Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2018 Mar 12;9(7):1803–1808. doi: 10.1021/acs.jpclett.7b03182

Testing Electronic Friction Models: Vibrational De-excitation in Scattering of H2 and D2 from Cu(111)

Paul Spiering 1, Jörg Meyer 1,*
PMCID: PMC5890313  PMID: 29528648

Abstract

graphic file with name jz-2017-03182b_0005.jpg

At present, molecular dynamics with electronic friction (MDEF) is the workhorse model to go beyond the Born–Oppenheimer approximation in modeling dynamics of molecules at metal surfaces. Concomitant friction coefficients can be calculated with either the local density friction approximation (LDFA) or orbital-dependent friction (ODF), which, unlike LDFA, accounts for anisotropy while relying on other approximations. Due to the computational cost of ODF, extensive high-dimensional MDEF trajectory calculations of experimentally measurable observables have hitherto only been performed based on LDFA. We overcome this limitation with a continuous neural-network-based representation. In our first application to the scattering of vibrationally excited H2 and D2 from Cu(111), we predict up to three times higher vibrational de-excitation probabilities with ODF than with LDFA. These results indicate that anisotropic electronic friction can be important for specific molecular observables. Future experiments can test for this “fingerprint” of different approximations underlying state-of-the-art MDEF.

The motion of atomic and molecular adsorbates on metal surfaces underlies every elementary reaction step in heterogeneous catalysis. Due to the absence of an energy gap between valence and conduction band electrons, these motions can result in the excitation of electron–hole pairs (EHPs) and thus violate the Born–Oppenheimer approximation.1−3 A growing number of experiments points to the importance of this nonadiabatic energy loss channel.4 On the other hand, the development of suitable theoretical models to account for these nonadiabatic effects is still an ongoing process.5−9 For systems with weak nonadiabatic coupling, molecular dynamics with electronic friction (MDEF)10 is currently the most popular approach.11 MDEF relies on a potential energy surface (PES), nowadays typically obtained from density functional theory (DFT),3 and accounts for the effects of the EHPs on the motion of the nuclei by electronic friction coefficients.10 One state-of-the-art technique for calculating these coefficients as functions of the adsorbate positions relies on mapping to an atom-in-jellium model for which only the surface electron density is considered (local density friction approximation, LDFA12,13). Alternatively, the electronic states of the molecule–surface system can be taken into account (orbital-dependent friction, ODF10,14). For the inelastic scattering of H atoms from Au(111), millions of MDEF trajectories based on a high-dimensional PES15 and LDFA have recently been demonstrated to yield accurate scattering probabilities in excellent agreement with experimental data.16

The situation is quite different for molecules. Due to its combination with the independent atom approximation, the LDFA completely neglects any molecular effects.12 ODF on the other hand accounts for the anisotropic tensorial character of friction coefficients on corrugated metal surfaces and along adsorbate-internal bonds,17−19 which is why ODF has been argued to be “theoretically” more accurate for (diatomic) molecules.20 However, this discussion12,20,21 has still remained inconclusive because an evaluation of ODF comes at very high computational costs. Consequently, extensive MDEF trajectory calculations for molecules including all relevant degrees of freedom (DOF) can be easily performed with LDFA,11,12 whereas only two molecular DOF have so far been included for ODF.22 The very recent on-the-fly evaluation of ODF within ab initio molecular dynamics by Maurer et al.19 is an important step, but the less than 20 calculated trajectories make direct validation via molecular beam experiments impossible. Modeling the nonadiabatic contribution to vibrational lifetimes of molecules adsorbed on metal surfaces on the other hand does not require such extensive statistical averaging.23,24 The most recent implementations of LDFA and ODF both yield results that agree with experimental data within the error bars.25,26 Furthermore, Novko et al. have shown recently in this context27 that the numerical evaluation of friction tensors within ODF19,26 effectively includes potentially spurious electronic memory effects with unclear consequences for MDEF.28 Given this situation, theoretical understanding and modeling relying on MDEF faces an important question: Is the molecular anisotropy as described by ODF important for any observables that can be validated by high-precision molecular beam experiments like for atoms?16

In this work, we provide an answer to this question using H2 and D2 on Cu(111). For this system, weak nonadiabatic coupling and static surface approximation are well justified,29−31 and an accurate DFT-based PES relying on the semiempirically constructed “SRP” exchange–correlation functional is available.32,33 We construct a six-dimensional neural-network-based continuous representation of ODF that allows us to perform extensive MDEF trajectory calculations on equal footing with LDFA. While dissociative sticking probabilities are hardly affected in general and by the type of electronic friction coefficients used, we find vibrational de-excitation probabilities to be a “fingerprint” that can be used to distinguish and validate LDFA and ODF in future experiments.

