Relativistic DFT investigation of electronic structure effects arising from doping the Au25 nanocluster with transition metals.

View Article Online View Journal

Nanoscale Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: F. Alkan, A. Muñoz-Castro and C. M. Aikens, Nanoscale, 2017, DOI: 10.1039/C7NR05214F. Volume 8 Number 1 7 January 2016 Pages 1–660

Nanoscale www.rsc.org/nanoscale

This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication. Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available. You can find more information about Accepted Manuscripts in the author guidelines.

ISSN 2040-3364

PAPER Qian Wang et al. TiC2: a new two-dimensional sheet beyond MXenes

Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the ethical guidelines, outlined in our author and reviewer resource centre, still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

rsc.li/nanoscale

Page 1 of 21

Nanoscale View Article Online

DOI: 10.1039/C7NR05214F


Published on 03 October 2017. Downloaded by Freie Universitaet Berlin on 03/10/2017 19:13:11.

Graphical Abstract

Relative energetics of the dopant d levels and super-atomic orbitals influence the isomers and states available in monolayer-protected clusters

1

Nanoscale

Page 2 of 21 View Article Online

DOI: 10.1039/C7NR05214F

Relativistic DFT Investigation of Electronic Structure Effects Arising from Doping the Au25 Nanocluster with Transition Metals

a) Department of Chemistry, Kansas State University, Manhattan, KS 66506, USA b) Grupo de Química Inorgánica y Materiales Moleculares, Universidad Autonoma de Chile, El Llano Subercaseaux 2801, Santiago, Chile c) Doctorado en Fisicoquímica Molecular, Universidad Andres Bello, Av. Republica 275, Santiago, Chile *[email protected]; 1-785-532-0954

Abstract We perform a theoretical investigation using density functional theory (DFT) and timedependent DFT (TDDFT) on the doping of the Au25(SR)18-1 nanocluster with group IX transition metals (M = cobalt, rhodium and iridium). Different doping motifs, charge states and spin multiplicities were considered for the single-atom doped nanoclusters. Our results show that the interaction (or the lack of interaction) between the d-type energy levels that mainly originate from the dopant atom and the super-atomic levels plays an important role in the energetics, the electronic structure and the optical properties of the doped systems. The evaluated MAu24(SR)18q (q= -1, -3) systems favor an endohedral disposition of the doping atom typically in a singlet ground state, with either a 6- or 8-valence electron icosahedral core. For the sake of comparison, the role of the d energy levels in the electronic structure of a variety of doped Au25(SR)18-1 nanoclusters was investigated for dopant atoms from other families such as Cd, Ag and Pd. Finally, the effect of spin-orbit coupling (SOC) on the electronic structure and absorption spectra was determined. The information in this study regarding the relative energetics of the d-based and super-atom energy levels can be useful to extend our understanding of the preferred doping modes of different transition metals in protected gold nanoclusters.

1



Fahri Alkana, Alvaro Muñoz-Castrob,c, Christine M. Aikensa,*

Page 3 of 21


INTRODUCTION Atomically precise gold nanoparticles have raised a considerable amount of research interest over the last decade due to their distinct electronic, optical and magnetic properties.1-12 These nanoparticles show promising applications in areas such as catalysis,13-15 biosensors16-18 and solar cell research.19 In particular, there have been significant advances in synthesis, characterization and theoretical understanding of thiolate-protected gold nanoparticles (with a general formula of Aun(SR)mq).4, 7, 20 These systems typically consist of an inner gold core that is protected by several Au-SR staple units. Some of the most prominent examples of thiolateprotected gold nanoparticles include Au25(SR)18-1,1, 21 Au38(SR)2422, 23 and Au102(SR)44.9 For a more comprehensive list of Aun(SR)mq nanoclusters with experimentally-resolved crystal structures, readers are referred to the recent reviews by Jin et al.10 and Pradeep et al.24 Among the studied thiolate-protected gold nanoparticles, the Au25(SR)18-1 system (Au25-1 for short) has attracted the most attention both in theoretical and experimental research, because it provides a prototypical example of a stable system exhibiting both interesting geometric and electronic stability.25 The structure of this system displays an icosahedral Au13 core, which is protected by six Au2(SR)3 dimeric staple units.1, 21 A super-atom model26, 27 is often utilized successfully to explain the ground state electronic structure and the optical, structural and magnetic properties of the Au25 cluster.28 In this sense, Au25(SR)18-1 exhibits a 1S2|1P6|1D0| electronic configuration with 8 valence electrons, which is analogous to the noble gas electronic configuration. Thus, the resulting closed super-atomic shell configuration is often related to the large HOMO-LUMO gap and chemical stability of this cluster.29, 30 Recently, other theoretical models regarding the stability of thiolate-protected gold nanoparticles have been proposed as well, as useful predictive tools for ligand-protected clusters.31-33 Besides changes in properties that arise from different sizes or nuclearities of the nanoclusters,6 the introduction of a different metal atom in these structures has been shown to offer an interesting approach to increasing the versatility of these systems and enabling the ability to tune cluster properties.4, 34 The possibility of metal doped Au25 clusters was first confirmed experimentally in 2009 for the PdAu24(SR)18 system by the Murray group.35 In the same year, Jiang and Dai36 reported a theoretical investigation for various M@Au24(SR)18q systems by applying the super-atom concept, Walter and Moseler37 considered Au25 doping with Pd, Ag, and Cd, Kacprzak et al.38 examined Pd doping, and Jiang and Whetten39 investigated magnetic doping of this system. These investigations led to considerable interest in the doping studies of the Au25 cluster. Since then, several doped Au25 clusters with one or more heteroatoms such as Pd,40-44 Pt,41, 42 Cu,45, 46 Ag,45, 47-49 Cd,50 Hg48, 50 and Ir51 have been synthesized and characterized successfully, underlining the present experimental capabilities to incorporate atoms from different groups. Doped Au25 clusters have been investigated both experimentally and theoretically to enable fine-tuning of HOMO-LUMO gaps, optical and magnetic properties, and catalytic activity.37-39, 52-54 Another important issue affecting the properties of doped structures is the location of the foreign atom in the gold nanocluster. Recently, X-ray crystallography and other experimental techniques along with DFT methods have been utilized to reveal the geometric structures of the doped Au25 nanoclusters.42, 43, 45, 48-50, 55 These studies often conclude that the doping motif in the nanocluster depends heavily on the nature of the dopant atom. However, the mechanisms that drive the dopant atom to the preferred location within the cluster are still subject to debate. Herein, our attention focuses primarily on elucidating both the structural and electronic modifications by doping the Au25 nanocluster with group IX transition metals, which has not

