First-principle density functional theory (DFT) and a periodic-slab model have been utilized to investigate the structure of the CuCl2(100) surface. Structural parameters of the bulk CuCl2 are reported and compared with the experimental values. The structure of the CuCl2(100) is calculated using a (2×2) supercell. Structural parameters in terms of bond lengths and bond angle are calculated. Electronic properties of the CuCl2(100) surface are investigated by calculating the density of state (DOS) and the projected density of state for a slab containing five layers.
1. Introduction
Copper plays a prominent role in many fields. This is primarily due to its unique mechanical properties and excellent electrical and thermal conductivity. Its applications spans heating and cooling piping, heat exchanging, electrical appliances, architecture, and catalysis [1–6].
Adsorption of halides on metallic surfaces is of particular importance due to their relevance for many industrial processes and its role in the formation of notorious environmental pollutants such as dioxins [7]. Central to a comprehensive understanding of various phenomena such as corrosion, electrodeposition, and catalytic cycles is addressing adsorption of chlorine on copper surfaces on a precise atomic-scale.
Motivated by its diverse application, the chlorine-copper interaction has attracted great deal of experimental [8–14] and theoretical [15–18] studies. Mounting experimental evidence shows a strong interaction between Cl and Cu atoms as evident from electron diffraction (LEED) [8], normal-incident X-ray standing wave (NIXSW) [9], surface-extended X-ray adsorption fine structure (SEXAFS) [10, 11], shadow-cone-enhanced secondary-ion mass spectrometry (SIMS) [6, 12], and Scanning Tunnelling microscopy (STM) [13, 14] methods. STM studies found that the chlorine molecules adsorb dissociatively on the Cu(111) surface, leading to formation of various well-ordered phases. It is also found that Cl atoms prefer to adsorbs onto face-centered cubic (fcc) hollow positions over the hexagonal close-pack (hcp) hollow, bridge, and top positions [13, 18].
Computational data for the Cl/Cu(111) system based on cluster [15] or a periodic DFT-based calculation [16, 17, 19] has been presented by different groups Doll and Harrison [16], Migani et al. [17], and Peljhan and Kokalj [19]. It was found that the preferred adsorption of the Cl atoms is the face-centered cubic (fcc) [16]. Peljhan et al. presented an extensive DFT calculation of Cl adsorption on Cu(111), considering a wide range of Cl coverages ranging from 1/16 to 1 ML. The strong interaction between Cl and Cu atoms results in the formation of two distinctive different materials; namely, CuCl and CuCl2. The crystal structure of the CuCl2 follows a base-centered monoclinic space group C2/m [20–22]. As demonstrated by Burns and Hawthrone [23]. Cu atoms are positioned in an axially distorted octahedral environment with four equatorial nearest neighbors Cl atoms and two axial Cu atoms.
First-principle simulation, based on density functional theory (DFT), is now one of the key tools for studying and developing a key understanding of solid surfaces. DFT calculations can provide an atomic-based insight into the detailed bonding arrangement of structures, their energetic stability, and charge densities. To this end, this contribution investigates the electronic structure of the Copper (II) chloride, CuCl2, mainly in the (100) surface orientation.
2. Methodology
The calculations reported in this paper have been performed using the plane wave pseudopotential density functional theory (DFT) method as implemented in the VASP program [24–27]. The generalized gradient approximation (GGA) for exchange and correlation as developed by Perdew and Wang (PW91) [28] was used to perform the spin-polarized calculations. Projector augmented wave (PAW) potentials [24, 25, 29] are used to represent the ionic potentials. A unit cell comprising four atomic layers has been employed together with a vacuum thickness of at least 14 Å to separate each slab from its neighboring images along the z-direction (normal to the surface). The three top-most layers of the slab were allowed to fully relax. The Brillouin-Zone (BZ) integrations were performed using automatic generation of (3 × 3 × 1) and (5 × 5 × 1) k-point set of Monkhorst-Pack random sampling [30]. Total energy was converged to an accuracy of 1 × 10−3 eV. Deployed energy cut-off was set at 300 eV. The choice of this value, stems from the maximum value of the kinetic-energy for the Cl and Cu atoms, that is, 280 and 273 eV, respectively.
3. Results3.1. Bulk Properties
CuCl2 forms a base-centered Monoclinic Bravies lattice. As shown in Figure 1, each Cu atom is in an axially distorted octahedral environment with four equatorial nearest neighbors Cl atoms and two axial Cl atoms. The lattice constants were determined (as illustrated in Table 1) by plotting the variation in the energy with respect to the volume of the unit cell of bulk CuCl2. The lattice constants of bulk CuCl2 are compared with calculated data in reference [19] and experiment data in references [23, 31] in Table 1. This comparison shows that our results are in a relative good agreement with those of earlier experiment and computational studies.