Quasi-classical trajectory calculations32 within MDEF rely on a generalized Langevin equation10

graphic file with name jz-2017-03182b_m001.jpg 1

where i,j indicate atoms and α,β Cartesian coordinates. Atomic masses and positions are denoted by mi and Ri, respectively. For a H2 or D2 molecule on a static surface, the total number of moving atoms N is two, resulting in six DOF. In addition to the forces from the PES Inline graphic, which yield the adiabatic dynamics, nonadiabatic effects on the nuclear dynamics originate from electronic friction forces Fiαfric(R) and thermal white noise Inline graphic, respectively. In this work, V(R) is mainly taken to be the static surface PES based on the SRP48 exchange–correlation functional from ref (33), but we also compare with the PW91-PES from earlier work.34 The friction forces are linear in nuclear velocities Ṙjβ and are in general given by a symmetric 6 × 6 friction tensor ηiαjβ(R), which consists of 21 independent elements each depending on six nuclear coordinates. These coordinates can be Cartesian R = (R1,R2) or expressed in the center-of-mass centered spherical coordinate system R = (X,Y,Zd,θ,ϕ), which is commonly used for diatomics and described by Figure 1A.

Figure 1.

Figure 1

Open in a new tab

(A) Molecular coordinate system denoting the center of mass positions (X,Y,Z), bond length d, as well as spherical orientation (θ,ϕ). (B) Top view of a reference configuration with Inline graphic, Y = 0, and θ0 = ϕ0 = 90° from the minimum-energy reaction path for dissociative chemisorption over the bridge site,34 where a denotes the surface lattice constant. Cu atoms in the first, second, and third layer are depicted by increasing transparence. Note that X,Y,Z = 0 corresponds to the position of a Cu atom in the surface plane (top site).

Within ODF, these 21 friction coefficients are obtained according to a Fermi golden rule-like expression resulting from time-dependent perturbation theory, which can be written in the quasi-static limit as10,24,26,35

graphic file with name jz-2017-03182b_m004.jpg 2

The electron–phonon matrix elements gkabiα describe the nonadiabatic coupling between two electronic states of the molecule at the metal surface with band indices a and b at wave vector k due to the motion of (adsorbate) atom i along direction α. In general, the ODF tensor can have different diagonal elements even for the same atom. This anisotropy yields very different friction forces when the atoms move (with the same velocity) in different directions. Its generally nonzero off-diagonal elements are responsible for coupling the motion in different directions and between both atoms in a way that is not accounted for by the PES. In particular, this can lead to a strong damping of the molecular stretch vibration of a diatomic molecule and thus a pronounced molecular anisotropy.19,20,22

In order to use the so-calculated ηiαjβODF(R) in MDEF trajectory calculations of generic experimentally measurable observables, a continuous representation of this 6 × 6 tensor is required that can be evaluated at low computational cost. We have designed such a representation based on a symmetry-adapted neural network fit that is briefly described in the Supporting Information and will be discussed extensively in a forthcoming publication.

Within LDFA, friction coefficients for hydrogen atoms ηH(ρ) are obtained from a spherical atom-in-jellium model with background density ρ, which is solved via DFT at the level of the local density36 or generalized gradient approximation.37 Mapping of the actual surface problem is accomplished by taking the electron density of the bare surface (without the molecule) at each atom’s position ρ(Ri) as the background density of the jellium.12 This independent-atom approximation (IAA) results in electronic friction coefficients that are isotropic for each atom and depends on its own three coordinates alone. In Cartesian coordinates, only diagonal elements of the friction tensor are nonzero

graphic file with name jz-2017-03182b_m005.jpg 3

A continuous representation of ηiαjβLDFA(R) for extensive MDEF trajectory calculations can be easily constructed.12,38

Going beyond the IAA within LDFA is possible, for example, by determining the background electron density using an atoms-in-molecules technique (LDFA-AIM).25 However, this approach does not lift the isotropy and, as detailed in the Supporting Information, cannot be applied to H2 and D2 molecules. The other way round, isotropic friction can be constructed from ODF by neglecting the coupling between different directions and atoms plus averaging the remaining (generally anisotropic) friction over different directions

graphic file with name jz-2017-03182b_m006.jpg 4

This ODF-iso allows one to disentangle the influence of anisotropy from the very different electronic structure inherent to ODF and LDFA.

Figure 2A–C shows ηZd, ηdd, and ηZZ, respectively, as obtained from eqs 2–4 along the minimum-energy reaction path for dissociative chemisorption over the bridge site, as depicted in Figures 2D and 1B. We focus here on these three particular friction coefficients in order to compare with the earlier two-dimensional ODF calculations.20,22 The agreement is quite good except for some differences close to the transition state for ηZZODF. As the molecule approaches the surface, each model yields increasing friction for the six diagonal elements of the friction tensor, and the absolute values of the off-diagonal elements increase likewise in the case of ODF. Furthermore, ODF directly reflects the strong rearrangement of (Kohn–Sham) orbitals when approaching the dissociation barrier by significantly higher friction along the molecular bond (and thus reaction) coordinate, resulting in ηdd ≈ 3ηddLDFA at the transition state, in agreement with earlier work.20 For the observables calculated below, friction beyond the dissociation barrier is not relevant. Quite remarkably, ηdd (ηZZODF-iso) and ηdd (ηZZLDFA) are almost identical up to the transition state and thus much more alike than what has originally been found for the diffusion of H atoms on Pd(100).17,39

Figure 2.