2



DOI: 10.1039/C7NR05214F

Nanoscale


been described in detail in the literature. We report a computational study using relativistic DFT methods for the ground and excited state properties of singly doped Au25, i.e. MAu24q, where M = Co, Rh and Ir. Different isomers, charge states and spin multiplicities are considered for the investigated systems. We also extend our investigation to other possible MAu24q systems to find general trends in the electronic structure and stability of doped gold clusters. Our aim is to explore the differences in energetics, electronic structure and optical properties induced by doping with foreign metals.



DOI: 10.1039/C7NR05214F

3

Page 5 of 21


COMPUTATIONAL METHODS Computations were performed on the native Au25-1 and doped clusters (MAu24q) where a single Au atom is replaced with a dopant (M = Co, Rh and Ir; q = -3, -1). The charges of the clusters are selected on the basis of the overall electron count in the experimentally characterized clusters, and are considered in relation to the parent Au25-1 cluster. Three different positions are considered for the dopant atom: the center of the cluster (isomer I), the inner shell of the cluster (isomer II) and the ligand shell of the cluster (isomer III). The resulting isomers of the MAu24q system are illustrated in Figure 1.

Figure 1. Different isomers for MAu24(SR)18q clusters (R=H). The gold atoms in the core of the nanoparticle are shown with balls and the ligand shell is illustrated using sticks for clarity. The blue ball represents the dopant atom. All calculations were performed using density functional theory (DFT) with the PBE functional56, 57 and a triple-ζ polarized (TZP) basis set. The frozen core approximation was applied to the core electrons (1s-2p for S, 1s-3p for Co, 1s-4p for Rh, Pd, Ag, and Cd, 1s-4d for In, and 1s-4f for Au and Ir) of the elements. The geometry optimizations were performed in the gas phase. The energy and gradient convergence thresholds were set to 1x10-4 and 1x10-3, respectively. For the isomer I structure, the Ci point-group symmetry was adapted in the starting geometries, whereas no symmetry constraint was employed for isomer II or III structures. The vertical excitations were computed by employing time-dependent DFT (TDDFT) with the relaxed geometries of the clusters. In order to further evaluate the effect of solvent (toluene) on the calculated absorption spectra, calculations were done by including the conductor-like screening model (COSMO) model for solvation. The inclusion of solvent effects in test cases only exhibits a uniform increase in the intensity of the peaks, while it has negligible effects on the excitation energies. Therefore, hereafter solvent effects are not discussed. The relativistic effects were included using the zeroth-order regular approximation (ZORA) Hamiltonian at the scalar or spin-orbit level.58-60 The computations were carried out by using the Amsterdam Density Functional (ADF) 2016 package.61-63

4



DOI: 10.1039/C7NR05214F

Nanoscale


RESULTS AND DISCUSSION 1. Electronic structure and stability In Table 1, we tabulate the relative energies and HOMO-LUMO gaps for the three possible isomers of MAu24q (M = Co, Rh and Ir) systems. For the trianionic charge state (q=-3), only the singlet configuration is considered for the doped systems; this charge state is isoelectronic to the parent Au25-1 cluster. In comparison, both singlet and triplet configurations are investigated for systems with a minus one charge. For q=-3, isomer I is predicted to be the energetically most favorable compared to isomers II and III. The energy differences between the most stable isomer (isomer I) and the other isomers are quite large and range between 35.9-59.1 kcal/mol, indicating that systems II and III are high-lying isomers, in comparison to the preferred endohedral incorporation of the doping metal. There is a similar correlation between the dopant location and the predicted HOMO-LUMO gaps, which also provide a prediction of isomer stability. For isomer I, the clusters exhibit the largest HOMO-LUMO gaps, which range between 1.33 eV and 1.84 eV. The calculated HOMO-LUMO gaps also systematically increase from CoAu24-3 to IrAu24-3. Additionally, the calculated HOMO-LUMO gaps of isomer I are predicted to be larger than the HOMO-LUMO gap obtained for the native Au25-1 cluster (1.21 eV) calculated at the same level of theory. For isomers II and III, the HOMO-LUMO gaps decrease significantly, which correlates with our findings regarding to relative isomer energies. In the case of isomer III of the CoAu24-3 cluster, geometry optimization did not reach the convergence threshold due to SCF convergence problems in the electronic singlet state. In contrast, the geometry convergence for the triplet configuration of the same system was successful. The resulting relative energy of this system was 31.8 kcal/mol, which also shows a significant destabilization from the isomer I (singlet) structure. Table 1. Relative isomer energies and HOMO-LUMO gaps for different isomers of MAu24q systems, calculated at the PBE-ZORA/TZP scalar relativistic level of theory. HOMO-LUMO gap (eV) Relative Energies (kcal/mol)a System Isomer Isomer Isomer Isomer Isomer Isomer III I II III I II q=-3 CoAu24 0.0 39.0 31.8b 1.33 0.25 0.19b RhAu24 0.0 35.9 43.3 1.55 0.71 0.61 IrAu24 0.0 50.1 59.1 1.84 0.63 0.49 q=-1 Singlet CoAu24 RhAu24 IrAu24 q=-1 Triplet CoAu24 RhAu24 IrAu24

