Excitonic Circular Dichroism of Chiral Quantum Rods - Journal of the

Jun 5, 2017 - Shenzhen Key Laboratory of Micro-Nano Photonic Information Technology, College of Electronic Science and Technology, Shenzhen University...
1 downloads 19 Views 4MB Size
Download PDF
Article pubs.acs.org/JACS

Excitonic Circular Dichroism of Chiral Quantum Rods Xiaoqing Gao,† Xiuwen Zhang,‡ Ke Deng,† Bing Han,† Luyang Zhao,§ Minghui Wu,# Lin Shi,† Jiawei Lv,† and Zhiyong Tang*,† †

CAS Key Laboratory of Nanosystem and Hierarchical Fabrication, CAS Center for Excellence in Nanoscience, National Center for Nanoscience and Technology, Beijing 100190, People’s Republic of China ‡ Shenzhen Key Laboratory of Micro-Nano Photonic Information Technology, College of Electronic Science and Technology, Shenzhen University, Guangdong 518060, People’s Republic of China § National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Science, Beijing 100190, People’s Republic of China # Department of Physics, South University of Science and Technology of China, Shenzhen, Guangdong 518055, People’s Republic of China S Supporting Information *

ABSTRACT: As an emerging type of optically active materials, chiral molecules-stabilized semiconductor quantum dots (QDs) have achieved extensive attention. Unfortunately, understanding of the optical characteristics of chiral QDs observed by circular dichroism (CD) spectroscopy remains a great challenge due to their rather weak signals. Herein, we successfully achieve much enhanced CD responses from L- or D-cysteinestabilized wurtzite CdSe quantum rods (QRs) thanks to their unique optical anisotropy. Furthermore, the optical activity of CdSe QRs is explored to be improved and subsequently become stable with the geometrical aspect ratio (AR) increasing, and such change matches well with alternation of the polarization factor of CdSe QRs. A non-degenerate coupled-oscillator (NDCO) model is established to elucidate the optical activity of chiral QRs, and the positive and negative natures of the CD peaks appearing at the first exciton band are clearly assigned to different transition polarization along 4pz,Se → 5sCd and 4p(x,y),Se → 5sCd, respectively. This work opens the door toward comprehension and design of optically active semiconductor nanomaterials.

■

INTRODUCTION Colloidal semiconductor nanocrystals (NCs) of unique optical properties have attracted much scientific and industrial interest for the last three decades because they not only offer an ideal platform for fundamental understanding of quantum confinement effect1 but also show many potential applications in displays and lighting,2 photovoltaics,3 lasers,4 fluorescent biological labels,5 and so on. Among varied types of NCs, of special interest is recently developed chiral semiconductor NCs, since their optical activity provides the additional opportunity for advanced applications in chiral recognition and separation, stereoselective catalysis, and high-performance optical devices, etc.6 Unfortunately, the optical activity of the currently reported semiconductor NCs is far from satisfaction. Generally, the development of chiral semiconductor NCs experiences two stages. In the first stage, chiral semiconductor NCs, referring to spherical quantum dots (QDs), were prepared by direct aqueous-phase synthesis with chiral molecules as stabilizers.6d,e,h,i However, no circular dichroism (CD) signals at the excitonic band of QDs were observed, and a possible reason was attributed to the wide size distribution of as-prepared © 2017 American Chemical Society

products. Therefore, in the second stage, QDs with improved narrow size distribution were obtained by organic-phase synthesis followed by ligand exchange with chiral molecules.6a,7 Yet, very weak CD signals at the characteristic absorption band of QDs were discerned. In parallel, many theoretical methods were developed for understanding and prediction of the optical activity of chiral QDs, such as the discrete dipole method8 and time-dependent density functional theory (TDDFT).6a It should be pointed out that though great progress has been made, fundamental knowledge, for instance, the relationship between the CD response and the electronic state of chiral QDs as well as the key factors to determine the CD response of chiral QDs, remains unexplored, which is reasonable considering their weak optical activity. To solve this challenge, herein we suggest a new strategy to enhance the CD intensity of semiconductor NCs by increasing their anisotropy. One-dimensional (1D) CdSe quantum rods (QRs) stabilized by chiral molecules are selected, and the greatly increased CD intensity is successfully achieved. More Received: April 29, 2017 Published: June 5, 2017 8734

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739

Article

Journal of the American Chemical Society importantly, the principle about generating and improving the optical activity of semiconductor nanostructures is clarified.

