Chiral Au25 Nanospheres and Nanorods: Synthesis and Insight into

LETTER pubs.acs.org/NanoLett

Chiral Au25 Nanospheres and Nanorods: Synthesis and Insight into the Origin of Chirality Manzhou Zhu,*,† Huifeng Qian,‡ Xiangming Meng,† Shenshen Jin,† Zhikun Wu,‡ and Rongchao Jin*,‡ † ‡

Department of Chemistry, Anhui University, Hefei, Anhui 230039, People's Republic of China Department of Chemistry, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States

bS Supporting Information ABSTRACT: Chirality in nanoparticles is an intriguing phenomenon. Herein, we have devised a well-defined gold nanoparticle system for investigating the origin of chirality in nanoparticles. We have designed chiral thiols (R- and S-isomers) and synthesized chiral gold nanoparticles composed of 25 gold atoms and 18 ligands, referred to as Au25(pet*)18, where pet* represents chirally modified phenylethylthiolate
SCH2C*H(CH3)Ph at the 2-position. These optically active nanoparticles are close analogues of the optically nonactive phenylethylthioaltecapped Au25(pet)18 nanoparticles, and the latter’s crystal structure is known. On the basis of the atomic and electronic structures of these well-defined Au25 nanoparticles, we have explicitly revealed that the ligands and surface gold atoms of Au25(pet*)18 play a critical role in effecting the circular dichroism responses from the nanoparticles. Similar effects are also observed in chiral Au25 rods. The mixing of electronic states of ligands with those of surface gold atoms constitutes the fundamental origin of chirality in such nanoparticles. KEYWORDS: Chirality, gold, nanoparticles, Au25

C

hirality (or handedness) is one of the intriguing and inspiring phenomena in nature. An early observation of chirality dates back to 1848, when Pasteur noticed that the (+)-tartaric acid was consumed more rapidly by microorganisms than the (-)enantiomer of tartaric acid.1 He also found that tartrate crystals had two non-superimposable forms in a mirror image relationship just as are the left and right hands. To date, apart from small molecules, a variety of chiral structures have been discovered, such as DNA, RNA, and proteins, as well as nanoparticles. In biology, chirality exerts profound effects on the molecular recognition and interaction. In catalysis, the design of chiral catalysts has led to valuable products such as chiral drugs. Chiral metal nanoparticles, such as gold and silver, are particularly interesting as bulk Au and Ag are of face-centered cubic (fcc) structure and hence are achiral. Chiral nanoparticles hold great promise in photonics, catalysis, and many other potential applications.2
6 From the viewpoint of basic science, an interesting question pertains to the fundamental origin of chirality in chiral metal nanoparticles. Chirality has been found in a variety of gold nanoparticles.7
22 Whetten and co-workers reported that ultrasmall Au nanoparticles (∼1 nm diameter) protected by chiral L-glutathione ligands exhibited strong optical circular dichroism (CD) activity.7 An interesting aspect is that the observed CD signals, apart from the inherent CD response from the ligand itself (whose peak position is at 215 nm for glutathione), are also manifested in the metal-based electronic transitions in the visible region. The latter is quite r 2011 American Chemical Society

intriguing, as it implies some kind of electronic coupling effect.8 Whetten and co-worker proposed three possible mechanisms: (i) the structure of the gold core is inherently chiral; (ii) an inherently achiral core but the adsorption of the
SG thiolates on the particle surface results in a chiral pattern of ligand interactions with the achiral core; (iii) chiral elements of the
SG adsorbates induce optical activity in the core electronic structure, even though neither the adsorption pattern nor the core structure is chiral.7 Theoretical calculations indicate that the chirality can arise from an intrinsically chiral gold core,10
13 as well as from an achiral core if the latter is placed in a chiral environment.8,14,15 Recent work has indeed discovered chiral arrangements of metal atoms in nanoparticles, such as the Au102(SR)44 and Au38(SR)24 nanoparticles.16,17 But chirality in many other cases of gold nanoparticles are still not well understood. Experimentally, ligand exchange on the particle surface with chiral ligands often led to chiroptical signals apart from the original response of ligands, but the mechanism was not clear since the metal core structure being chiral or achiral was not known.18
22 Another complication is that in some cases ligand exchange (e.g., thiolate to phosphine, or vice versa) may result in alteration of the metal core size and/or structure.23 Nevertheless, experimental work by a number of groups has provided important information on Received: July 6, 2011 Revised: August 4, 2011 Published: August 11, 2011 3963

