The chiroptical signature of achiral metal clusters induced by dissymmetric adsorbates

Abstract

Using a dissymmetrically-perturbed particle-in-a-box model, we demonstrate that the induced optical activity of chiral monolayer protected clusters, such as Whetten’s Au₂₈(SG)₁₆ glutathione-passivated gold nanoclusters (J. Phys. Chem. B, 2000, 104, 2630–2641), could arise from symmetric metal cores perturbed by a dissymmetric or chiral field originating from the adsorbates. This finding implies that the electronic states of the nanocluster core are chiral, yet the lattice geometries of these cores need not be geometrically distorted by the chiral adsorbates. Based on simple chiral monolayer protected cluster models, we rationalize how the adsorption pattern of the tethering sulfur atoms has a substantial effect on the induced CD in the NIR spectral region, and we show how the chiral image charge produced in the core provides a convenient means of visualizing dissymmetric perturbations to the achiral gold nanocluster core.

---

Introduction

The interactions between metal surfaces or nanoclusters and chiral adsorbates are the subject of intense interest for applications in nanotechnology, molecular biosensing, spintronics, and catalysis.^1–8,11,12 Passivation of gold or silver nanoclusters with chiral admolecules results in unique electronic and chiroptic response properties that are unlike those of the component parts.^6,7 For example, small nanoclusters (<1.5 nm) with chiral adsorbates exhibit a giant induced CD spectrum.⁶ The chiroptic signatures vary with the chiral admolecule, the nanocluster core size, and the cluster composition.^6–9 One explanation proposed for the distinctive chiroptic response of chiral monolayer protected clusters⁶ (MPCs) is an adsorbate-induced distorted chiral core lattice structure, and its corresponding chiral electronic structure.^9,10 We explore the hypothesis that the CD response may arise from a symmetric core lattice with an electronic structure that is dissymmetrically perturbed by the chiral adsorbate. We show that the CD signal induced in this way accounts for the experimental observations.

Method

We used a charge-perturbed particle-in-a-box model (PIB)^13,14 to explore the dissymmetrically perturbed symmetric sub-nm metal core.^6,10,12 The Au_n core was modeled with non-interacting electrons confined to a cubic box, and the surrounding adsorbates were described using point charges. The point charge electrostatic perturbation was
Here q_i is the ith perturbing point-charge at (x_i,y_i,z_i) and the sum is performed over all m charges. A schematic representation of this model is shown in Fig. 1. Chiral adsorbates or dissymmetrically placed achiral adsorbates induce a chiral perturbation on the core making it optically active, whereas symmetric achiral adsorbates or symmetric arrangements result in symmetric perturbations and are thus optically inactive. In this study we focussed primarily on dissymmetric perturbations.

Fig. 1 Schematic representation of (a) a molecular adsorbate and Au cluster (i.e. Au₁₄(R-methylthiirane) represented by (b) a system of point charges.

With this model, we determined the first-order response of the cluster electronic states in a perturbation theory framework: an achiral cluster in the presence of a chiral perturbation. Then, we computed the electric and magnetic transition moments ([small mu, Greek, vector]_n0 and m⃑_n0, respectively) for the PIB states. From these transitions moments we calculated the rotational strength R_n0, given by

and the oscillator strength F_n0,
These quantities are required for prediction of the CD and absorption spectra of the assembly, enabling us to explore the influence of chiral adsorption patterns and chiral adsorbates with achiral adsorption patterns on CD spectra. We also computed the polarized electron-density (i.e., the induced chiral image charge) inside the box.¹⁴

Results and discussion

The chiroptical response in the PIB model changes with the handedness of the perturbation. Negative point charges (2, 4, or 8 charges in total) of magnitude −0.8e placed in dissymmetric arrangements produced chiral image charges in the box. The induced absorption and CD spectra are shown in Fig. 2 for charges chosen to represent achiral methylthiolate adsorbates placed dissymmetrically on the surface of the cluster.¹⁹ Expected quantum size effects (i.e. red-shifts upon box size increase corresponding to Au₁₄ to Au₂₈ cluster size changes) are observed in the electronic spectra with similar dissymmetric point-charge arrangements. Similarly, in the case of symmetrically placed point-charges, symmetric or achiral image charges result. In this case, the absorption spectrum is identical with respect to the surface coverage of adsorbates and core size, although the core is optically inactive (i.e. CD = 0). Both induced image-charge in the core arising from the symmetric/achiral case, and demonstration of the distance-dependence of CD arising from dissymmetric/chiral adsorption patterns appear in the supporting information.†

Fig. 2 Representative chiral point-charge systems and their induced chiral image charges for cores of Au₁₄ with 2, 4 and 8 negative point-charges (a–c, respectively) and similarly with an Au₂₈ core (d–f, respectively). The computed electronic absorption (top) and CD spectra (bottom) are shown to the right. As surface coverage increases, CD spectra change significantly more than absorption spectra. Similarly, as cluster size increases, the transition frequency drops, as seen experimentally.⁶ Red point-charges correspond to negative point-charges, red surfaces in the core represent regions of increased electron density and blue indicates regions of decreased electron density.