Figure 2

Open in a new tab

(A–C) ηdZ, ηdd, and ηZZ in the molecular coordinate system (see Figure 1A), respectively, along the minimum-energy reaction path for dissociative chemisorption over the bridge site (see Figure 1B) as depicted in (D) together with the corresponding two-dimensional PES cut. The blue, red, and purple lines indicate the continuous representation from this work for ODF, LDFA, and ODF-iso, respectively, as obtained from eqs 2–4. Blue (red) dots show the ODF (LDFA) results from previous work of Luntz et al.,20,22 and the reaction coordinate is defined in the same way as in that work. The barrier and thus the transition state for dissociation are located at the vertical gray line (i.e. 0 Å) in (A–C) and indicated by the empty circle in (D). Negative numbers up to the transition state denote the approach from the gas phase (i.e., decreasing heights Z above the surface).

In order to study the effect of the different friction models on actual experimental observables, we perform MDEF calculations according to the quasi-classical trajectory method.28 In view of the short interaction time of the molecules with the Cu(111) surface during all simulated trajectories, we neglect the fluctuating forces in eq 1.

Figure 3A,B shows the results for the dissociative chemisorption probability S0 for both H2 and D2 molecular beams based on the SRP48-PES, respectively. Due to the construction of the latter,32,33 already the adiabatic calculations yield good agreement with the experimental data.40,41 Inclusion of electronic friction reduces S0, leading to even better agreement with the experimental data, in particular, at high incidence energies. The reduction is strongest for ODF and weaker for LDFA and ODF-iso, which are very similar to each other. It can be rationalized by the differences of the friction models for the friction ηdd along the reaction coordinate close to the dissociation barrier (see Figure 2B). This effect of ηddODF on S0 for H2 and D2 on Cu(111) has not been reported for two-dimensional ODF calculations.22 Consequently, a high-dimensional treatment of ODF in MDEF, on an equal footing with LDFA,21 is important. However, the overall small effect of electronic friction on S0 makes this not an optimal observable for experimental validation of the different friction models.

Figure 3.

Figure 3

Open in a new tab

Calculated reaction probabilities S0 based on the SRP48-PES for dissociative chemisorption of (A) H2 and (B) D2 molecular beams as a function of average normal incidence energy ⟨Ei⟩ for the indicated nozzle temperatures Tnozzle in comparison to experimental data (brown unfilled squares) from refs (40) and (41), respectively. The calculations are adiabatic (filled black squares) or employ the LDFA (red plusses), ODF (blue circles), and ODF-iso (purple crosses) models for the electronic friction coefficients.

Instead, we have identified vibrational de-excitation probabilities Ptransition to yield a clearly distinguishable difference between LDFA and ODF. On the basis of 50000 MDEF trajectories, we calculate Ptransition as a function of incidence energy Ei from the scattered trajectories by a conventional binning procedure. The concomitant average gain in translational energy ⟨ΔEtrans⟩ is calculated from the final center-of-mass velocities. As detailed in the Supporting Information, the error bars reflect the error due to statistical sampling of the initial conditions. Only by employing our continuous representation to compute a large amount of trajectories were we able to reduce these errors so that the different electronic friction models can be distinguished. We focus the discussion on de-excitation from vibrational state ν = 2, J = 1 (2) to ν = 1, J = 1 (2) for H2 (D2), respectively, as shown in Figure 4A,C (B,D). Unlike for other vibrational transitions,42 for this transition, we obtain results that are not only qualitatively but even almost quantitatively identical to corresponding results obtained with the PW91-PES (see the Supporting Information).

Figure 4.

Figure 4

Open in a new tab

Vibrational de-excitation probabilities Ptransition (lower row) and concomitant average gain in translational energy ⟨ΔEtrans⟩ (upper row) as a function of normal incidence energy Ei for state-to-state scattering using the SRP48-PES. Panels (A,C) [(B,C)] are for the transition from the rovibrational state ν = 2, J = 1 [2] to ν = 1, J = 1 [2] for H2 [D2]. Shown are results from adiabatic calculations (filled black squares), as well as those including electronic friction according to the LDFA (red pluses), ODF (blue circles), or ODF-iso (purple crosses) models. The error bars indicate the error due to statistical sampling as described in detail in the Supporting Information.