dx.doi.org/10.1021/nl202288j | Nano Lett. 2011, 11, 3963–3969

Nano Letters

LETTER

Scheme 1. One-Pot Synthesis of Chiral Au25(pet*)18 Nanoparticles (pet* = SCH2C*H(Me)Ph)

nanoparticle charlity, which has led to several plausible mechanisms, including the chiral footprint model18,19 and the dissymmertric (i.e., asymmetric) field effect.24,25 B€urgi and co-workers reported the synthesis of Au nanoparticles covered with N-isobutyrylL-cysteine and N-isobutyryl-D-cysteine and found that the electrophoresis-separated AunLm nanoclusters (L = ligands) exhibited strong optical activity in the metal-based transitions;19 they attributed the origin of the optical activity to a chiral footprint on the metal core imparted by the adsorbed thiolates. Yao et al.24 achieved chiral phase transfer of aqueous racemic penicillamineprotected gold nanoclusters into chloroform by hydrophobizing the anionic nanocluster with a chiral ephedrinium cation, and the as-transferred product exhibited interesting CD signals in the metal-based electronic transitions. They proposed that the origin of CD roots in the ligand dissymmetric field due to the surface stereostructures.25 For Ag nanoclusters, Kitaev et al. recently observed intense CD signals from Ag nanoclusters prepared by using different combinations of captopril and glutathione.26 It would be interesting to see if such chiroptical behavior can be observed in other types of Ag nanoclusters.27 All the above scenarios pertain to ultrasmall Au or Ag nanoparticles that are nonplasmonic. Chiroptical behavior has also been observed in plasmonic nanoparticles in the surface plasmon resonance region (i.e., ∼520 nm for Au and ∼400 nm for Ag) caused by chiral molecules such as peptides and DNA,28,29 which is attributed to dissipative currents in the plasmonic particle induced by the molecular dipole.28 Although chirality in gold nanoparticles has been reported extensively and some possible mechanisms have been proposed, the fundamental origin of the chiroptical activity of such nanoparticles in the metal-based electronic transitions is still unclear, which is largely hurdled by the unknown atomic structure of the nanoparticles. As for the chiral Au102 and Au38 nanoparticles, their atomic structures are known, but they exist in racemic mixtures and have not been separated into individual isomers thus far.16,17 Herein we are motivated to devise a well-defined, atomically monodisperse chiral nanoparticle system for a detailed study on the origin of chirality. We chose the 25-gold-atom nanoparticles (including Au25 nanopheres and nanorods) as a model system to gain insight into the origin of chirality in gold nanoparticles. The spherical Au25 nanoparticles are capped by phenylethylthiolate (abbreviated as pet). Previous work has shown that pet-capped Au25(pet)18 has no CD activity,30 although the gluathionate-capped Au25(SG)18 was reported to show strong CD signals.7,30 It should be clarified

Figure 1. UV
vis absorption spectra of Au25 nanoparticles capped by R- and S-pet*, respectively.

that the null CD signals from Au25(pet)18 is not due to a racemic mixture, because if racemic Au25(pet)18 nanoparticles were present, there would be a pair of mirror-imaged particles in the crystal structure, but we did not find such a pair in the anionic or neutral Au25 structure.31,32 In contrast, Au38(pet)24 clusters are racemic, and a pair of mirror-imaged particles was indeed found in the crystal structure.17 Furthermore, NMR analysis of Au25(pet)18 did not show any chiral splittings in the 1H signals of Au25(pet)18, whereas racemic Au38(pet)24 shows distinct NMR splittings. Thus, Au25(pet)18 is not racemic, rather it is achiral. B€urgi and co-workers23 prepared chiral Au25 nanoclusters via ligand exchange with R/S-1,10 -binaphthyl-2,20 -dithiol and N-isobutyrylL-cysteine/N-isobutyryl-D-cysteine, but the metal core size was unfortunately altered in these reactions. To utilize the pet-capped Au25 system, we have devised R- and S-2-phenylpropane-1-thiols (abbreviated as pet* hereafter), which are close analogues of the phenylethylthiol (pet) ligand except the chiral center at the 2-position. With such chiral pet* thiols, we have successfully synthesized chiral Au25(pet*)18 nanoparticles with the Au25 core structure inherited from the pet-capped Au25(pet)18 counterpart. Our results explicitly show that the chirality in Au25 is not caused by the metal core, but by the surface ligands and surface gold atoms of the Au25 nanoparticles. Similar effects are also observed in the chiral Au25 rods. Synthesis and Characterization of Chiral Au Nanoparticles. The synthesis of thiolate-capped Au25 nanoparticles follows a kinetically controlled method reported previously.33,34 The R- or S-pet* protected Au25 nanoparticles, as well as racemic pet*-protected Au25 nanoparticles were synthesized using the same method (Scheme 1). Details of the synthesis of chiral Ror S-pet* and of nanoparticles are provided in the Supporting Information. The reaction of Au25 nanoparticle synthesis involves an interesting size focusing process spontaneously occurred during a prolonged process at room temperature.34,35 Right after the addition of NaBH4 to reduce the Au(I) intermediate to Au(0), the product consisted of polydisperse Aun(SR)m nanoparticles, which exhibited a featureless decaying UV
vis absorption spectrum. After ∼40 h of aging, the final crude product (without purification) already showed distinct absorption bands at 670, 450, and 400 nm that are characteristic of Au25(SR)18 nanoparticles (Figure 1), indicating high purity of the as-prepared chiral 3964