Next, we placed chiral adsorbates (for the model chiral structures Au₁₄(MTI)₆, Au₂₈(MTI)₆, Au₁₄(SG)₆, Au₂₈(SG)₆, where MTI = R-methylthiirane, and SG = glutathione) on each face of the box in an achiral pattern (Fig. 3) (the ligands were placed in a topmost position in which the S-atom tethered to the surface was normal to each face at a separation distance of 2.48 Å). The induced chiral image charges, and the computed CD spectra are qualitatively consistent with those seen experimentally.⁶

Fig. 3 (a) Au₂₈(R-methylthiirane)₆ and (b) Au₂₈(glutathione)₆ using partial charges for the admolecules and their image charges. Top right are the computed electronic absorption and CD spectra (bottom). As the system size increases, the CD spectra change substantially: varying ligands produce distinctive CD spectra. Similar quantum size effects are observed as in Fig. 2 upon increasing the metal core size. The color scheme used is the same as in Fig. 2.

We also used extended-Hückel and DFT methods to explore the induced chiral image charges and CD spectra arising from the core. The core image charges computed at the extended-Hückel level are similar to those found with the PIB model.¹⁶Fig. 4 shows the results for the Au₁₄(MTI)₆ monolayer protected cluster.

Fig. 4 Induced chiral-image charge using (a) extended-Hückel theory and (b) the chirally perturbed PIB model for the (c) Au₁₄(MTI)₆ system. Negatively charged sulfur groups (yellow) dissymmetrically perturb electron density in the box leaving positive image charges in the central regions. Negative charge density moves to the box edges and corners, and appears to be mostly due to the adsorbing atoms. The color scheme for the chiral image charge in the core is as in Fig. 2.

The CD spectra derived from time-dependent density functional theory (RI-TD-DFT,^17b 10 RPA singlet excitations, BP-86 functional and an effective core potential^17c for gold where appropriate, otherwise the SV(P) basis was used) were also qualitatively similar to the PIB model (Fig. 5).^15,17 In addition to qualitative differences among the models, the PIB model does not permit charge-transfer between admolecules and core cluster atoms. Nonetheless, the simple model predicts the same CD signature and associated Cotton effect.

Fig. 5 A comparison of TD-DFT single-excitation RPA calculation for Au₁₄(R-methylthiirane) in the topmost face-centered position tethered via sulfur. The three curves indicate: (a) TD-DFT including explicit adsorbate (BP-86 functional, ECP for Au and SV(P) basis for adsorbate, 10 RPA singlet excitations); (b) TD-DFT (BP-86, ECP for Au, 10 RPA singlet excitations) using a point charge adsorbate model; and (c) the chirally perturbed PIB calculation. Magnitudes are similar and, moreover, the qualitative shape and sign of the CD bands match. Details are provided in the supporting information.†

In summary, the asymmetrically-perturbed PIB model describes one possible origin of induced CD for monolayer protected clusters composed of chiral molecules.¹⁵ Trends in the electronic transition frequencies with cluster size also agree quantitatively with experiment.⁶ From our simple models it appears that chiral adsorption patterns induce both considerably larger CD and chiral image charges in the metal core. Similarly, inherently chiral adsorbates placed in achiral adsorption patterns also induce CD and chiral image charges in the core. Thus, the nature of the distance-dependent dissymmetric electrostatic perturbation, influenced by both the dissymmetry of the adsorption pattern and the centers of point asymmetry on the ligands (i.e. chiral ligands), is effective in inducing chirality even in a symmetric metal core.

Similar experimental studies of enantiomer specific second-harmonic generation of chiral adsorbates on metal surfaces reveal that a potential prerequisite for inducing chirality on the electronic states of metal surfaces is that the adsorption places a chiral footprint on the surface that is directly related to the proximity of the adsorbate to the surface.¹⁸ Likewise, in our study the distance-dependent adsorbate perturbation produces a significantly enhanced effect on the chirality of the induced image charge and on the magnitude of the CD spectra in the case of chiral adsorption patterns, more so than chiral adsorbates placed “symmetrically”. Nonetheless, achiral adsorption patterns of chiral adsorbates also induce substantial changes in CD spectra.