0.0 0.0 0.0

----14.7 21.7

-------------

0.10 0.45 0.40

----0.37 0.44

-------------

-8.9 4.3 3.6

-7.6 17.9 27.7

-8.4 23.3 34.5

0.33 0.07 0.08

0.72 0.19 0.02

0.71 0.18 0.05

a) The calculated energies of (-3) singlet states and (-1) singlet states of isomer I for each doped cluster are set to zero while the energies of the rest of the systems are

5



DOI: 10.1039/C7NR05214F

Page 7 of 21


DOI: 10.1039/C7NR05214F

For the minus one charge state of the MAu24-q clusters, isomer I is also predicted as the most stable isomer from the comparison of relative energies. For RhAu24-1 and IrAu24-1, the most stable configuration is the singlet configuration, whereas for CoAu24-1, the triplet configuration is found to be more stable by 8.9 kcal/mol. This observation denotes differences between lighter and heavier members of group IX. It should also be noted that all isomers I-III of CoAu24-1 in the triplet configuration have quite similar energies which lie in a range of only 1.3 kcal/mol. In comparison, the triplet configuration of isomer I is less stable than the singlet one by ~4 kcal/mol in the case of RhAu24-1 and IrAu24-1. For these systems, isomers II and III exhibit a large destabilization in bonding energies compared to isomer I. For a better understanding of the trends observed in the energetics and HOMO-LUMO gaps of MAu24q clusters, we analyzed the electronic structure of the different isomers of doped systems. The results are illustrated in Figure 2 for the q=-3 cases. A detailed analysis of the final molecular orbital (MO) populations using fragment orbitals is given in the electronic supplementary information (ESI). In the electronic configuration of isomer I, low-lying occupied MOs (HOMO to HOMO-7) consist of two distinct bands: three energy levels which mainly have super-atomic P orbital character that are delocalized on the MAu12 kernel (1P), and five levels which have significant atomic d character originating from the dopant atoms (nd). The occupied nd levels also have considerable super-atomic D orbital character. In comparison, the LUMO and LUMO+1, denoted 1D, have near double degeneracy and possess mainly super-atomic D character with some contribution from atomic (dopant) d orbitals. In CoAu24-3, the HOMOHOMO-4 levels are nearly 5-fold degenerate and originate from the atomic d orbitals of Co. The energy separation between the nd and 1P levels is about 0.1 eV. In the case of RhAu24-3 and IrAu24-3, the energy ordering of the nd and 1P shells is switched compared to the case in CoAu243 . For all doped systems that exhibit the isomer I structure, there is no mixing allowed in the electronic structure between d and P orbitals, due to the symmetry of the system.

Figure 2. The electronic configuration of low-lying states for different isomers of a) CoAu24-3, b) RhAu24-3, and c) IrAu24-3 systems. The 1P and 1D bands are shown in red boxes, and the nd

6



reported accordingly. b) The calculated energy and HOMO-LUMO gap correspond to the triplet configuration of this system instead of the singlet configuration.

Nanoscale


DOI: 10.1039/C7NR05214F

As expected from the calculated energies, the electronic structures of the doped clusters are altered considerably when the dopant atom is not located at the center of the cluster. Due to the lowering of symmetry, energy levels that are degenerate in isomer I are mostly split into singly degenerate levels in the electronic structures of isomer II or III. Another important change in the electronic structure is the significant relative destabilization of the nd levels. As a result, for Rh and Ir doped systems the energy ordering of the nd band and 1P band switches in isomers II and III compared to the case in isomer I. We also note that there is a notable mixing between atomic d orbitals and super-atomic P orbitals in some of the MOs of isomers II and III. An example of this interaction is illustrated in Figure 3 for the IrAu24-3 case. In comparison, the super-atomic D character in the nd levels of isomer II and III is somewhat reduced. Overall, a change in the dopant location alters the contributions of atomic d orbitals and super-atomic P and D orbitals in low-lying states significantly due at least in part to the symmetry considerations between these levels.

Figure 3. Selected MOs for isomer I (a and b) and isomer II (c and d) of the IrAu24-3 system. For a) and b), the selected MOs belong to the 1P and nd levels, respectively, and illustrate the case with no mixing between P and d orbitals. In comparison, MOs shown in c) (1P) and d) (nd) demonstrate a case with significant mixing between atomic d orbitals and super-atomic P orbitals.

7



band is shown in a blue box. Each diamond in the nd and 1P levels indicates fully occupied orbitals.

Page 9 of 21


In Figure 4, we show the singlet and triplet electronic configurations of low-lying states for the isomer I structure of the MAu24q systems for the case of q=-1. Surprisingly, the electronic structure of singlet CoAu24-1 (Figure 4a) exhibits a 1P4|nd10|1Pʹ0| configuration with the third 1Pʹ orbital designating the LUMO level, despite the fact that the HOMO is predicted to be an nd level in the q=-3 case of the same system. In comparison, the electronic configuration becomes nd8|1P6|ndʹ0| for the triplet state, where nd′ denotes the unoccupied nd orbital. As shown in Table 1, the triplet configuration is energetically more favorable compared to the singlet configuration for this system. For RhAu24-1 and IrAu24-1 (Figure 4b and 4c respectively), the changes in the electronic structure with respect to the spin state are rather small. In these cases, both singlet and triplet states exhibit the nd10|1P4|1Pʹ0| electronic configuration. However, the HOMO-LUMO gaps are roughly 0.4 eV larger for the singlet states compared to the triplet states.