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969

Nano Letters

LETTER

Figure 2. MALDI mass spectra of the R- and S-Au25 nanoparticles (panel A and B, respectively. Observed, m/z = 7643; expected, MW = 7643). Note that the less intense peak at lower m/z 6409 is a fragment of the nanoparticle caused by MALDI-MS.

Au25(pet*)18 nanoparticles. The nanoparticles were precipitated by adding ethanol in order to remove excess thiol and other byproducts. The Au25 nanoparticles protected by R-pet*, S-pet*, and racemic thiolates, respectively, show superimposable UV
vis absorption spectra, and these spectra are also identical to that of pet-capped Au25,34 demonstrating that the ligand, being chiral or not (i.e., R- or S-pet* vs achiral pet), does not affect the ordinary optical absorption spectra of Au25 nanoparticles. Mass spectrometry analysis by matrix-assisted laser desorption ionization (MALDI) confirms that the as-prepared nanoparticles are Au25(SR)18 with atomic monodispersity (Figure 2). The MALDI mass spectra of Au25 nanoparticles protected by R- and S-pet*, respectively, are essentially identical. Their isotope patterns show a spacing of unity (Figure 2, insets), indicating the ionized nanoparticle bears
1 charge (negative mode used in the MS analysis). Thus, the peak centered at 7643 m/z corresponds to the molecular mass of Au25(pet*)18 nanoparticles, which agrees well with the calculated mass (7643 Da). It is worth noting that a fragment of Au25(pet*)18 is also observed at m/z = 6251, which is assigned to Au21(pet*)14 (i.e., after the loss of a Au4(pet*)4 unit). This fragmentation pathway is the same as the case of nonchiral Au25(pet)18.36 The MS results of these R-/S-Au25 nanoparticles, together with their UV
vis absorption spectra being identical to that of Au25(pet)18, indicate that these chiral Au25(pet*)18 nanoparticles share a common Au25 core structure as that of Au25(pet)18.31,32,37 Thermogravimetric analysis further confirms the composition and purity of the as-prepared chiral nanoparticles (Supporting Information Figure S1). Weight losses of 38.9, 39.02, and 38.9% were observed for the nanoparticles capped by R-, S-, and racemic pet* ligands, respectively. These results agree well with the expected organic weight percentage (39.27%) for the Au25(pet*)18
TOA+ formula of the nanoparticles. Note that these nanoparticles bear
1 charge and the counterion is tetraoctylammonium (TOA+).31,37 NMR analysis (Supporting Information Figure S2) also shows identical spectra for the R- and S-Au25 nanoparticles. TOA+ signals were observed. Quantification of peak areas implies that the ratio of TOA+ to pet* is 1:18, consistent with the formula. On the basis of these results, the as-prepared chiral Au25 nanoparticles are anions and bear one equivalent of TOA+. Chiroptical Activity of Gold Nanoparticles. The most interesting aspect of the R- and S-Au25 nanoparticles is their chiroptical properties. Figure 3 shows the CD spectra of these nanoparticles. Intense bands at 310 and 425 nm are observed in the CD spectra with positive sign for the peaks of the R-Au25

Figure 3. The CD spectra of the R- and S-Au25 nanoparticles (y-axis (θ) is in units of mdeg).