At low incidence energies, with increasing Ei, more and more molecules come close enough to the surface so that the curvature of the PES and electronic friction lead to an increase of Ptransition. Both effects are additive and result in de-excitation probabilities that are up to 6 (2) times larger for H2 (Figure 4C) and up to 3 (2) times larger for D2 with ODF (LDFA), respectively (Figure 4D). At high incidence energies, the dissociation channel (see Figure 3A,B) becomes more effective, which is why Ptransition decreases again in all cases. For the adiabatic simulations on the static surface, ⟨ΔEtrans⟩ is equal to the rovibrational energy loss of one vibrational quantum and thus by about √2 larger for H2 than that for D2 (Figure 4A,B). Electronic friction reduces the energy gain. The reduction is almost twice as large for ODF compared to that for LDFA. The fact that it does not very strongly depend on Ei for the energy range considered here suggests that it is dominated by ηdd and thus directly reflects the differences observed along the minimum-energy path depicted in Figure 2A–D. Consequently, when comparing MDEF with other nonadiabatic models43−45 for vibrational de-excitation, our results suggest that it is crucial to also take into account whether the friction coefficients include any molecular anisotropy. Unfortunately, because molecular beam experiments for this system have hitherto focused on rovibrational excitation rather than de-excitation,46−48 experimental verification of this effect is still pending.

Although ODF-iso inherits the spurious memory effects as well as going beyond the independent atom approximation from ODF, quite remarkably, for Ei > 15 (20) kJ/mol for H2 (D2), we obtain results with ODF-iso that are almost identical to those with LDFA. That means that (at least in this energy range) these do not affect the dynamics and the molecular anisotropy is the most important difference. For lower incidence energies, scattering over the top site has been found to dominate vibrational de-excitation from ν = 1 in adiabatic calculations.42 Indeed, for top sites, ηddODF-iso is rather different from ηdd so that the difference in electronic structure inherent to LDFA and ODF also becomes visible in the dynamics in this case. If the molecular anisotropy could be experimentally validated, it would greatly encourage future theoretical work to develop extensions to LDFA that might be able to (at least approximately) account for it.

In summary, we have obtained different observables for H2 and D2 on Cu(111) from extensive MDEF trajectory calculations for the first time using full-dimensional friction tensors based on both LDFA and ODF. The molecular anisotropy as described by ODF and absent from LDFA leads to strongly enhanced friction for motion along the molecular axis when the molecules are close to the surface. The dissociative sticking probability is almost negligibly reduced compared to adiabatic simulations. The effect is slightly stronger for ODF compared to LDFA and improves the agreement with experimental data in both cases. For the state-to-state scattering of vibrationally excited molecules (from ν = 2, J = 1 (2) to ν = 1, J = 1 (2) for H2 (D2)), we predict up to six (two) times larger vibrational de-excitation probabilities with ODF (LDFA) compared to adiabatic simulations. Remarkably, isotropicalization of ODF yields results almost identical to LDFA for incidence energies larger than 15 (20) kJ/mol for H2 (D2). The predicted differences between the vibrational de-excitation probabilities are a “fingerprint” of the molecular anisotropy as described by ODF. Recently suggested techniques to prepare H2 molecular beams with 1 ≤ ν ≤ 449 should allow testing for this fingerprint. This would provide unprecedented insight into the accuracy of state-of-the-art electronic friction models for molecules and allow analysis of the importance of concomitant approximations.

Computational Details

In eq 2, we calculate the electron–phonon matrix elements Inline graphic from the change of the Kohn–Sham potential Inline graphic with respect to nuclear coordinate Riα, which is obtained from density functional perturbation theory (DFPT)50 employing the PW9151 exchange–correlation functional as implemented in the Quantum Espresso package.52 Surfaces are modeled by 2 × 2 Cu(111) slabs with four layers and 10 Å of vacuum. A plane-wave cutoff energy of 816 eV is used, together with ONCV pseudopotentials53 from the SG1554 library and an 18 × 18 k-point grid. These settings reproduce the PW91-PES from ref (34) up to a few meV. They also enable an accurate evaluation of the sum over electronic states in eq 2 at the Fermi level using an equivalent Gaussian envelope technique to broaden the δ-function with a width of 0.6 eV as suggested in ref (26). We note that this implies the possible presence of spurious electronic memory effects, as argued in ref (27).

The neural network fits for the 21 independent elements of ηiαjβODF(R) are based on ∼30000 ODF coefficients obtained from DFT calculations on the same seven lateral sites that have been used to construct the SRP48-PES.33 For LDFA, we extract the background electron density ρ(Ri) from a DFT calculation with the same computational setup as described above. Employing the functional form for ηH(ρ) suggested in ref (38), we then construct three-dimensional neural network interpolations for ηH(ρ(Ri)) based on the symmetry-adapted coordinates55,56 in order to obtain a continuous representation of ηiαjβ(R).