Table 1. Data Obtained from UV−Vis Absorption Spectra of CdSe NCs before and after Ligand Exchange with L- or DCysteinea

■

RESULTS AND DISCUSSION A series of wurtzite CdSe QRs with an identical diameter of ∼3.0 nm but different aspect ratio (AR) were synthesized in trioctylphosphine oxide (TOPO) following the classic thermal injection technique with slight modification.9 Three types of organic stabilizers including tetradecylphosphonic acid (TDPA), octadecylphosphonic acid (ODPA), and hexylphosphonic acid (HPA) were used to control the growth rate of CdSe QRs along the [001] axis. Figure 1 presents the typical

HWHMb (nm)

band-edge transition (nm) AR

A

B

C

A

B

C

1.0 1.7 1.9 2.7 4.2 7.3

516 551 553 571 576 578

517 541 543 560 564 565

+1 −10 −10 −11 −12 −13

15 13 15 14 14 18

15 15 16 16 16 23

+0 +2 +1 +2 +2 +5

a

A: before ligand exchange with L- or D-cysteine. B: after ligand exchange. C: different value between A and B. bHalf width at halfmaximum of the absorption peak.

reasonable because the longer QRs have a wider diameter distribution.1c An obvious feature in Figure S6 and Table 1 is the red shift of the band-edge transition with AR increasing, which is coincident with theoretical prediction.12 Moreover, the red shift range of the band-edge transition becomes smaller, for instance, from 541 to 560 nm or from 564 to 565 nm, when the AR of CdSe QRs alters from 1.7 to 2.7 or from 4.2 to 7.3, respectively. This phenomenon is caused by transition from three-dimensional (3D) to two-dimensional (2D) quantum confinement.12 CD response, the differential absorption of left-hand circular (LHC) and right-hand circular (RHC) polarized light, is a primary means to characterize the materials’ chirality.13 As for CdSe NCs capped by chiral cysteine molecules, a complete CD spectrum should include both the electronic transition of chiral ligands on NC surfaces at the UV band and the excitonic transition of semiconductors at the visible band.6i,j In the UV band from 190 to 400 nm, two peaks of I and II appear in the CD spectra (Figure 2a and 2b, and S7 and S8).

Figure 1. TEM images of CdSe NCs with a diameter of ∼3.0 nm and different AR of (a) 1.0, (b) 1.7, (c) 1.9, (d) 2.7, (e) 4.2, and (f) 7.3.

transmission electron microscope (TEM) images of six different products, whose diameter and length are determined by measuring more than 200 individual particles (Figures S2 and S3). Notably, all as-prepared samples are monodispersed in shape and size (Figure S2), and their ARs are 1.0, 1.7, 1.9, 2.7, 4.2, and 7.3. Subsequent ligand exchange with L- or D-cysteine allows QRs fully transferring from organic solvent to water without destruction of geometry and crystal structures, as confirmed by optical photos, TEM images, powder X-ray diffraction (XRD) patterns, and nuclear magnetic resonance (NMR) spectra (Figures S1−S5). The narrow size and shape distribution is known to be an essential prerequisite for study of the size- and shapedependent optical properties of NCs. Figure S6 records the UV−vis absorption spectra of both CdSe QDs and QRs before and after ligand exchange with chiral cysteine molecules. Impressively, the discrete energy level and tunable band gap can be clearly distinguished from the absorption spectra, thanks to the size and shape uniformity of all products (Figures 1 and S2 and S3). Table 1 further quantitatively summarizes the information drawn from the UV−vis absorption spectra of CdSe NCs before and after ligand exchange. For all samples with varied AR, the first excitonic absorption peaks, which correspond to the band-edge transition, display only slightly increased half width at the half-maximum (HWHM) after ligand exchange, likely owing to a change of solvent and surface ligands.10 Furthermore, the HWHM values of all samples except for CdSe QRs with an AR of 7.3 are small enough to be about 15 nm, highlighting the high quality of as-prepared products even after the ligand exchange process.11 A larger HWHM value of 23 nm for CdSe QRs with an AR of 7.3 is

Figure 2. CD and UV absorption spectra (from 190 to 400 nm) and anisotropic g-factor of L- or D-cysteine-stabilized CdSe NCs with different AR in neutral solution (pH ≈ 7). (a) CD spectra of Lcysteine-stabilized CdSe NCs with different AR. (b) CD spectra of Dcysteine-stabilized CdSe NCs with different AR. (c) UV absorption spectra of L-cysteine-stabilized CdSe NCs with different AR. (d) Linear fit of the anisotropic g-factor at peak II (255 nm) of L-cysteinestabilized (blue line) or D-cysteine-stabilized (red line) CdSe NCs with different AR. 8735