isomer and negative for the S-Au25 isomer. The observed CD signals are not from the R- or S-pet* ligands, as the latter’s signals are in the vacuum UV region, rather than in the measured wavelength region. The R- and S-Au25 nanoparticles exhibit a precise mirror image relationship (so-called Cotton effect). The results of chiral Au25(pet*)18 nanoparticles are particularly interesting when compared with their “parent” Au25(pet)18 particle with the only difference in the chiral carbon at the 2-position; the latter particle is not optically active. These results demonstrate that the incorporation of a chiral carbon center in the ligand (i.e.,
SCH2
C*H(CH3)
Ph) can indeed strongly induce chiral character in the metal-based electronic transitions of Au25, while the ligand itself shows no absorption nor CD signals in this wavelength range. Of note, we have measured the CD spectra of the Au(I)(pet*) complexes before reduction into Au25(pet*)18. As shown in the Supporting Information Figure S3, Au(I)(pet*) complexes show a very weak band at 350 nm (with highly concentrated samples, positive for the R-complex and negative for the S-complex). This feature is different from the CD spectra of Au25(pet*)18 nanoparticles; the latter exhibit bands at 310 and 425 nm. It is worthwhile to compare the CD spectra of the chiral Au25(pet*)18 with glutathionate-capped Au25(SG)18 nanoparticles (where, the
G group bears chiral carbon centers as well).7,30 Note that their normal absorption spectra are superimposable,30 indicating that the electronic transitions are not affected by chirality of the
SG or
pet* ligand; however, their CD bands 3965

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969

Nano Letters

Figure 4. Core
shell-type geometric and electronic structures of Au25(SR)18 nanoparticles. In (A), only 6 ligands are shown in full (for clarity), while the remaining 12 ligands are only represented by sulfur atoms.

show some interesting differences (c.f. Figure 3 and Supporting Information Figure S4). Au25(SG)18 with L-glutathione shows two prominent positive bands at 375 and 500 nm as well as two weak negative bands at 310 and 425 nm. The latter two negative bands seems parallel with the two negative bands observed in S-Au25(pet*)18, but the two positive CD bands are not observed in R- or S-Au25. Furthermore, if one compares the CD spectra of Au25(SG)18, R- and S-Au25(pet*)18 with their normal nondistinguishable absorption spectra (all showing ordinary absorption peaks at λ = 400, 450, and 670 nm), it is interesting to notice that the CD peak wavelengths are not exactly coincident with the normal absorption peaks (i.e., with the 400 and 450 nm peaks). This implies some subtle effects of the types of ligands on the CD spectra of Au25 nanoparticles, though no effects on their normal absorption spectra. This dependence of CD signals on the ligand type is interesting, as the CD spectral features may serve as spectroscopic “fingerprints” for the identification of different types of chiral Au25 nanoparticles. Taken together, the above CD spectral features clearly indicate that the chiral ligands strongly affect the CD spectra of the nanoparticles in a subtle way, although their normal absorption spectra are not affected. CD spectroscopy measures the difference in absorbance of left- and right-circularly polarized light by the analyte, circular dichroism can only occur within a normal absorption band manifested in the normal absorption spectrum