Comparison of our calculated properties of the bulk CuCl2 bulk with previous calculated data [19] and experiment data [23, 31].
Reference
Lattice constant
a (Å)
b (Å)
c (Å)
β (deg.)
This work
7.35
3.40
7.20
117
Previous calculation (ref. [19])
7.52
3.35
7.29
119
Experiment (ref. [23, 31])
6.90
3.30
6.82
122
Unit cell of CuCl2. In online version, green circle represent Cl atoms while red and violet circles represent Cu atom.
The calculated total density of states (DOS) and atomic-projected DOS (pDOS) for bulk CuCl2 are presented in Figure 2; Figure 2(a) shows the Density of state (DOS) of the CuCl2 bulk, while Figure 2(b) shows the projected density of state for Cu atom and Figure 2(c) shows the projected density of state for Cl atom.
Density of state (DOS) of the bulk of the CuCl2; (a) total DOS, projected DOS (PDOS) of (b) Cu atom and (c) Cl atom.
There are two distinct bands in the valence region (Figure 2(a)). The Cu band (predominantly Cu 3d and little contributions of Cu (3p and 4s)) occupies the region (0 to −5) eV, and the Cl band (predominantly Cl (3s and 3p)) occupies that of (−14 to −15) eV. The surface state band resides just above the Fermi level where two regions are highlighted. The first one is in the region from 3.0 to 7.0 eV, while the second one is from 7.5 to 9.0 eV. The first region above the Fermi level is predominantly the empty state Cu 4s and the contribution in this region from Cu 3p and Cl (3s and 3p) is minimal (see Figures 2(b) and 2(c)). The second region above the Fermi level is predominantly the state Cu (3d and 3p) and Cl (3s and 3p).
3.2. CuCl2(100) Surface
A supercell of the CuCl2(100) is constructed from the optimized bulk unitcell and also based on the data in [21, 22, 32]; the unit cell is shown in Figure 3. This unit cell is a base-centered Monoclinic with one Cu atom located in the center of each base and four Cl atoms in the corner of each base. The Cu atom has four nearest neighbors Cl atoms in the same plane (plane of a and b) and two Cl atoms in c direction.
the unit-cell of the CuCl2(100) surface: it is distorted in the (c) direction, which represents the height. In online version, green circles represent Cl atoms, while red circles represent Cu atom.
Test is performed using automatic generation of (3 × 3 × 1) and (5 × 5 × 1) k-point set of Monkhorst and Pack. For each optimization using these two schemes, the total energy was converged to an accuracy of 1 × 10−3 eV and the forces on each ion to an accuracy of 0.015 eV Å−1. A test using (5 × 5 × 1) k-point changed the total energies of the (3 × 3 × 1) k-point by a only few meV. All the calculations discussed here are using a 3 × 3 × 1 k-point grid (unless otherwise specified).
Transformation upon building the surface can be viewed in terms of the relaxation, that is, the change in interatomic distances and angle B between optimized bulk unit cell and the extended surface. We found that Cu-Cl distances become longer in the extended surface than the bulk one by only 0.009 Å. While the angle B on the unit cell unit cell varies between 117° in the bulk and 111.3° in the surface (as shown in Figure 3). It is worthwhile also mentioning that optimized surface retains the symmetry exhibited by the unit cell. A noticeable difference in c and a lattice constants could be rationalized based on the shortcoming of standard DFT methods in descrying states relating to long-term weak interactions (Van der Walls type) [33, 34].
The surface has a space group C2/m (as shown in Figure 3) the same as that of the CuCl2 bulk. Unit cells are shifted from each other by 4.06 Å. The super-cell of the surface is illustrated in Figure 4; the top and side views are also shown in Figure 5. All bond lengths and bond angles are illustrated in Table 2 based on Figure 5(b). The distance between each two atomic layer is 3.212 Å. The surface has been optimized also by the (5 × 5 × 5) k-point set, and some of the bond lengths and bond angles are illustrated also in Table 2 based on Figure 5(b). As can be noticed from Table 2, some of the bond length shrinks slightly by 0.02 Å while the others stay the same. Also some of the bond angle did not change, while the others changed by only 0.2°. Angle B in the unit cell illustrated in Figure 3 becomes smaller also by 0.8°.
Calculated structural parameters of the CuCl2(100) surface using (3 × 3 × 1) and (5 × 5 × 1) k-point set of Monkhorst and Pack, angle B is the one in the unit cell as in Figure 3.