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739

Article

Journal of the American Chemical Society Peak I corresponds to nO→ πC*= O transition from the lone pair on the CO to the antibonding πC*= O valence orbital in cysteine molecule,14 while peak II originates from the 3pS → 5sCd electronic transition of the Cd−S bond on the surface of NCs.15 Keeping the same absorption intensity at 200 nm (Figure 2c), the CD signal change upon these two peaks is obviously different. Contrary to the small intensity fluctuation of peak I against AR, peak II gradually appears and its intensity steadily increases with AR increasing. Quantitative analysis on the CD peaks is further implemented by calculating and comparing their anisotropic g-factor (Table S1). Figure 2d and Table S2 indicate the linear increase of the g-factor of peak II against AR, resulting from the array of Cd−S bonds along the long axis of QRs. This increased CD intensity of peak II against AR could be described by a non-degenerate coupled-oscillator model and will be discussed latter. In the visible band from 400 to 700 nm, both CD spectra and visible absorption spectra show a similar red shift against AR (Figure 3a−c). However, in comparison with only two (QRs)

Figure 4. CD and visible absorption spectra of CdSe QRs with AR of 1.7. First exciton transition (peak III) in the visible absorption spectrum corresponds to two opposite peaks (IIIa and IIIb) in the CD spectrum.

the anisotropic g-factor against AR also shows the same tendency with the optical activity change (Figure 3d and Table S3), which is very similar to the reported AR-dependent linear polarized emission that is also related to the anisotropy of QRs.1c,12a The greatly enhanced CD intensity of QRs enables us to understand the intrinsic optical activity of chiral semiconductor NCs. The first question that we must answer is how the optical activity is generated inside chiral NCs. Molecules with chromophores are known to exhibit a CD effect when possessing no reflection symmetry. Here, we suggest that the whole hybrid of the optically active molecule-stabilized CdSe NC is a huge artificial chiral “molecule”, which has no symmetric plane and center due to a “chiral surface”. In other words, chiral molecules on the surface of NC destroy its intrinsic symmetry and endow the whole hybrid optical activity. Therefore, in CdSe NC, the Cd−S bond on its surface and C O group in cysteine are considered as three chromophores in this huge chiral “molecule”, respectively. The asymmetric carbon center in cysteine offers a chiral geometry in the chromophores. With the above assumption the rich optical activity of chiral cysteine-stabilized CdSe QRs is understood through a nondegenerate coupled-oscillator (NDCO) model (part S4 in the Supporting Information).16 In this model, the CD signal is produced by coupling of the electric dipole transition moments in different chromophores (Figure 5a). The CD strength of chromophore A is determined by coupling with a different chromphore C, which is expressed as16

Figure 3. CD and visible absorption spectra (from 400 to 700 nm) and anisotropic g-factor of L- or D-cysteine-stabilized CdSe NCs with different AR. (a) CD spectra of L-cysteine-stabilized CdSe NCs with different AR. (b) CD spectra of D-cysteine-stabilized CdSe NCs with different AR. (c) Visible absorption spectra of L-cysteine-stabilized CdSe NCs with different AR. (d) Tendency of g-factor at the largest CD peak (523−582 nm) of L-cysteine-stabilized (blue curve) and Dcysteine-stabilized (red curve) CdSe NCs with different AR.

or three (QDs) peaks appearing in absorption spectra (Figure 3c), CD signals are more complex and comprise multiple peaks (Figure 3a and 3b), which originate from the complicated energy band structure of the semiconductor under strong quantum confinement.6j For instance, the first exciton absorption peak III of QRs in the visible absorption spectrum well correlates with their two opposite peaks IIIa and IIIb in the CD spectrum (Figure 4 and S9). Impressively, the higher exciton absorption peaks in the visible absorption spectrum are hardly discerned at room temperature but can be clearly detected by CD spectroscopy (Figure S9). This fact demonstrates that the CD spectrum is much more sensitive than the UV−vis spectrum, owing to the selection rule of polarized transition under circular polarized light. When AR increases from 1.0 to 1.7, the CD intensity shows almost 10 times enhancement under the same absorption intensity, whereas with further increase of AR from 1.7 to 7.2, the CD intensity only slightly changes (Figure 3a and 3c). The plot of

R(|00⟩ → |10⟩) =

∑ C

−εA εCV1c ℏ(εC2 − εA2)

(μC0c × μA01 ·rAC)