LETTER

(measured by UV
vis spectroscopy). Ordinary UV
vis measures the absorbance of unpolarized (i.e., isotropic) light by the sample. In the case of Au25(pet*)18, the 420 nm CD band locates on the blue side of the normal absorption (λpeak∼450 nm) band. The CD spectrum of Au25(SG)18 nanoparticles seems to involve a splitting (i.e., positive and negative CD sub-bands) corresponding to the normal absorption band at λpeak∼450 nm. These effects need to be explained by performing theoretical calculations.13
15 Below we discuss the origin of chirality in such gold nanoparticles. Origin of Chirality in Gold Nanoparticles. Generally speaking, the observation of CD signals indicates either an inherently asymmetric chromophore or a symmetric chromophore in an asymmetric environment. In the R- and S-Au25(pet*)18, as well as chiral Au25(SG)18 nanoparticles, the observation of CD signals corresponding to the metal-based electronic transitions is quite interesting, since the intrinsic CD signal of HS-G is at λ ∼ 210 nm and HS
CH2C*H(CH3)Ph is far below 200 nm. The metal-based electronic transitions could originate in the Au25 metal core itself or charge transfer between the metal core and surface thiolate ligands. In either case, the absorbing center should have different molar extinction coefficients (i.e., εL and εR) for the left- and right-polarized light, and this unequality results in unequal magnitudes of the L- and R-rotating electric field components of light (i.e., electric dipole-allowed absorption transitions), thus, giving rise to an ellipticity (θ), which relates to the differential absorbance (ΔA) by ΔS = (εL
εR)Cl = θ/32.98, where, C is the concentration of the analyte, and l is the optical beam path. As for the electronic transitions of the Au25 nanoparticles, theoretical calculations have revealed a quantized sp band (as opposed to the quasicontinuous sp band in bulk gold or Au nanocrystals) (Figure 4);31 note that there are two d-bands (upper and lower ones)38 and the splitting is caused by the geometric core
shell structure of Au13/Au12(SR)18 (Figure 4A). The Au25 nanoparticle consists of an icosahedral Au13 core (or kernel) plus the exterior Au12 shell with 18 thiolate ligands. The quantized sp band includes the HOMO, LUMO, and LUMO+1, etc., and each is comprised almost exclusively of atomic orbital (AO) contributions from the 13 Au atoms in the icosahedral core rather than the 12 exterior Au atoms,31 whereas the upper d-band (Figure 4B) is mainly contributed by the 12 shell Au atoms and the thiolate ligands; note that the lower set of d-band lies deeper (not shown in Figure 4B) and is principally contributed by the Au13 core (corresponding to the traditional d-band).38 DFT simulations have identified that the 670 nm peak (Figure 1) is due to a LUMO r HOMO transition, which is essentially an intraband transition (within the sp band, labeled a in Figure 4B). The 450 nm peak (Figure 1) arises from mixed intraband and interband (sp r d) transitions (labeled b in Figure 4B). Peaks at even shorter wavelength (e.g., the less prominent peaks at λpeak ∼ 400 nm and ∼320 nm) arise principally from sp r d interband transitions. Since the HOMO, LUMO, and LUMO+1 are constructed almost exclusively from original Au(6s) atomic orbitals of the Au13 icosahedral core, the HOMO
LUMO transition (i.e., the 670 nm peak) in the normal absorption spectrum can be viewed as a transition occurring in the Au13 core. But, interestingly, no CD signals were observed around the ∼670 nm peak (see Supporting Information Figure S4 for Au25(SG)18 and Supporting Information Figure S5 for Au25(pet*)18), which implies a symmetric HOMO
LUMO transition dipole. This is indeed reasonable in consideration of the geometric structure of Au25, for that the Au13 core possesses an 3966

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969

Nano Letters

LETTER

Figure 5. The CD and normal UV
vis absorption spectra of chiral Au25 rods.

icosahedral symmetry (Ih) and should not be chirally active. It should be noted that the icosahedral Au13 kernel of Au25(pet)18 is not an ideal icosahedron, instead, it has some distortions31,32 but the HOMO
LUMO transition occurring in the Au13 kernel is not optically active, indicating that slight structural distortion of the Au13 kernel does not lead to CD signals (at least not detectable). In contrast, strong CD signals are observed at higher energy electronic transitions (λ < 600 nm), which all involve more or less character of surface thiolate ligands since these peaks start from the Au12 shell’s d-band (Figure 4B). Therefore, the observed chiroptical responses clearly indicate the critical role of ligands (i.e., the type and configuration of thiolate) in effecting the circular dichroism of optical signals from the nanoparticles. Our results explicitly show that the CD signals of chiral ligandcapped Au25 nanoparticles arise from the electronic state mixing of the ligands and surface gold atoms. We have also investigated the effect of Au25(pet*)18 charge state on the CD responses. Using the S-Au25 nanoparticle as an example, the native anionic S-Au25
nanoparticles are oxidized to neutral S-Au250 using H2O2 as in the case of nonchiral Au25(pet)18 nanoparticles.32 Interestingly, the CD spectrum is essentially the same as that of the anionic particle (Supporting Information Figure S5). This is consistent with the electronic properties of Au25 nanoparticles, for that the
1 charge resides primarily at the Au13 kernel, rather than on the Au12(pet*)18 surface, hence, the chiroptical response is not affected. With respect to the oxidation of Aun(SR)m nanoclusters, one would speculate that the surface
SR ligands should be oxidized first as indicated in previous thiol SAM research on bulk gold. But in the case of nanoclusters, the situation is the opposite. For example, the HOMO orbital of [Au25(SR)18]q is primarily contributed by the Au13 kernel, rather than by surface Au-SR bonds. The charge q primarily resides at the Au13 kernel, rather than on the surface. This is supported by experimental studies on paramagnetic [Au25(pet)18]0 and theory as well.39 Recently, Garzon et al15 performed DFT calculations on chiral cysteine-capped Au25(SCys)18 and found that the origin of the optical activity is due to the distortion in the metal core enhanced by the dissymmetric location of the chiral ligands. Tsukuda et al. prepared chiral phosphine-capped [Au11(R-BINAP)4X2]+ and [Au11(S-BINAP)4X2]+ nanoparticles, where BINAP represents bidentate phosphine ligand 2,20 -bis(diphenylphosphino)-1,10 binaphthyl and observed CD signals;40 they attributed the origin of the chiroptical activity to the structural deformation of the