Bond length (Å)
(3 × 3 × 1)
(5 × 5 × 5)
Bond angle (°)
(3 × 3 × 1)
(5 × 5 × 5)
Cu1–Cl1
2.287
2.275
Cu1–Cl1–Cu2
94
94
Cl1–Cu2
2.280
2.273
Cl1–Cu2–Cl2
93.8
94
Cu1–Cl3
2.270
2.270
Cl1–Cu1–Cl3
86
86
Cl3–Cu2
2.270
2.275
Cu1–Cl3–Cu2
94
94
Cu1–Cu3
7.190
7.190
Cl1–Cu2–Cl3
86.2
86
Cl3–Cl5
4.090
4.090
B
111.3
110.5
Cu2–Cu4
7.190
7.190
—
—
—
Cl4–Cu6
4.080
4.080
—
—
—
(a) Side view of the (2 × 2) unit-cell of the CuCl2(100) surface and (b) side view of the (2 × 2) unit-cell of the CuCl2(100) surface but the atoms in Figure 4(b) are rotated with small angle to show the atoms in each layer. The (b) direction is perpendicular to both (a) and (c). The unit-cell is distorted in the (c) direction, which represents the height.
(a) Side view and (b) top view of the CuCl2(100) surface; Figure 5(a) shows part of the surface in Figure 5(b) but rotated to show the structure. Dimensions are in angstrom.
3.3. Electronics Properties
The electronics properties of the CuCl2(100) surface are investigated by calculating the density of state (DOS) for a slab containing five layers using the automatic generation of (3 × 3 × 1) k-point. Figure 6 shows the total DOS and projected DOS of the surface. The investigation of the DOS shows that there are three main regions in the valence band below the Fermi level (0 eV in Figure 6(a)). The first and second regions occupy the regions of (−2 to −6) eV and (−6 to −8) eV. These two regions are manly composed of Cu 3d (Figure 6(b)), with a little contribution of Cl 3p, but the Cu (4s and 3p) and Cl 3s states are minimal (Figures 6(b) and 6(c)).
Density of state (DOS) of the CuCl2(100) surface: (a) total dos, (b) projected DOS (PDOS) of the surface Cu atom, and (c) projected DOS (PDOS) of the surface Cl atom.
The last region in the valence band is starting at −17 eV; it is composed of Cl 3s states. The conduction band has two distinct regions starting at 1.5 eV with a wide of approximately 2.5 eV for each one. Each of these regions has small peaks at 3.2 eV and 5.2 eV, respectively. The first region is predominantly Cu (4s and 3p) and Cl (3s and 3p), while the second region is predominantly of Cu 4s and the Cl (3s and 3p).
4. Conclusions
The structural, electronic properties of bulk and (100) surface of CuCl2 were investigated by means of periodic quantum chemical calculation based on the first-principle DFT approach. For the bulk, the comparison of optimized structural parameters with experimental and previous DFT study shows that the lattice constants are in a good agreement. The supercell of the CuCl2(100) surface showed to be axially distorted in the z direction. The interlayer distance, all the bond lengths, and the bond angle are calculated.
Total and projected Density of State (DOS) for the bulk and the extended surface is calculated. For the (100) surface, the valence band is having two regions: the first one composed of Cu (3d) and Cl (3p), while the second region is composed of Cl (3s) states. The surface states above the Fermi level are composed of manly Cu (4s and 3p) and Cl (3s and 3p). An extension to this work is to study the adsorption and dissociation of small molecules such as H2 and H2O on this surface.
SubramanianK. S.Determination of tin in lead/tin solder leachates from copper piping by graphite platform furnace atomic-absorption spectrometry19893611107510802-s2.0-0037784306SmirnovV. V.RothJ. P.Mechanisms of electron transfer in catalysis by copper zinc superoxide dismutase20061285116424164252-s2.0-3384592710510.1021/ja066369rPaxtonW. F.SpruellJ. M.StoddartJ. F.Heterogeneous catalysis of a copper-coated atomic force microscopy tip for direct-write click chemistry200913119669266942-s2.0-6765052363510.1021/ja9015974SongJ.ZhaoT.DuY.Supported copper catalysts for direct vapor-phase oxycarbonylation of methanol20062753863902-s2.0-3375016059110.1016/S1872-2067(06)60023-5CenY.LiX.LiuH.Preparation of copper-based catalysts for methanol synthesis by acid-alkali-based alternate precipitation method20062732102162-s2.0-3374945107610.1016/S1872-2067(06)60014-4LooiM. H.LeeS. T.Abd-HamidS. B.Use of citric acid in synthesizing a highly dispersed copper catalyst for selective hydrogenolysis20082965665702-s2.0-4894911831210.1016/S1872-2067(08)60050-9AltarawnehM.DlugogorskiB. Z.KennedyE. M.MackieJ. C.Mechanisms