Figure 4. The electronic configuration of singlet and triplet states for the isomer I structure of a) CoAu24-1 b) RhAu24-1 and c) IrAu24-1 systems. For the triplet state, α levels are illustrated with solid lines, whereas β levels are illustrated with dotted lines. Each diamond indicates fully occupied orbitals; in singlet configurations this denotes two electrons, and in triplet states (openshell) the diamond represents one-electron. 2. Comparison with other doped systems It is quite evident from our electronic structure analysis of the doped MAu24q systems that the interplay between atomic d levels and the super-atomic P and D levels plays an important role in the calculated energies and the relative stabilities of different isomers. In this study, isomer I is theoretically predicted to be the most stable isomer for MAu24q clusters, where M=Co, Rh and Ir. Among other investigated doped MAu24 systems in the literature, the same result has been shown theoretically for Pd and Pt doped clusters as well.38, 64 Additionally, X-ray crystallography and other experimental techniques suggest that the structures of Pd and Pt doped MAu24 clusters are consistent with DFT predictions.41-44 On the other hand, Yao et al.50 have recently shown that the X-ray crystal structure of the Cd doped MAu24 cluster exhibits the isomer II structure. In order to understand the nature of the atomic arrangement in doped species, we investigate the energetics and the electronic structure for different isomers of MAu24q that are isoelectronic to the parent Au25-1 cluster, where M=Rh, Pd, Ag, Cd, In and q=-3, -2, -1, 0, 1,

8



DOI: 10.1039/C7NR05214F

Nanoscale


respectively. Figure 5 shows the relative energies for different isomers of the investigated systems. In this comparison, the calculated relative energy of isomer I for each cluster is set to zero, whereas the rest are scaled accordingly. For Rh and Pd doped clusters, isomer I is the most stable structure whereas for Ag, Cd and In doped clusters, isomer II becomes the most stable isomer. We note that the difference in the calculated energies of isomer I compared to isomer II or III is quite large for RhAu24-3 and ranges around 40 kcal/mol. In comparison, the same difference for PdAu24-2 is considerably lower, at about 10 kcal/mol. Interestingly, for Cd and Ag doped clusters, isomer I is predicted to be the least stable isomer. For these systems, calculated isomer energies lie within a range of 16-18 kcal/mol. In the case of InAu24+1, isomer III is predicted as the least stable isomer, and the relative energies of different isomers for this system lie within a range of 18 kcal/mol, similar to the case in Cd and Ag doped isomers.

Figure 5. Calculated relative energies of isomer II and III structures for gold clusters doped with different atoms. In each system, the calculated energy of the isomer I structure is set to zero, whereas the energies of isomers II and III are reported accordingly. To understand the extent of the effects of atomic d levels on the electronic structure and the energetics of the doped Au25 cluster, we compare the electronic structures of Pd, Ag and Cd doped clusters for the occupied levels for isomers I and II. As shown in Figure 6, the result of doping on the energy levels leads to three different cases in these systems. For isomer I of PdAu24-2, the atomic d levels sit between super-atomic P levels and the Au d band, similar to the case of Rh or Ir doped clusters. In isomer II, we note that there is a considerable amount of mixing between atomic d levels and super-atomic P levels, which leads to an overall destabilization of the frontier energy levels. In comparison, the atomic d levels of Ag in both isomers I and II of AgAu24-1 contribute to the d band of the system, and mix with Au d orbitals as shown in Figure 6. We note that no mixing is observed between atomic d levels and superatomic P or D levels in either isomer. Instead, Ag 5p levels contribute to the super-atomic P and D levels in isomer I and isomer II respectively. In CdAu24, the atomic d levels of Cd lie lower in energy than the 1S level. For this system, the nd levels mostly stay inert as we switch from isomer I to II. Similar to the case in the Ag doped system, Cd 5p orbitals contribute to the super-

9



DOI: 10.1039/C7NR05214F

Page 11 of 21


DOI: 10.1039/C7NR05214F

Figure 6. The electronic configuration schematics of occupied levels for Cd, Ag and Pd doped systems. The 1S and 1P bands are shown in red boxes, and the nd band is shown in a blue box.

3. Excited state properties The effect of doping on the optical properties of small gold nanoclusters is an active research area. Some recent experimental and theoretical work includes doping of the Au25 cluster with Ag, Cd, Pd, Pt, Hg and group XIV elements.38, 40-42, 48-50, 52 Herein, we calculated the excited states of Co, Rh and Ir doped clusters using TDDFT in order to provide useful information in further explorative synthesis efforts, allowing an initial characterization of the sample prior to structural characterization via X-ray experiments. The resulting theoretical spectra in the near IR and UV/Vis region of MAu24-3 species along with the undoped Au25 cluster are shown in Figure 7.

10



atomic P and D levels in the electronic structure. These results suggest that the interaction, or the lack of interaction, between the dopant atom’s orbitals and the super-atomic P or D levels plays an important role in determining the relative stabilities of the different isomers. It should be noted that for other dopants such as Ag and Cd, other factors such as the size of the dopant, or the electronegativity differences between dopant atom and Au, can be a determining factor for relative stabilities as well.

Nanoscale


Figure 7. The calculated UV/vis/NIR spectra of native Au25-1 and doped MAu24-3 clusters. The calculated spectrum for the Au25-1 cluster mainly show four features in the lowenergy region (1.0-3.0 eV). As discussed previously in the literature,1, 65, 66 the lowest energy peak (A) is the result of the electronic transitions between super-atomic occupied 1P and lower unoccupied 1D levels, which is usually designated as the HOMO → LUMO transition. The second feature (B) occurs around 2.2 eV and originates mainly from the transition between Au d + thiolate based orbitals and lower 1D levels. There is also some contribution to this feature from 1P → upper 1D transitions as well. The third feature (C) observed at 2.5 eV mainly has 1P → 1D character. Finally, the peak at 2.7 eV originates almost entirely from the transitions between Au d + thiolate based orbitals and 1D levels. In the case of Co, Rh or Ir doped clusters, two main differences are observed in the optical absorption spectra when compared to the spectrum of the native Au25-1 cluster. First, the position of the first peak consistently moves to higher energies (blue-shifts) in the Au25 -1 → CoAu24-3 → RhAu24-3 → IrAu24-3 series, as expected from the changes in HOMO-LUMO gaps in the doped species (Table 1). The origin of this peak is also a 1P →1D transition in the doped systems, similar to the case in the Au25-1 cluster. Secondly, peak B becomes more intense and the splitting between B and C becomes more apparent for the doped species. For all doped systems, B originates mainly from the 1P → upper 1D transition. In the case of CoAu24-3, there is also some contribution to this peak from the transitions between nd levels and orbitals that originate from Au d + thiolate units. In this system, peak C (2.46 eV) originates from 1P → 1D transitions similar to the Au25-1 case. In comparison, nd → 1F transitions contribute significantly to peak C in the case of RhAu24-3 and IrAu24-3. In Figure 8, we illustrate a selected example of an occupiedunoccupied orbital pair that contributes to these transitions. In all doped species, the higherenergy end of the spectra (2.8-3.0 eV) is dominated by the transitions originating from Au d + thiolate band. A more detailed analysis of the excited states responsible for the peaks, oscillator strengths and contributing electronic configurations is given in the SI.