Au11 core. In relevant theoretical work, Aikens et al. found that the phosphine ligands induce a slightly chiral feature of the metal core structure, and that the CD signal intensity is further enhanced by the presence of the chiral phosphine ligands.14 Here we comment on the structural distortion in Au25. X-ray crystallography shows that anionic Au25(pet)18 shows larger distortions than neutral Au25(pet)18.32 Should the slight distortion be responsible for the observed CD signals, one would observe stronger CD signals from the anionic Au25(pet*)18 than from the neutral Au25(pet*)18, but this is not the case. Our data shows equal intensities of CD signals from anionic and neutral Au25(pet*)18 (Supporting Information Figure S5). Overall, we deem that the small distortions in the metal core31,32 are less important, instead, the electronic state mixing of ligand and surface gold atoms should play a critical role in effecting the CD responses from the nanoparticles. For instance, structural distortions were observed in Au25 nanoparticles capped achiral phenylethylthiolate ligand, but such nanoparticles do not exhibit any measurable CD signals, albeit theoretical calculations give rise to CD signals of a distorted Au25 framework.15 One would argue that the null CD signal of Au25(pet)18 is due to a racemic mixture, but no mirrorimaged Au25(pet)18 pair exists in the crystal structure. To gain further insight into the origin of chirality in Au25 nanoparticles, we have synthesized another type of 25-atom (Au25) nanoparticles capped by mixed achiral triphenylphosphine and chiral R- or S-pet* following a literature method.41 The crystal structure of this Au25 rod has been previously reported by Tsukuda et al.,42 which is a vertex-sharing biicosahedral structure, resembling a rod, with the five thiolate ligands located at the waist of the rod and ten phosphines evenly distributed at the two icosahedra. Interestingly, we observed distinct CD signals from the [Au25(PPh3)10(pet*)5Cl2]2+ (counterion: Cl
or Br
) rods after incorporating five chiral R- or S-pet* ligands (Figure 5A). In contrast, the Au25 rod capped by achiral pet and PPh3 shows no CD response at all, though the distortions in this Au25 structure are quite large compared to the ideal biicosahedron; note that our recent X-ray crystallographic analysis has confirmed the same structure of [Au25(PPh3)10(pet)5Cl2]2+ as [Au25(PPh3)10(SC2H5)5Cl2]2+ reported in ref 42; thus, the replacement of SC2H5 by pet thiolate does not alter the particle structure. The CD spectrum of S-Au25 rods shows peaks at 490 nm (+ sign), 430 nm (+), and 330 nm (+, intense), whereas the R-isomer shows completely inversed peaks (i.e., mirror image relationship with the S-isomer). Similar to the case of spherical Au25(pet*)18 3967