Acknowledgments

J.M. gratefully acknowledges financial support from The Netherlands Organisation for Scientific Research (NWO) under Vidi Grant No. 723.014.009. We would like to thank Geert-Jan Kroes and Reinhard Maurer for very insightful and stimulating discussions.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.7b03182.

  • LDFA-AIM for H2 and D2, continuous representation of 6 × 6 friction tensors, errors due to statistical sampling, and vibrational de-excitation for PW91-PES (PDF)

The authors declare no competing financial interest.

Supplementary Material

jz7b03182_si_001.pdf (313.6KB, pdf)

References

  1. Wodtke A. M.; Tully J. C.; Auerbach D. J. Electronically Non-Adiabatic Interactions of Molecules at Metal Surfaces: Can We Trust the Born-Oppenheimer Approximation for Surface Chemistry?. Int. Rev. Phys. Chem. 2004, 23, 513–539. 10.1080/01442350500037521. [DOI] [Google Scholar]
  2. Kroes G.-J. Frontiers in Surface Scattering Simulations. Science 2008, 321, 794–797. 10.1126/science.1157717. [DOI] [PubMed] [Google Scholar]
  3. Golibrzuch K.; Bartels N.; Auerbach D. J.; Wodtke A. M. The Dynamics of Molecular Interactions and Chemical Reactions at Metal Surfaces: Testing the Foundations of Theory. Annu. Rev. Phys. Chem. 2015, 66, 399–425. 10.1146/annurev-physchem-040214-121958. [DOI] [PubMed] [Google Scholar]
  4. Wodtke A. M. Electronically Non-Adiabatic Influences in Surface Chemistry and Dynamics. Chem. Soc. Rev. 2016, 45, 3641–3657. 10.1039/C6CS00078A. [DOI] [PubMed] [Google Scholar]
  5. Shenvi N.; Roy S.; Tully J. C. Nonadiabatic Dynamics at Metal Surfaces: Independent-Electron Surface Hopping. J. Chem. Phys. 2009, 130, 174107. 10.1063/1.3125436. [DOI] [PubMed] [Google Scholar]
  6. Meyer J.; Reuter K. Electron-Hole Pairs during the Adsorption Dynamics of O2 on Pd(100): Exciting or Not?. New J. Phys. 2011, 13, 085010. 10.1088/1367-2630/13/8/085010. [DOI] [Google Scholar]
  7. Ryabinkin I. G.; Izmaylov A. F. Mixed Quantum-Classical Dynamics Using Collective Electronic Variables: A Better Alternative to Electronic Friction Theories. J. Phys. Chem. Lett. 2017, 8, 440–444. 10.1021/acs.jpclett.6b02712. [DOI] [PubMed] [Google Scholar]
  8. Dou W.; Miao G.; Subotnik J. E. Born-Oppenheimer Dynamics, Electronic Friction, and the Inclusion of Electron-Electron Interactions. Phys. Rev. Lett. 2017, 119, 046001. 10.1103/PhysRevLett.119.046001. [DOI] [PubMed] [Google Scholar]
  9. Rittmeyer S. P.; Meyer J.; Reuter K. Nonadiabatic Vibrational Damping of Molecular Adsorbates: Insights into Electronic Friction and the Role of Electronic Coherence. Phys. Rev. Lett. 2017, 119, 176808. 10.1103/PhysRevLett.119.176808. [DOI] [PubMed] [Google Scholar]
  10. Head-Gordon M.; Tully J. C. Molecular Dynamics with Electronic Frictions. J. Chem. Phys. 1995, 103, 10137–10145. 10.1063/1.469915. [DOI] [Google Scholar]
  11. Alducin M.; Díez Muiño R.; Juaristi J. I. Non-Adiabatic Effects in Elementary Reaction Processes at Metal Surfaces. Prog. Surf. Sci. 2017, 92, 317–340. 10.1016/j.progsurf.2017.09.002. [DOI] [Google Scholar]
  12. Juaristi J. I.; Alducin M.; Díez Muiño R.; Busnengo H. F.; Salin A. Role of Electron-Hole Pair Excitations in the Dissociative Adsorption of Diatomic Molecules on Metal Surfaces. Phys. Rev. Lett. 2008, 100, 116102. 10.1103/PhysRevLett.100.116102. [DOI] [PubMed] [Google Scholar]
  13. Novko D.; Blanco-Rey M.; Alducin M.; Juaristi J. I. Surface Electron Density Models for Accurate Ab Initio Molecular Dynamics with Electronic Friction. Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 93, 245435. 10.1103/PhysRevB.93.245435. [DOI] [Google Scholar]