11



DOI: 10.1039/C7NR05214F

Page 13 of 21


Figure 8. An example of an occupied-virtual MO pair that contributes significantly to the nd (a) → 1F (b) transitions. It is important to assess how the optical spectra of the doped systems evolve with respect to charge state since the native Au25 system can carry different charges (q=+1, 0, -1).67 In Figure 9, we present the optical absorption spectra of singlet state MAu24-1 clusters along with the spectra of MAu24-3 clusters. In general, the peaks in the spectra of the q=-1 charge state system become broader compared to the q=-3 case. For CoAu24-1 and RhAu24-1, the first and second peaks shift slightly to higher energies compared to the spectra of their q=-3 counterparts. Similar to the q=-3 case, these peaks also originate mainly from the transitions between super-atomic 1P and 1D levels. The shift in energies can be attributed to the 1S2|1P6|→1S2|1P4|1Pʹ0| transformation in the electronic structure with the change in charge state and the resulting splitting of the degeneracies in P levels as a result of the Jahn-Teller effect, similar to the case observed in the Au25+1 cluster.67 In the case of IrAu24-1, the spectrum exhibits an additional peak in the 1.5 eV region. This feature results due to the transitions from the Au nd band to the unoccupied 1P level. In comparison, the super-atomic 1P→1D transition contributes to the peak at 2.0 eV and the peaks in the 2.5 eV region. We note that the lower-energy excited states in the 0.0-1.0 eV region (not shown in the spectra) mainly originate from the transitions between nd levels to the unoccupied 1P level in all systems. However, the oscillator strengths of these excitations are quite small compared to the features observed in the 1.5-3.0 eV region. The observed differences in the calculated UV/vis/NIR spectra expose the modification of the optical absorption properties driven by the change between -3 and -1 charge states, revealing useful characteristic patterns for further experimental characterization.

12



DOI: 10.1039/C7NR05214F

Nanoscale




DOI: 10.1039/C7NR05214F

Figure 9. The comparison of calculated UV/vis/NIR spectra for MAu24q systems where q=-1 case is shown with solid lines, and q=-3 case is shown with dotted lines.

13

Page 15 of 21


4. Effect of spin-orbit coupling Inclusion of spin-orbit coupling (SOC) has important consequences in the electronic configuration and optical properties of systems containing heavy atoms.68 In the case of the Au251 cluster, it is shown that the triply degenerate super-atomic 1P level splits into an energetically higher and doubly degenerate 1P3/2 level and a single lower energy 1P1/2 level.69 Consequently, this splitting of super-atomic levels has effects on the optical absorption spectra as well. In Figure 10, we illustrate the effect of SOC on the electronic structure of doped MAu24-3 clusters. Under SOC, the otherwise triply degenerate P orbitals split into two energetically different levels, similar to the case in the electronic structure of the Au25-1 cluster. The calculated magnitudes of the SOC energy level splittings are 0.26, 0.22 and 0.21 eV in IrAu24-3, RhAu24-3 and CoAu24-3 respectively. The calculated SOC constants for the doped systems are comparable to the value obtained for Au25-1 (0.23 eV) at the same level of theory.68 In comparison, the effect of SOC on the d levels depends strongly on the nature of the dopant atom as expected. In the case of CoAu24-3, the atomic d levels are barely affected by the SOC. For RhAu24-3, the d levels shift slightly (~0.1 eV) to higher energies without notable splitting. For IrAu24-3, the higher-lying doubly degenerate d orbitals shift to higher energies, similar to the case in RhAu24-3. On the other hand, the triply degenerate d orbitals split into two levels with an energy difference around 0.1 eV.

Figure 10. The effect of SOC on the energy levels of MAu24-3 systems While the SOC affects the electronic structure considerably, it has very little effect on the relative stabilities of different isomers. The relative energies of the isomers (for q=-3) only change by 1-2 kcal/mol when SOC is included in the geometry optimization, compared to the case when only scalar relativistic effects are included. In comparison, HOMO-LUMO gaps become smaller by 0.01-0.25 eV with SOC, due to the splitting of the levels. These results are shown in Table S3 of the supplementary information.

14



DOI: 10.1039/C7NR05214F

Nanoscale


DOI: 10.1039/C7NR05214F

In Figure 11, we show the comparison of the optical absorption spectra of the doped MAu24 clusters and native Au25-1 cluster calculated with or without the SOC coupling. Due to the broken degeneracy of the P levels under the SOC regime, the first peak in the spectrum of the Au25-1 cluster is split into two peaks as shown by experimental low-temperature measurements (78 K),70 owing to the presence of 1P1/2 → 1D and 1P3/2 → 1D transitions. Similarly, this splitting is readily observable for the absorption spectra of the doped clusters as well. The magnitude of the splitting is ~0.2 eV in all cases, which is in agreement with the calculated energy differences between 1P3/2-1P1/2 levels from the core. In the spectra of doped systems, the splitting of the second peak (B) becomes more visible compared to the case in Au25-1. The splitting of this peak is also related to the aforementioned splitting of super-atomic P levels under SOC, since peak B originates mainly from transitions between P levels and higher-lying D levels.