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969

Nano Letters nanoparticles, the HOMO
LUMO transition at 675 nm is not chirally active since this transition does not involve surface gold atoms, while the higher energy transitions (λ < 600 nm) are strongly CD active due to the involvement of surface gold atoms and ligands. The differences in the CD spectra of Au25(pet*)18 nanospheres and Au25(PPh3)10(pet*)5Cl2 nanorods are due to their different electronic structures as revealed by DFT calculations.43 The normal UV
vis absorption spectra of R- and S-Au25 rods are identical (Figure 5B), indicating again that the thiolate ligand, being chiral or not, does not affect the normal optical absorption spectra as in the case of spherical Au25 nanoparticles. Finally, it is worth commenting on the size dependence of chirality in gold nanoparticles. For all the reported Aun(SR)m nanoparticle structures (n = 25, 38, 102), their kernels are indeed achiral, and the observed chirality is caused by surface atom arrangement (e.g., the cases of n = 38 and 102) and/or ligand induction (e.g., the case of n = 25). Therefore, the chiral feature in Aun(SR)m nanoparticles is expect to become less prominent with increasing size. In summary, we have devised a well-defined nanoparticle system for investigating the origin of chirality. We choose the Au25 (sphere vs rod) nanoparticles as a model system since their structures (being achiral), electronic, and optical properties are known. Specifically, we have designed chiral 2-phenylpropane-1thiol (pet*), which is a close analogue of phenylethylthiol (pet), and successfully prepared chiral thiolate-capped Au25 nanoparticles. In all cases, the chiral thiolate ligands do not affect the normal optical absorption spectrum of the nanoparticles, but the CD spectra of these nanoparticles critically depend on the ligand type (e.g., pet* vs SG, where SG = glutathionate). Our results explicitly reveal the major roles of ligands and surface gold atoms in giving rise to CD responses in the metal-based electronic transitions. Interestingly, the HOMO
LUMO transition is not optically active. The metal-based electronic transitions with higher energy than the HOMO
LUMO transition are imparted with ligand character, hence, giving rise to strongly liganddependent CD responses that reflect “fingerprints” of the ligand. Overall, the chirality is resulted from a mixing of ligand orbitals with those of the surface gold atoms of the nanoparticles, while the contribution from the slight distortion in the structure of the Au25 nanoparticles is not detectable. This scenario is in contrast with the inherently chiral metal core in the Au102 and Au38 cases. Understanding of the origin of chirality in nanoparticles is expect to benefit future exploration of applications of chiral nanoparticles, for example, chiral catalysis, separation, and many others.

’ ASSOCIATED CONTENT

bS

Supporting Information. Experimental procedures and supporting figures S1
5. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: (M.Z.) [email protected]; (R.J.) [email protected].

’ ACKNOWLEDGMENT M.Z. acknowledges financial support by NSFC (20871112, 21072001), 211 Project of Anhui University. R.J. acknowledges

LETTER

support by the Air Force Office of Scientific Research under AFOSR Award No. FA9550-11-1-9999 (FA9550-11-1-0147).