  14. Hellsing B.; Persson M. Electronic Damping of Atomic and Molecular Vibrations at Metal Surfaces. Phys. Scr. 1984, 29, 360–371. 10.1088/0031-8949/29/4/014. [DOI] [Google Scholar]
  15. Janke S. M.; Auerbach D. J.; Wodtke A. M.; Kandratsenka A. An Accurate Full-Dimensional Potential Energy Surface for H-Au(111): Importance of Nonadiabatic Electronic Excitation in Energy tTansfer and Adsorption. J. Chem. Phys. 2015, 143, 124708. 10.1063/1.4931669. [DOI] [PubMed] [Google Scholar]
  16. Bünermann O.; Jiang H.; Dorenkamp Y.; Kandratsenka A.; Janke S. M.; Auerbach D. J.; Wodtke A. M. Electron-Hole Pair Excitation Determines the Mechanism of Hydrogen Atom Adsorption. Science 2015, 350, 1346–1349. 10.1126/science.aad4972. [DOI] [PubMed] [Google Scholar]
  17. Askerka M.; Maurer R. J.; Batista V. S.; Tully J. C. Role of Tensorial Electronic Friction in Energy Transfer at Metal Surfaces. Phys. Rev. Lett. 2016, 116, 217601. 10.1103/PhysRevLett.116.217601. [DOI] [PubMed] [Google Scholar]
  18. Inoue K.-i.; Watanabe K.; Sugimoto T.; Matsumoto Y.; Yasuike T. Disentangling Multidimensional Nonequilibrium Dynamics of Adsorbates: CO Desorption from Cu(100). Phys. Rev. Lett. 2016, 117, 186101. 10.1103/PhysRevLett.117.186101. [DOI] [PubMed] [Google Scholar]
  19. Maurer R. J.; Jiang B.; Guo H.; Tully J. C. Mode Specific Electronic Friction in Dissociative Chemisorption on Metal Surfaces: H 2 on Ag(111). Phys. Rev. Lett. 2017, 118, 256001. 10.1103/PhysRevLett.118.256001. [DOI] [PubMed] [Google Scholar]
  20. Luntz A. C.; Makkonen I.; Persson M.; Holloway S.; Bird D. M.; Mizielinski M. S. Comment on “Role of Electron-Hole Pair Excitations in the Dissociative Adsorption of Diatomic Molecules on Metal Surfaces. Phys. Rev. Lett. 2009, 102, 109601. 10.1103/PhysRevLett.102.109601. [DOI] [PubMed] [Google Scholar]
  21. Juaristi J. I.; Alducin M.; Muiño R. D.; Busnengo H. F.; Salin A. Juaristi et al. Reply. Phys. Rev. Lett. 2009, 102, 109602. 10.1103/PhysRevLett.102.109602. [DOI] [Google Scholar]
  22. Luntz A. C.; Persson M. How Adiabatic is Activated Adsorption/Associative Desorption?. J. Chem. Phys. 2005, 123, 074704. 10.1063/1.2000249. [DOI] [PubMed] [Google Scholar]
  23. Krishna V.; Tully J. C. Vibrational Lifetimes of Molecular Adsorbates on Metal Surfaces. J. Chem. Phys. 2006, 125, 054706. 10.1063/1.2227383. [DOI] [PubMed] [Google Scholar]
  24. Forsblom M.; Persson M. Vibrational Lifetimes of Cyanide and Carbon Monoxide on Noble and Transition Metal Surfaces. J. Chem. Phys. 2007, 127, 154303. 10.1063/1.2794744. [DOI] [PubMed] [Google Scholar]
  25. Rittmeyer S. P.; Meyer J.; Juaristi J. I.; Reuter K. Electronic Friction-Based Vibrational Lifetimes of Molecular Adsorbates: Beyond the Independent-Atom Approximation. Phys. Rev. Lett. 2015, 115, 046102. 10.1103/PhysRevLett.115.046102. [DOI] [PubMed] [Google Scholar]
  26. Maurer R. J.; Askerka M.; Batista V. S.; Tully J. C. Ab Initio Tensorial Electronic Friction for Molecules on Metal Surfaces: Nonadiabatic Vibrational Relaxation. Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 94, 115432. 10.1103/PhysRevB.94.115432. [DOI] [Google Scholar]
  27. Novko D.; Alducin M.; Blanco-Rey M.; Juaristi J. I. Effects of Electronic Relaxation Processes on Vibrational Linewidths of Adsorbates on Surfaces: The Case of CO/Cu(100). Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 94, 224306. 10.1103/PhysRevB.94.224306. [DOI] [Google Scholar]