Figure 11. The calculated UV/vis/NIR spectra with the inclusion of SOC (solid lines) and the comparison to the scalar-only case (dotted lines) for MAu24-3 systems. The impact of SOC on the electronic structure and the theoretical spectra of MAu24-1 clusters are less obvious compared to the case of MAu24-3, since the P levels in the former are already split into two sets as a result of the intermediate core valence electron count (6e) followed by further distortion by the Jahn-Teller effect. As an example, we show the changes in the electronic structure (Figure 12a) and the optical absorption spectra (Figure 12b) of RhAu24-1 with the inclusion of SOC as an example. In this case, the effects of SOC on the energetics of atomic d levels and the unoccupied P level are very small. The occupied P levels, which are nearly degenerate at the scalar relativistic level of theory, split into two levels under SOC. However, the energy separation between these two levels is less than 0.1 eV. As a result, the calculated spectrum with SOC shows only one peak at 1.7 eV, almost identical to the spectrum calculated with only scalar relativistic effects. The same observation can be extended to other

15



-3

Page 17 of 21


DOI: 10.1039/C7NR05214F

Figure 12. a) The effect of SOC on the energy levels for the RhAu24-1 system in the singlet state and b) calculated UV/vis/NIR spectra with (solid line) and without (dotted line) the inclusion of SOC for the same system. CONCLUSION The energetics, electronic structure and optical absorption spectra of Co, Rh and Ir doped MAu24q clusters were theoretically investigated using DFT and TDDFT methods, where distinctive properties were observed exposing the differences arising from inclusion of different metal atoms from group IX. Our results show that the most stable isomer occurs when the dopant atom is located at the center of the clusters, namely isomer I. The calculated energy differences between isomer I and other possible isomers are quite large, ranging between 35.9-59.1 kcal/mol for clusters with the q=-3 charge state. For each dopant, isomer I also exhibits the largest HOMO-LUMO gap compared to the other isomers. For the q=-1 charge state, isomer I is predicted to be the most energetically favorable isomer as well. However, the differences in isomer energies are smaller compared to the case with q=-3, especially for the CoAu24-1 cluster. For the MAu24-1 systems, the singlet configuration of isomer I is energetically more favorable than the triplet configuration for Rh and Ir doped clusters, whereas the opposite is the case for the Co doped cluster. The low-lying energy levels in the electronic configuration of MAu24-3 systems (HOMOHOMO-7) consist of five levels that mainly originate from the dopant d orbitals and three levels that originate from the super-atomic P orbitals. It is shown that the interaction between atomic d levels and super-atomic P and D levels plays a significant role in the relative stabilities of isomers. The same interaction is also important for other doped systems such as PdAu24-2 and PtAu24-2. In comparison, there is no mixing observed between atomic d orbitals and super-atomic orbitals in the electronic configuration of Ag or Cd doped clusters, due to the large energy differences between these levels. For Ag or Cd doped clusters, it is possible that other

16



doped systems and the native Au25+1 cluster, which also exhibits an intermediate 1S21P4 electronic configuration.

Nanoscale


mechanisms become more pronounced for the relative stabilities of the isomers, such as symmetry or the interaction between ligands and the dopant atom. Overall, these results may provide some insights for future experimental efforts that aim for synthesis and characterization of relevant doped species. The calculated UV/vis/NIR spectra of MAu24-3 systems show some resemblance to spectra of the native Au25-1 cluster, especially in the overall spectral shape and the nature of the transitions. On the other hand, some changes occur in the peak positions and intensities upon doping. Among these changes in the spectra, the most visible one is the consistent blue-shift of the first peak from Co to Ir doping. Other changes in the spectra can be attributed to the transitions that involve the atomic d levels of the dopant. Under spin-orbit coupling, the degeneracy of the super-atomic P levels in the electronic configuration of doped systems is lifted, in a similar manner to the native Au25-1 cluster. The calculated SOC constants for the P levels of doped systems are also in good agreement with the SOC constant of Au25-1. Due to the broken degeneracy, the transitions that involve the P levels become multiplets under SOC, resulting in several new features in the spectra. The effect of SOC on the atomic d levels strongly depends on the nature of the dopant atom, and it is most pronounced in the case of Ir doping as expected. The SOC effects on the electronic structure and the spectra depend on the charge state of the cluster. We expect that our results can be useful for further understanding and characterization of singly doped thiolate-protected clusters, which can serve as guidance for possible explorative synthesis efforts. Conflict of Interest There are no conflicts to declare. Acknowledgements This material is based on work supported by the National Science Foundation under Grant CHE-1507909. A.M.-C. acknowledges FONDECYT 1140359. The computing for this project was performed on the Beocat Research Cluster at Kansas State University, which is funded in part by NSF grants CNS-1006860, EPS-1006860, and EPS-0919443. The authors thank Prof. Chris Ackerson and Prof. Ken Knappenberger for interesting discussions regarding doped nanocluster synthesis.

17



DOI: 10.1039/C7NR05214F

Page 19 of 21


DOI: 10.1039/C7NR05214F

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. 28.