(1)

here R is the CD intensity, the ground and excited states of A are denoted as 0 and 1, the ground and excited states of C are denoted as 0 and c, ℏ is the reduced Planck constant, εA(C) is the transition energy of A or C, μA(C) is the electric dipole transition moment of A or C, rAC is the distance vector from A to C, and V1c is attributed to the Coulombic interaction between A and C (eq S5). Note that in eq 1 the boldface represents vector, while the sum stands for all chromophores C around A. Clearly, eq 1 discloses that the negative or positive 8736

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739

Article

Journal of the American Chemical Society

give a rise to an opposite CD sign at nO → πC*= O transition in cysteine (Figure 5c). As a result, the final intensity of peak I is the superposition of the nO → πC*= O transition coupling with each transition in the Cd−S bond and CdSe QR and its intrinsic electric and magnetic dipole coupling. To confirm the validity of the NDCO model, the possible conformation of L- or D -cysteine on the QR surface is established, and its corresponding dipole moment is estimated by density functional theory (part S5 in the Supporting Information). Remarkably, as demonstrated in Table S4, the calculated result about peaks II and IIIa based on the NDCO model, regardless of peak sign and intensity, well matches the experimental result (Figures 2a, 2b, 3a, and 3b). In addition, based on the fact that peak IIIa is of stronger but opposite CD response compared with peak II, it reasonable to conclude that peak I corresponding to the nO → πC*= O transition shows the same CD sign with peak II because of stronger coupling (Figure 5c). The second issue of concern is about the CD signals in the UV region, i.e., why the CD intensity of peak II (3pS → 5sCd transition in Cd−S bond) increases against AR increasing (Figure 2a and 2b). Since the CD strength of the Cd−S bond comes from its coupling with chiral carboxyl groups in L- or Dcysteine, dipole−dipole interaction should include all cysteine molecules along the long axis of QRs, and their number is noted as nCys. Once the length of QRs increases, nCys increases simultaneously and the electric dipole intensity of Cd−S bond enhances, finally leading to the improved CD intensity (eq S14). The gradually enhanced absorption in Figure 2c supports this assumption. Furthermore, beside the CD intensity, the theory calculation discloses that the anisotropic g-factor of the Cd−S bond is also related with nCys (eqs S15 and S16), resulting in a linear increase of the g-factor of peak II against AR observed in experiments (Figure 2d). It needs to be pointed out that both the CD intensity and the g-factor of peak I corresponding to the nO → πC*= O transition only show a slight change with AR increasing (Figure 2a and 2b and Table S1). This is results from the fact that although the g-factor of peak II linearly increases against AR, other peaks in the visible band also have some changes, especially the opposite peaks IIIa and IIIb. Interestingly, the sum of the g-factor of IIIa and IIIb also increases against AR increasing to a first approximation, but the sign is opposite with that of peak II (Table S3). Therefore, such a counteraction causes peak I to remain almost independent with AR varying. The third question arises in the visible region, that is, why Lor D-cysteine-stabilized QRs possess a much larger optical activity than QDs (Figure 3). The cross over between the highest occupied levels with the NC shape transformation from QDs to QRs is responsible for this phenomenon.1c,19 As for bulk CdSe, the highest occupied molecular orbital (HOMO) is known to come from the Se 4p orbital, while the lowest unoccupied molecular orbital (LUMO) originates from the Cd 5s orbital. The 4pz orbital has greater momentum projected onto the [001] axis of the CdSe crystal with respect to 4px and 4py orbitals.1c Hence, the fine structure of CdSe NCs under quantum confinement effect is changed when QDs grow to QRs along the c axis, and the highest energy level of the valence band is switched from 4px,y to 4pz.1c,12a,19b As a result, the large increase of the CD intensity of QRs compared with QDs is ascribed to the transition from plane-polarized optics inside the x−y plane to highly linearly polarized optics along the z

Figure 5. Schematic diagrams of the NDCO model, and qualitative analysis on the optical activity of L-cysteine-stabilized CdSe QR. (a) Scheme of coupling between two electric dipole transition moments in a chiral system, where coupling of two dipoles in different chromophores results in two opposite CD signs. (b) Simplified view of L-cysteine-stabilized CdSe QR, where blue dotted arrows present coupling between two electric dipoles. Note that coupling between Cd−S and CdSe has no contribution to CD signal (red cross). (c) Charting CD spectrum of L-cysteine-stabilized QRs. CD signals of the Cd−S bond and CdSe QR come from their coupling with CO double bonds in cysteine, and the CD sign of CO results from superposition of each coupling (dotted curves). (d) Quantum transition in the system. Solid vertical (horizontal) arrows represent light (coulomb)-induced transitions. Dotted vertical arrows stand for relaxation processes.