’ REFERENCES (1) Pasteur, L. C. R. Acad. Sci. 1848, 26, 535–538. (2) Tamura, M.; Fujihara, H. J. Am. Chem. Soc. 2003, 125, 15742– 15743. (3) Corma, A.; Serna, P. Science 2006, 313, 332–334. (4) Zhu, Y.; Qian, H.; Drake, B. A.; Jin, R. Angew. Chem., Int. Ed. 2010, 49, 1295–1298. (5) Zhu, Y.; Wu, Z.; Gayathri, C.; Qian, H.; Gil, R. R.; Jin, R. J. Catal. 2010, 271, 155–160. (6) Shukla, N.; Bartel, M. A.; Gellman, A. J. J. Am. Chem. Soc. 2010, 132, 8575–8580. (7) Schaaff, T. G.; Whetten, R. L. J. Phys. Chem. B 2000, 104, 2630–2641. (8) Goldsmith, M. R.; George, C. B.; Zuber, G.; Naaman, R.; Waldeck, D. H.; Wipf, P.; Beratan, D. N. Phys. Chem. Chem. Phys. 2006, 8, 63–67. (9) Gautier, C.; B€urgi, T. ChemPhysChem 2009, 10, 483–492. (10) Garzon, I. L.; Reyes-Nava, J. A.; Rodriguez-Hernandez, J. I.; Sigal, I.; Beltran, M. R.; Michaelian, K. Phys. Rev. B 2002, 66, 073403. (11) Roman-Velazquez, C. E.; Noguez, C.; Garzon, I. L. J. Phys. Chem. B 2003, 107, 12035–12038. (12) Noguez, C.; Garzon, I. L. Chem. Soc. Rev. 2009, 38, 757–771. (13) Lopez-Acevedo, O.; Tsunoyama, H.; Tsukuda, T.; Hakkinen, H.; Aikens, C. M. J. Am. Chem. Soc. 2010, 132, 8210–8218. (14) Provorse, M. R.; Aikens, C. M. J. Am. Chem. Soc. 2010, 132, 1302–1310. (15) Sanchez-Castillo, A.; Noguez, C.; Garzon, I. L. J. Am. Chem. Soc. 2010, 132, 1504–1505. (16) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Science 2007, 318, 430–433. (17) Qian, H.; Eckenhoff, W. T.; Zhu, Y.; Pintauer, T.; Jin, R. J. Am. Chem. Soc. 2010, 132, 8280–8281. (18) Gautier, C.; B€urgi, T. J. Am. Chem. Soc. 2008, 130, 7077–7084. (19) Gautier, C.; B€urgi, T. J. Am. Chem. Soc. 2006, 128, 11079– 11087. (20) Yao, H.; Miki, K.; Nishida, N.; Sasaki, A.; Kimura, K. J. Am. Chem. Soc. 2005, 127, 15536–15543. (21) Yao, H.; Fukui, T.; Kimura, K. J. Phys. Chem. C 2007, 111, 14968–14976. (22) Knoppe, S.; Dharmaratne, A. C.; Schreiner, E.; Dass, A.; B€urgi, T. J. Am. Chem. Soc. 2010, 132, 16783–16789. (23) Si, S.; Gautier, C.; Boudon, J.; Taras, R.; Gladiali, S.; B€urgi, T. J. Phys. Chem. C 2009, 113, 12966–12969. (24) Yao, H.; Fukui, T.; Kimura, K. J. Phys. Chem. C 2008, 112, 16281–16285. (25) Yao, H. Curr. Nanosci. 2008, 4, 92–97. (26) Cathcart, N.; Kitaev, V. J. Phys. Chem. C 2010, 114, 16010–16017. (27) Kumar, S.; Bolan, M. D.; Bigioni, T. P. J. Am. Chem. Soc. 2010, 132, 13141–13143. (28) Slocik, J. M.; Govorov, A. O.; Naik, R. R. Nano Lett. 2011, 11, 701–705. (29) Molotsky, T.; Tamarin, T.; Moshe, A. B.; Markovich, G.; Kotlyar, A. B. J. Phys. Chem. C 2010, 114, 15951–15954. (30) Wu, Z.; Gayathri, C.; Gil, R. R.; Jin, R. J. Am. Chem. Soc. 2009, 131, 6535–6542. (31) Zhu, M.; Aikens, C. M.; Hollander, F. J.; Schatz, G. C.; Jin, R. J. Am. Chem. Soc. 2008, 130, 5883–5885. (32) Zhu, M.; Eckenhoff, W. T.; Pintauer, T.; Jin, R. J. Phys. Chem. C 2008, 112, 14221–14224. (33) Zhu, M.; Lanni, E.; Garg, N.; Bier, M. E.; Jin, R. J. Am. Chem. Soc. 2008, 130, 1138–1139. (34) Wu, Z.; Suhan, J.; Jin, R. J. Mater. Chem. 2009, 19, 622–626. (35) Jin, R.; Qian, H.; Wu, Z.; Zhu, Y.; Zhu, M.; Mohanty, A.; Garg, N. J. Phys. Chem. Lett. 2010, 1, 2903–2910. 3968

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969

Nano Letters

LETTER

(36) Dass, A.; Stevenson, A.; Dubay, G. R.; Tracy, J. B.; Murray, R. W. J. Am. Chem. Soc. 2008, 130, 5940–5946. (37) Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R. W. J. Am. Chem. Soc. 2008, 130, 3754–3755. (38) Aikens, C. M. J. Phys. Chem. Lett. 2011, 2, 99–104. (39) Zhu, M.; Aikens, C. M.; Hendrich, M. P.; Gupta, R.; Qian, H.; Schatz, G. C.; Jin, R. J. Am. Chem. Soc. 2009, 131, 2490–2492. (40) Yanagimoto, Y.; Negishi, Y.; Fujihara, H.; Tsukuda, T. J. Phys. Chem. B 2006, 110, 11611–11614. (41) Qian, H.; Zhu, M.; Lanni, E.; Zhu, Y.; Bier, M. E.; Jin, R. J. Phys. Chem. C 2009, 113, 17599–17603. (42) Shichibu, Y.; Negishi, Y.; Watanabe, T.; Chaki, N. K.; Kawaguchi, H.; Tsukuda, T. J. Phys. Chem. C 2007, 111, 7845–7847. (43) Nobusada, K.; Iwasa, T. J. Phys. Chem. C 2007, 111, 14279– 14282.

3969

dx.doi.org/10.1021/nl202288j |Nano Lett. 2011, 11, 3963–3969