M. Zhu, C. M. Aikens, F. J. Hollander, G. C. Schatz and R. Jin, J. Am. Chem. Soc., 2008, 130, 5883-5885. M. Z. Zhu, C. M. Aikens, M. P. Hendrich, R. Gupta, H. F. Qian, G. C. Schatz and R. C. Jin, J. Am. Chem. Soc., 2009, 131, 2490-2492. C. M. Aikens, J. Phys. Chem. Lett., 2011, 2, 99-104. R. C. Jin, Nanoscale, 2015, 7, 1549-1565. G. L. Wang, T. Huang, R. W. Murray, L. Menard and R. G. Nuzzo, J. Am. Chem. Soc., 2005, 127, 812-813. R. C. Jin, Nanoscale, 2010, 2, 343-362. D. E. Jiang, Nanoscale, 2013, 5, 7149-7160. H. Hakkinen, Chem. Soc. Rev., 2008, 37, 1847-1859. P. D. Jadzinsky, G. Calero, C. J. Ackerson, D. A. Bushnell and R. D. Kornberg, Science, 2007, 318, 430-433. R. C. Jin, C. J. Zeng, M. Zhou and Y. X. Chen, Chem. Rev., 2016, 116, 10346-10413. A. Fernando, K. L. Dimuthu, M. Weerawardene, N. V. Karimova and C. M. Aikens, Chem. Rev., 2015, 115, 6112-6216. C. M. Aikens, Mol. Simul., 2012, 38, 607-614. Y. Zhu, H. Qian and R. Jin, J. Mater. Chem., 2011, 21, 6793-6799. M. Stratakis and H. Garcia, Chem. Rev., 2012, 112, 4469-4506. C. L. Yu, G. Li, S. Kumar, H. Kawasaki and R. C. Jin, J. Phys. Chem. Lett., 2013, 4, 28472852. Z. Wu, M. Wang, J. Yang, X. Zheng, W. Cai, G. Meng, H. Qian, H. Wang and R. Jin, Small, 2012, 8, 2028-2035. Y. Zhang, W. Chu, A. D. Foroushani, H. Wang, D. Li, J. Liu, C. J. Barrow, X. Wang and W. Yang, Materials, 2014, 7, 5169-5201. X. Yang, M. Yang, B. Pang, M. Vara and Y. Xia, Chem. Rev., 2015, 115, 10410-10488. Y. S. Chen, H. Choi and P. V. Kamat, J. Am. Chem. Soc., 2013, 135, 8822-8825. Y. Negishi, K. Nobusada and T. Tsukuda, J. Am. Chem. Soc., 2005, 127, 5261-5270. J. Akola, M. Walter, R. L. Whetten, H. Haekkinen and H. Groenbeck, J. Am. Chem. Soc., 2008, 130, 3756-3757. H. F. Qian, W. T. Eckenhoff, Y. Zhu, T. Pintauer and R. C. Jin, J. Am. Chem. Soc., 2010, 132, 8280-8281. O. Lopez-Acevedo, H. Tsunoyama, T. Tsukuda, H. Hakkinen and C. M. Aikens, J. Am. Chem. Soc., 2010, 132, 8210-8218. I. Chakraborty and T. Pradeep, Chem. Rev., 2017, 117, 8208-8271. H. Hakkinen, M. Walter and H. Gronbeck, J. Phys. Chem. B, 2006, 110, 9927-9931. D. E. Bergeron, P. J. Roach, A. W. Castleman, N. Jones and S. N. Khanna, Science, 2005, 307, 231-235. A. W. Castleman and S. N. Khanna, J. Phys. Chem. C, 2009, 113, 2664-2675. M. Walter, J. Akola, O. Lopez-Acevedo, P. D. Jadzinsky, G. Calero, C. J. Ackerson, R. L. Whetten, H. Gronbeck and H. Hakkinen, Proc. Natl. Acad. Sci. U.S.A., 2008, 105, 91579162.

18


REFERENCES

Nanoscale



29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.

M. W. Heaven, A. Dass, P. S. White, K. M. Holt and R. W. Murray, J. Am. Chem. Soc., 2008, 130, 3754-3755. H. F. Qian, C. Liu and R. C. Jin, Sci. China Chem., 2012, 55, 2359-2365. W. W. Xu, B. E. Zhu, X. C. Zeng and Y. Gao, Nat. Commun., 2016, 7, 13574. M. G. Taylor and G. Mpourmpakis, Nat. Commun., 2017, 8, 15988. W. W. Xu, X. C. Zeng and Y. Gao, Chem. Phys. Lett., 2017, 675, 35-39. R. C. Jin and K. Nobusada, Nano Res., 2014, 7, 285-300. C. A. Fields-Zinna, M. C. Crowe, A. Dass, J. E. F. Weaver and R. W. Murray, Langmuir, 2009, 25, 7704-7710. D. E. Jiang and S. Dai, Inorg. Chem., 2009, 48, 2720-2722. M. Walter and M. Moseler, J. Phys. Chem. C, 2009, 113, 15834-15837. K. A. Kacprzak, L. Lehtovaara, J. Akola, O. Lopez-Acevedo and H. Hakkinen, Phys. Chem. Chem. Phys., 2009, 11, 7123-7129. D. E. Jiang and R. L. Whetten, Phys. Rev. B, 2009, 80. Y. Negishi, W. Kurashige, Y. Niihori, T. Iwasa and K. Nobusada, Phys. Chem. Chem. Phys., 2010, 12, 6219-6225. K. Kwak, Q. Tang, M. Kim, D.-e. Jiang and D. Lee, J. Am. Chem. Soc., 2015, 137, 1083310840. S. B. Tian, L. W. Liao, J. Y. Yuan, C. H. Yao, J. S. Chen, J. L. Yang and Z. K. Wu, Chem. Commun., 2016, 52, 9873-9876. M. A. Tofanelli, T. W. Ni, B. D. Phillips and C. J. Ackerson, Inorg. Chem., 2016, 55, 9991001. Y. Negishi, W. Kurashige, Y. Kobayashi, S. Yamazoe, N. Kojima, M. Seto and T. Tsukuda, J. Phys. Chem. Lett., 2013, 4, 3579-3583. S. Yamazoe, W. Kurashige, K. Nobusada, Y. Negishi and T. Tsukuda, J. Phys. Chem. C, 2014, 118, 25284-25290. Y. Negishi, K. Munakata, W. Ohgake and K. Nobusada, J. Phys. Chem. Lett., 2012, 3, 2209-2214. C. Yao, J. Chen, M.-B. Li, L. Liu, J. Yang and Z. Wu, Nano Lett., 2015, 15, 1281-1287. N. Yan, L. W. Liao, J. Y. Yuan, Y. J. Lin, L. H. Weng, J. L. Yang and Z. K. Wu, Chem. Mat., 2016, 28, 8240-8247. C. Kumara, C. M. Aikens and A. Dass, J. Phys. Chem. Lett., 2014, 5, 461-466. C. Yao, Y.-j. Lin, J. Yuan, L. Liao, M. Zhu, L.-h. Weng, J. Yang and Z. Wu, J. Am. Chem. Soc., 2015, 137, 15350-15353. S. Bhat, A. Baksi, S. K. Mudedla, G. Natarajan, V. Subramanian and T. Pradeep, J. Phys. Chem. Lett., 2017, 8, 2787–2793. S. A. Miller, C. A. Fields-Zinna, R. W. Murray and A. M. Moran, J. Phys. Chem. Lett., 2010, 1, 1383-1387. H. Qian, D.-e. Jiang, G. Li, C. Gayathri, A. Das, R. R. Gil and R. Jin, J. Am. Chem. Soc., 2012, 134, 16159-16162. S. Xie, H. Tsunoyama, W. Kurashige, Y. Negishi and T. Tsukuda, Acs Catal., 2012, 2, 15191523. J. Timoshenko, A. Shivhare, R. W. J. Scott, D. Y. Lu and A. I. Frenkel, Phys. Chem. Chem. Phys., 2016, 18, 19621-19630. 19


