical pyrolysis oil is used to pose the problem of oxygen removal from this oil and to motivate a discussion of the different reactions for oxygen removal that are thermodynamically favored. From a consideration of this thermodynamics analysis of the favored reactions, the surface chemistry literature is surveyed to reveal those materials that allow the adsorption of oxygen-containing species. Included in this survey are experimental and theoretical studies. This consideration of adsorption of oxygen-containing species prompts a further examination of the literature of reaction mechanisms and reaction sequences to reveal the conventional pathways to remove oxygen from the substrates. Moreover, the literature of hydroprocessing is reviewed to show how traditional hydrodesulfurization catalysts have been studied as a hydrodeoxygenation (HDO) catalyst of the reactive species in pyrolysis oils. Furthermore, aqueous-phase reforming is discussed as an alternative to HDO.

View all citing articles on Scopus

View full text

Adsorption of the formate species on copper surfaces: a DFT study

Abstract

Introduction

Section snippets

Method

Free HCO2 and HCO−2 species

Conclusions

Acknowledgements

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

J. Catal.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Surf. Sci.

Catal. Today

J. Mol. Struct. Theochem.

Chem. Phys. Lett.

J. Electroanal. Chem.

Free HCO₂ and HCO⁻₂ species