for formation, chlorination, dechlorination and destruction of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs)20093532452742-s2.0-6474909526410.1016/j.pecs.2008.12.001GoddardP. J.LambertR. M.Adsorption-desorption properties and surface structural chemistry of chlorine on Cu(111) and Ag(111)1977671180194KadodwalaM. F.DavisA. A.ScraggG.CowieB. C. C.KerkarM.WoodruffD. P.JonesR. G.Structural determination of the Cu(111)-(3×3) R30°-Cl/Br surface using the normal incidence X-Ray standing wave method19953242-31221322-s2.0-0029253711CrapperM. D.RileyC. E.SweeneyP. J. J.McConvilleC. F.WoodruffD. P.Complete adsorption site information for Cl on Cu(111) using X-ray absorption fine structure and photoelectron diffraction198621185710.1209/0295-5075/2/11/007WoodruffD. P.SeymourD. L.McConvilleC. F.RileyC. E.CrapperM. D.PrinceN. P.JonesR. G.Simple x-ray standing-wave technique and its application to the investigation of the Cu(111) (√3 √3) R30°-Cl structure19875814146014622-s2.0-3374470700710.1103/PhysRevLett.58.1460WayW. K.PikeA. C.RosencranceS. W.BraunR. M.WinogradN.Coverage-dependent bond length of chlorine adsorbed on Cu(111)19962421371412-s2.0-0030085665SakuraiT.HashizumeT.FI-STM study of metal surfaces1992331261322-s2.0-002689112210.1088/0957-4484/3/3/005SuggsD. W.BardA. J.Scanning tunneling microscopic study with atomic resolution of the dissolution of Cu(111) in aqueous chloride solutions199411623107251073310.1021/ja00102a042IgnaczakA.GomesJ. A.Interaction of halide ions with copper: the DFT approach19962575-660961510.1016/0009-2614(96)00603-3DollK.HarrisonN. M.Chlorine adsorption on the Cu(111) surface20003173–528228910.1016/S0009-2614(99)01362-7MiganiA.SousaC.IllasF.Chemisorption of atomic chlorine on metal surfaces and the interpretation of the induced work function changes20055742-32973052-s2.0-1134425343710.1016/j.susc.2004.10.041MiganiA.IllasF.A systematic study of the structure and bonding of halogens on low-index transition metal surfaces20061102411894119062-s2.0-3374579278910.1021/jp060400uPeljhanS.KokaljA.Adsorption of Chlorine on Cu(111): A Density Functional Theory Study20091133214363143762-s2.0-6874909512710.1021/jp902273kBergasovaL. B.FilatovS. K.The new mineral tolbachite CuCl219832704154172-s2.0-0021063994GmelinB.1958211Weinheim, GermanyChemie GmbHWellsA. F.333. The crystal structure of anhydrous cupric chloride, and the stereochemistry of the cupric atom194711670167510.1039/jr9470001670BurnsP. C.HawthroneF. C.Tolbachite, CuCl2, the first example of Cu2+ octahedrally coordinated by Cl-199378187189KresseG.HafnerJ.Ab initio molecular dynamics for liquid metals199347155856110.1103/PhysRevB.47.558KresseG.HafnerJ.Ab initio molecular-dynamics simulation of the liquid-metal–amorphous-semiconductor transition in germanium19944920142511426910.1103/PhysRevB.49.14251KresseG.FurthmullerJ.Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set199661155010.1016/0927-0256(96)00008-0KresseG.FurthmullerJ.Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set19965416111691118610.1103/PhysRevB.54.11169PerdewJ. P.ChevaryJ. A.VoskoS. H.JacksonK. A.PedersonM. R.SinghD. J.FiolhaisC.Atoms, molecules, solids, and surfaces: applications of the generalized gradient approximation for exchange and correlation19924611667166872-s2.0-2324446083810.1103/PhysRevB.46.6671BlöchlP. E.Projector augmented-wave method19945024179531797910.1103/PhysRevB.50.17953MonkhorstH. J.PackJ. D.Special points for Brillouin-zone integrations19761312518851922-s2.0-184281690710.1103/PhysRevB.13.5188LideD. R.19937thBoca Raton, Fla, USACRC PressFindJ.HereinD.UchidaY.SchlöglR.New three-dimensional structural model for the CuCl2 graphite intercalation compound1999379143114412-s2.0-003262414310.1016/S0008-6223(99)00005-6AltarawnehM.RadnyM. W.SmithP. V.MackieJ. C.KennedyE. M.DlugogorskiB. Z.2-Chlorophenol adsorption on Cu(1 0 0): first-principles density functional study20086028155415622-s2.0-4174912341110.1016/j.susc.2008.02.029LilienfeldO. A.TavernelliI.RothlisbergerU.SebastianiD.Optimization of effective atom centered potentials for London dispersion forces in density functional theory200493151530042-s2.0-1964440083210.1103/PhysRevLett.93.153004