DOI: 10.1039/C7NR05214F

Page 21 of 21



56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.

J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865-3868. J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B, 1996, 54, 16533-16539. E. Van lenthe, E. J. Baerends and J. G. Snijders, J. Chem. Phys., 1993, 99, 4597-4610. E. v. Lenthe, E. J. Baerends and J. G. Snijders, J. Chem. Phys., 1994, 101, 9783-9792. E. v. Lenthe, J. G. Snijders and E. J. Baerends, J. Chem. Phys., 1996, 105, 6505-6516. ADF2016, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands, http://www.scm.com/. C. F. Guerra, J. G. Snijders, G. te Velde and E. J. Baerends, Theor. Chem. Acc., 1998, 99, 391-403. G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. F. Guerra, S. J. A. Van Gisbergen, J. G. Snijders and T. Ziegler, J. Comput. Chem., 2001, 22, 931-967. S. L. Christensen, M. A. MacDonald, A. Chatt, P. Zhang, H. F. Qian and R. C. Jin, J. Phys. Chem. C, 2012, 116, 26932-26937. C. M. Aikens, J. Phys. Chem. C, 2008, 112, 19797-19800. C. M. Aikens, J. Phys. Chem. A, 2009, 113, 10811-10817. M. A. Tofanelli, K. Salorinne, T. W. Ni, S. Malola, B. Newell, B. Phillips, H. Hakkinen and C. J. Ackerson, Chem. Sci., 2016, 7, 1882-1890. J. Autschbach, J. Chem. Phys., 2012, 136, 150902. D. E. Jiang, M. Kuhn, Q. Tang and F. Weigend, J. Phys. Chem. Lett., 2014, 5, 3286-3289. M. S. Devadas, S. Bairu, H. F. Qian, E. Sinn, R. C. Jin and G. Ramakrishna, J. Phys. Chem. Lett., 2011, 2, 2752-2758.

20


DOI: 10.1039/C7NR05214F

Transition metal doping of Mg2FeH6--a DFT insight into synthesis and electronic structure.

Electronic Structure and Magnetic Properties of Dioxo-Bridged Diuranium Complexes with Diamond-Core Structural Motifs: A Relativistic DFT Study.

Theoretical investigation of the effects of doping on the electronic structure and thermoelectric properties of ZnO nanowires.

Thermodynamic and kinetic origins of Au25(0) nanocluster electrochemiluminescence.

Effects of 3d transition-metal doping on electronic and magnetic properties of MoS₂ nanoribbons.

Doping of Zinc Oxide with Selected First Row Transition Metals for Photocatalytic Applications.

DFT and TD-DFT assessment of the structural and optoelectronic properties of an organic-Ag14 nanocluster.

Cob(II)alamin: Relativistic DFT Analysis of the EPR Parameters.

Investigation of VO-salophen complexes electronic structure.

Communication: a density functional investigation of structure-property evolution in the tetrakis hexahedral C4Al14 nanocluster.

Electronic structure of cesium butyratouranylate(VI) as derived from DFT-assisted powder X-ray diffraction data.

Electronic structure investigation of the evanescent AtO(+) ion.

Understanding natural semiquinone radicals--multifrequency EPR and relativistic DFT studies of the structure of Hg(II) complexes.

DFT simulation, quantum chemical electronic structure, spectroscopic and structure-activity investigations of 2-benzothiazole acetonitrile.

Single-Crystal and Electronic Structure of a 1.3 nm Indium Phosphide Nanocluster.

Crystal structure and electronic properties of a thiolate-protected Au24 nanocluster.

Evolution of the electronic band structure of twisted bilayer graphene upon doping.

TDDFT Study of the Electronic Structure of Axially Pyridine Coordinated Metallocorroles.

Basis set error estimation for DFT calculations of electronic g-tensors for transition metal complexes.

Continuously tunable electronic structure of transition metal dichalcogenides superlattices.

DFT analysis on the molecular structure, vibrational and electronic spectra of 2-(cyclohexylamino)ethanesulfonic acid.

Structure and Electronic Properties of Transition Metal Doped Kaolinite Nanoclay.

The influence of doping with transition metal ions on the structure and magnetic properties of zinc oxide thin films.

Transition from Coherent to Stochastic electron heating in ultrashort relativistic laser interaction with structured targets.