Whetten and coworkers^6,10 have also suggested that dissymmetric structural distortions of the core could induce optical activity in analogous systems. Additionally, Whetten and coworkers suggested that chiral arrangements were expected to have a large effect on the gold electronic structure and on the resulting CD response, perhaps more so than the molecular chirality of the adsorbate itself. Large atomic charges of adsorbate atoms (i.e. sulfur) and reduced distances between core and adsorbate enhance the perturbation. Although inherently chiral adsorbates can induce CD spectra in an achiral metal core, the largest contribution to this effect, via the adsorption patterns, is in turn a function of the chirality and intermolecular interactions of the adsorbates; the two effects are coupled.⁷ As a material design principle, the largest chiral image charges and CD responses are expected to arise from chiral adsorbates with the largest surface coverage and charge magnitude, smallest distance to surface, and maximally dissymmetric adsorption pattern.

Concluding remarks

Induced chiroptic signatures in chiral monolayer protected clusters (<1.5 nm) can arise from chiral electronic state perturbations (described here), chiral nanocluster core distortions,^9,10 or both. Interestingly, achiral metal cores may exhibit induced CD spectra as a consequence of a dissymmetric distribution of adsorbates. In an ensemble of cluster-adsorbate species, the net influence of chiral adsorption of achiral adsorbates would average to zero. However, chiral monolayer protected clusters with enantiomerically-enriched adsorbate species can induce dissymmetrically perturbed electronic states of the gold cluster core, producing chiroptic signatures that would not average to zero and thus exhibit optical activity in the IR/NIR spectral region.

Furthermore, we have demonstrated that the dissymmetrically-perturbed PIB model is a viable simple tool to evaluate and interpret chiral MPC structure-optical activity relationships and visualize chiral image charge induction.

Acknowledgements

This work was supported in part by the National Science Foundation (CHE-0078944). M.-R. G. thanks Duke Structural Biology and Biophysics and the National Institutes of Health/National Institute of General Medical Sciences (NIH/NIGMS 5T32-GM08487) for a graduate fellowship, and also thanks Michael A. Peterson for parallelization and scripting assistance. G. Z. gratefully acknowledges a postdoctoral fellowship by the Swiss National Science Foundation. R. N. and D. H. W. acknowledge partial support by the US–Israel Binational Science Foundation. Additional support at Duke from the NEDO (Japan) and Keck Foundations is gratefully acknowledged.