  28. Kroes G.-J.; Juaristi J. I.; Alducin M. Vibrational Excitation of H2 Scattering from Cu(111): Effects of Surface Temperature and of Allowing Energy Exchange with the Surface. J. Phys. Chem. C 2017, 121, 13617–13633. 10.1021/acs.jpcc.7b01096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Diaz C.; Pijper E.; Olsen R. A.; Busnengo H. F.; Auerbach D. J.; Kroes G.-J. Chemically Accurate Simulation of a Prototypical Surface Reaction: H2 Dissociation on Cu(111). Science 2009, 326, 832–834. 10.1126/science.1178722. [DOI] [PubMed] [Google Scholar]
  30. Nattino F.; Genova A.; Guijt M.; Muzas A. S.; Díaz C.; Auerbach D. J.; Kroes G.-J. Dissociation and Recombination of D2 on Cu(111): Ab Initio Molecular Dynamics Calculations and Improved Analysis of Desorption Experiments. J. Chem. Phys. 2014, 141, 124705. 10.1063/1.4896058. [DOI] [PubMed] [Google Scholar]
  31. Kroes G.-J.; Díaz C. Quantum and Classical Dynamics of Reactive Scattering of H2 from Metal Surfaces. Chem. Soc. Rev. 2016, 45, 3658–3700. 10.1039/C5CS00336A. [DOI] [PubMed] [Google Scholar]
  32. Díaz C.; Olsen R. A.; Auerbach D. J.; Kroes G.-J. Six-Dimensional Dynamics Study of Reactive and Non Reactive Scattering of H2 from Cu(111) Using a Chemically Accurate Potential Energy Surface. Phys. Chem. Chem. Phys. 2010, 12, 6499–6519. 10.1039/c001956a. [DOI] [PubMed] [Google Scholar]
  33. Mondal A.; Wijzenbroek M.; Bonfanti M.; Díaz C.; Kroes G.-J. Thermal Lattice Expansion Effect on Reactive Scattering of H2 from Cu(111) at Ts = 925 K. J. Phys. Chem. A 2013, 117, 8770–8781. 10.1021/jp4042183. [DOI] [PubMed] [Google Scholar]
  34. Diaz C.; Olsen R. A.; Busnengo H. F.; Kroes G.-J. Dynamics on Six-Dimensional Potential Energy Surfaces for H2/Cu(111): Corrugation Reducing Procedure versus Modified Shepard Interpolation Method and PW91 versus RPBE. J. Phys. Chem. C 2010, 114, 11192–11201. 10.1021/jp1027096. [DOI] [Google Scholar]
  35. Butler W. H.; Pinski F. J.; Allen P. B. Phonon Linewidths and Electron-Phonon Interaction in Nb. Phys. Rev. B: Condens. Matter Mater. Phys. 1979, 19, 3708–3721. 10.1103/PhysRevB.19.3708. [DOI] [Google Scholar]
  36. Puska M. J.; Nieminen R. M. Atoms Embedded in an Electron Gas: Phase Shifts and Cross Sections. Phys. Rev. B: Condens. Matter Mater. Phys. 1983, 27, 6121. 10.1103/PhysRevB.27.6121. [DOI] [PubMed] [Google Scholar]
  37. Gerrits N.; Juaristi J. I.; Meyer J.. Atoms in jellium revisited - Implications for the Local Density Friction Approximation. In preparation. [Google Scholar]
  38. Saalfrank P.; Juaristi J. I.; Alducin M.; Blanco-Rey M.; Muiño R. D. Vibrational Lifetimes of Hydrogen on Lead Films: An Ab Initio Molecular Dynamics with Electronic Friction (AIMDEF) Study. J. Chem. Phys. 2014, 141, 234702. 10.1063/1.4903309. [DOI] [PubMed] [Google Scholar]
  39. Askerka M.; Maurer R. J.; Batista V. S.; Tully J. C. Erratum: Role of Tensorial Electronic Friction in Energy Transfer at Metal Surfaces [Phys. Rev. Lett. 116, 217601 (2016)]. Phys. Rev. Lett. 2017, 119, 069901. 10.1103/PhysRevLett.119.069901. [DOI] [PubMed] [Google Scholar]
  40. Rettner C. T.; Michelsen H. A.; Auerbach D. J. Quantum-State-Specific Dynamics of the Dissociative Adsorption and Associative Desorption of H2 at a Cu(111) Surface. J. Chem. Phys. 1995, 102, 4625–4641. 10.1063/1.469511. [DOI] [Google Scholar]
  41. Michelsen H. A.; Rettner C. T.; Auerbach D. J.; Zare R. N. Effect of Rotation on the Translational and Vibrational Energy Dependence of the Dissociative Adsorption of D2 on Cu(111). J. Chem. Phys. 1993, 98, 8294–8307. 10.1063/1.464535. [DOI] [Google Scholar]
  42. Muzas A. S.; Juaristi J. I.; Alducin M.; Díez Muiño R.; Kroes G.-J.; Díaz C. Vibrational Deexcitation and Rotational Excitation of H2 and D2 Scattered from Cu(111): Adiabatic versus Non-Adiabatic Dynamics. J. Chem. Phys. 2012, 137, 064707. 10.1063/1.4742907. [DOI] [PubMed] [Google Scholar]