References

1. (a) K. Ray, S. P. Ananthavel, D. H. Waldeck and R. Naaman, Science, 1999, 283, 814–816; (b) I. Carmeli, V. Skakalova, R. Naaman and Z. Vager, Angew. Chem., Int. Ed., 2002, 41, 761–764.
2. J. D. Aiken and R. G. Finke, J. Mol. Catal. A: Chem., 1999, 145, 1–44.
3. Z. Li, R. C. Jin and C. A. Mirkin, Nucleic Acids Res., 2003, 30, 1558–1562.
4. J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger and G. C. Schatz, J. Am. Chem. Soc., 2000, 122, 4640–4650.
5. R. L. Whetten, M. N. Shafigullin, J. T. Khoury, T. G. Schaaff, I. Vezmar, M. M. Alvarez and A. Wilkinson, Acc. Chem. Res., 1999, 32, 397–406.
6. T. G. Schaaff and R. L. Whetten, J. Phys. Chem. B, 2000, 104, 2630–2641.
7. T. Li, H. G. Park, H.-S. Lee and S.-H. Choi, Nanotechnology, 2004, 15, S660–S663.
8. (a) S. Link, A. Beeby, S. Fitzgerald, M. A. El-Sayed, T. G. Schaaff and R. L. Whetten, J. Phys. Chem. B, 2002, 106, 3410–3415; (b) T. P. Bigioni, R. L. Whetten and Ö. Dag, J. Phys. Chem. B, 2000, 104, 6983–6986.
9. C. E. Román-Velázquez, C. Noguez and I. L. Garzón, J. Phys. Chem. B, 2003, 107, 12035–12038.
10. There are three proposed explanations for the chiroptic signatures of chiral-monolayer protected clusters;⁶ The first explanation assumes a chiral core due to thiol-passivated lattice distortions (Garzon and coworkers described the expected CD effects⁹). Two additional explanations include an achiral nanoparticle core in which chirality may be induced via a chiral adsorption pattern of ligands or chiral adsorbates with achiral adsorption patterns.
11. (a) J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger and G. C. Schatz, J. Am. Chem. Soc., 2000, 122, 4640–4650; (b) T. Jensen, L. Kelly, A. Lazarides and G. C. Schatz, J. Cluster Sci., 2004, 10, 295–317; (c) A. A. Lazarides and G. C. Schatz, J. Phys. Chem. B, 2000, 104, 460–467.
12. J. Zheng, C. W. Zhang and R. M. Dickson, Phys. Rev. Lett., 2004, 93 , Art. No. 077402.
13. Semi-classical electrodynamic models and Mie theory (or Mie Scattering theory) are typically used to interpret and predict plasmon absorption bands of nanoparticles (≫1.5 nm) with biomolecular adsorbates.¹¹ In this spirit Garzon and coworkers used the discrete dipole approximation (DDA) to simulate CD spectra in sub-nanometer chiral clusters.⁹ However, the electronic absorption of small gold monolayer protected clusters (i.e. <1.5 nm size regime, for instance Au₅–Au₃₁) had previously been experimentally obtained and modeled by jellium approaches where it was shown that the collective plasmon band did not dominate the absorption characteristics, with distinct quantum size effects that suggested MPCs of this kind behave as “multielectron artificial atoms”.^6,12 Similarly, using first-order perturbation theory and a PIB approach we have chosen to evaluate the chiroptic signature of model MPCs under the assumption of a symmetric core, which thus represents the difference of an achiral metal core in the presence of a dissymmetric perturbation¹².
14. (a) The chiral image charge is the difference of the unperturbed electron density isosurface and the chirally-perturbed electron density isosurface that corresponds to a net displacement of approximately 1 × 10⁻⁵e. (b) The CPPIB model was implemented using gridMathemathica, and the chiral image charge surfaces were rendered using Math3GL viewer and Pov-ray. (c) gridMathemathica v. 5.0, Wolfram Inc. 2005. (d) Math3GL Viewer for Mathematica, Version 3.1beta copyright by Jens-Peer Kuska 1997–2003. (e) Pov-ray for MAC OS Shell 3.6.0 Build 0197, July 24 2004: copyright Persistence of Vision Raytracer Pty. Ltd. 2003–2004.
15. Performing calculations on conformationally averaged ligand geometries may allow a more direct comparison to experimental spectra. In these calculations, we have chosen ad hoc a given low energy conformation, however, none of these results are Boltzmann weighted to obtain thermodynamically averaged ensembles.
16. (a) R. Hoffmann, J. Chem. Phys., 1963, 39, 1397; (b) HyperChem 6.0 program, Hypercube Inc. Gainesville, USA.
17. (a) R. Ahlrichs, M. Bär, H.-P. Baron, R. Bauernschmitt, S. Böcker, M. Ehrig, K. Eichkorn, S. Elliott, F. Furche, F. Haase, M. Häser, H. Horn, C. Huber, U. Huniar, M. Kattannek, C. Kolmel, M. Kollwitz, K. May, C. Ochsenfeld, H. Öhm, A. Schäfer, U. Schneider, O. Treutler, M. von Arnim, F. Weigend, P. Weis and H. Weiss, Turbomole (version 5.6), Universität Karlsruhe: Karlsruhe, Germany, 2002; (b) These calculations were performed using TurboMole 5.6. Additional information on Au extended-core potential, functional, auxiliary basis sets, computation of CD using RPA single-excitations and the adsorbate DFT description is available in the supporting information; (c) K. Eichkorn, O. Treutler, H. Oehm, M. Haeser and R. Ahlrichs, Chem. Phys. Lett., 1995, 242, 652–660; (d) D. Andrae, U. Haeussermann, M. Dolg, H. Stoll and H. Preuss, Theor. Chim. Acta, 1990, 77, 123.
18. A. Mulligan, I. Lane, G. B. D. Rousseau, S. M. Johnston, D. Lennon and M. Kadodwala, Angew. Chem., Int. Ed., 2005, 44, 1830–1.
19. The charge of −0.8e was based on a Mulliken population analysis performed at the extended-Hückel level on the methylthiolate placed with 2.48 Å separation distance from the core cluster in the topmost position above the face-centered gold atom, subsequently coarse-grained to a single charge.

---

Footnote

† Electronic supplementary information (ESI) available: Details for chirally-perturbed PIB model, the extended-Hückel image charge calculations and time-dependent density functional theory (RPA singlet excitations) circular dichroism spectral simulation are provided in the supporting information, and are used in this context as benchmarking/validation tools for the chirally perturbed PIB model. Please contact michael.goldsmith@duke.edu for CPPIB gridMathematica source-code. See DOI: 10.1039/b511563a

---