  43. Luntz A. C.; Persson M.; Sitz G. O. Theoretical Evidence for Nonadiabatic Vibrational Deexcitation in H2(D2) State-to-State Scattering from Cu(100). J. Chem. Phys. 2006, 124, 091101. 10.1063/1.2177664. [DOI] [PubMed] [Google Scholar]
  44. Golibrzuch K.; Shirhatti P. R.; Rahinov I.; Kandratsenka A.; Auerbach D. J.; Wodtke A. M.; Bartels C. The Importance of Accurate Adiabatic Interaction Potentials for the Correct Description of Electronically Nonadiabatic Vibrational Energy Transfer: A combined Experimental and Theoretical Study of NO(v = 3) Collisions with a Au(111) Surface. J. Chem. Phys. 2014, 140, 044701. 10.1063/1.4861660. [DOI] [PubMed] [Google Scholar]
  45. Krüger B. C.; Bartels N.; Bartels C.; Kandratsenka A.; Tully J. C.; Wodtke A. M.; Schäfer T. NO Vibrational Energy Transfer on a Metal Surface: Still a Challenge to First-Principles Theory. J. Phys. Chem. C 2015, 119, 3268–3272. 10.1021/acs.jpcc.5b00388. [DOI] [Google Scholar]
  46. Hodgson A.; Moryl J.; Traversaro P.; Zhao H. Energy transfer and vibrational effects in the dissociation and scattering of D2 from Cu (111). Nature 1992, 356, 501. 10.1038/356501a0. [DOI] [Google Scholar]
  47. Rettner C. T.; Auerbach D. J.; Michelsen H. A. Observation of Direct Vibrational Excitation in Collisions of H2 and D2 with a Cu(111) Surface. Phys. Rev. Lett. 1992, 68, 2547–2550. 10.1103/PhysRevLett.68.2547. [DOI] [PubMed] [Google Scholar]
  48. Hodgson A.; Samson P.; Wight A.; Cottrell C. Rotational Excitation and Vibrational Relaxation of H 2 (υ = 1, J= 0) Scattered from Cu (111). Phys. Rev. Lett. 1997, 78, 963. 10.1103/PhysRevLett.78.963. [DOI] [Google Scholar]
  49. Perreault W. E.; Mukherjee N.; Zare R. N. Preparation of a Selected High Vibrational Energy Level of Isolated Molecules. J. Chem. Phys. 2016, 145, 154203. 10.1063/1.4964938. [DOI] [PubMed] [Google Scholar]
  50. Baroni S.; de Gironcoli S.; Dal Corso A.; Giannozzi P. Phonons and Related Crystal Properties from Density-Functional Perturbation Theory. Rev. Mod. Phys. 2001, 73, 515–562. 10.1103/RevModPhys.73.515. [DOI] [Google Scholar]
  51. Perdew J. P.; Chevary J. A.; Vosko S. H.; Jackson K. A.; Pederson M. R.; Singh D. J.; Fiolhais C. Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation. Phys. Rev. B: Condens. Matter Mater. Phys. 1992, 46, 6671–6687. 10.1103/PhysRevB.46.6671. [DOI] [PubMed] [Google Scholar]
  52. Giannozzi P.; Baroni S.; Bonini N.; Calandra M.; Car R.; Cavazzoni C.; Ceresoli D.; Chiarotti G. L.; Cococcioni M.; Dabo I.; et al. QUANTUM ESPRESSO: A Modular and Open-Source Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter 2009, 21, 395502. 10.1088/0953-8984/21/39/395502. [DOI] [PubMed] [Google Scholar]
  53. Hamann D. R. Optimized Norm-Conserving Vanderbilt Pseudopotentials. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 88, 085117. 10.1103/PhysRevB.88.085117. [DOI] [Google Scholar]
  54. Schlipf M.; Gygi F. Optimization Algorithm for the Generation of ONCV Pseudopotentials. Comput. Phys. Commun. 2015, 196, 36–44. 10.1016/j.cpc.2015.05.011. [DOI] [Google Scholar]
  55. Meyer J.Ab Initio Modeling of Energy Dissipation during Chemical Reactions at Transition Metal Surfaces. Ph.D. thesis, Freie Universität Berlin, Germany, 2012. [Google Scholar]
  56. Goikoetxea I.; Beltrán J.; Meyer J.; Juaristi J. I.; Alducin M.; Reuter K. Non-Adiabatic Effects during the Dissociative Adsorption of O2 at Ag(111)? A First-Principles Divide and Conquer Study. New J. Phys. 2012, 14, 013050. 10.1088/1367-2630/14/1/013050. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

jz7b03182_si_001.pdf (313.6KB, pdf)

Articles from The Journal of Physical Chemistry Letters are provided here courtesy of American Chemical Society

RESOURCES