nature of CD peak comes from the relative geometry of two coupled chromophores as manifested in Figure 5a and S11. Furthermore, the CD sign induced from chromophore A to C should be opposite in magnitude to that from chromophore C to A. In addition, the anisotropic g-factor is approximately presented as17 g≅

4R(|00⟩ → |10⟩) D(|0⟩ → |1⟩)

(2)

where D is the electric dipole intensity that is equal to the dot 10 product of μ01 A and μA . Consequently, the CD signs of CdSe QRs and the Cd−S bonds should be ascribed to the transition coupling between either of them with the amino acid. Also, because there is no chiral geometry between the Cd−S bond and CdSe NC, their coupling has no contribution to the CD spectrum (red cross in Figure 5b), that is, the peak II in Figure 2a and 2b comes from the 3pS → 5sCd transition in the Cd−S bond coupling with nO → πC*= O transition. The numerous up and down CD waves in the 400−600 nm region (Figure 3a and 3b; peaks IIIa, IIIb and others in Figure 4) belong to coupling of complicated excitonic transitions in QR with the nO → πC*= O transition. It turns out that CdSe QR possesses two dipole orientations due to geometry-dependent polarization:18 one is oriented parallel to the long axis (denoted as μQR,L), and the other is along the radial axis (denoted as μQR,R), as displayed in Figure 5b. It is reasonable that the oscillation of CD signals at the characteristic absorption of CdSe QRs comes from the different polarization of the excitonic transitions. Theoretically, each CD sign from the transition either in the Cd−S bond or CdSe QRs would 8737

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739

Journal of the American Chemical Society

■

direction. The polarized exciton transition would be one major impact factor on the strength of CD signals.20 A particularly noteworthy phenomenon is the AR-dependent change of g-factor (Figure 3d). The g-factor increases with AR increasing and reaches a saturate value when the AR is about 4. The analysis based on the NDCO model demonstrates that due to the quantum confinement, the effective number of cysteine coupling with QR increases with QR length increasing and would get the highest value when the length is about twice the exicton Bohr radius (eq S24). Since the exicton Bohr radius of bulk CdSe is known to be 5.6 nm21 and the prepared QRs have a diameter of about 3.0 nm, the critical AR to obtain the largest g-factor is expected to be 3.7, which is in good agreement with the experimental observation (Figure 3d). The last but not the least theme that must be addressed is attribution of the CD peaks at the characteristic absorption of CdSe QRs (Figure 4). CdSe QR of the wurtzite crystal structure has two dipole orientations, μQR,L and μQR,R,22 which give rise to opposite coupling with carboxyl groups (Figure 5b). Furthermore, μQR,L and μQR,R correspond to the excitonic transition of 4pz,Se → 5sCd and 4p(x,y),Se → 5sCd, respectively.12a,19b Therefore, in regard of the first two opposite peaks at the characteristic absorption of CdSe QRs (IIIa and IIIb in Figure 4), IIIa is ascribed to 4pz,Se → 5sCd and IIIb is attributed to 4p(x,y),Se → 5sCd. This peak assignment is quantitatively supported by experimental evidence. The splitting energy between peaks IIIa and IIIb is estimated to be about 45 meV for the CdSe QRs of AR no more than 4.2 (Figure S10, Table S5). This value is identical with the difference in the fine energy levels of 4pz,Se → 5sCd and 4p(x,y),Se → 5sCd in the first exciton transition.19b

AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Zhiyong Tang: 0000-0003-0610-0064 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge financial support from National Key Basic Research Program of China (2014CB931801 and 2016YFA0200700, Z.Y.T.), National Natural Science Foundation of China (21475029 and 91427302, Z.Y.T.), Frontier Science Key Project of the Chinese Academy of Sciences (QYZDJ-SSW-SLH038, Z.Y.T.), Instrument Developing Project of the Chinese Academy of Sciences (YZ201311, Z.Y.T.), CAS-CSIRO Cooperative Research Program (GJHZ1503, Z.Y.T.), K. C. Wong Education Foundation and “Strategic Priority Research Program” of Chinese Academy of Sciences (XDA09040100, Z.Y.T.), and National Key Research and Development Program of China Grant (2016YFB0700700). We thank Prof. Suhuai Wei and Jingbo Li for useful discussion.

■

REFERENCES

(1) (a) Klimov, V. I. Annu. Rev. Phys. Chem. 2007, 58, 635. (b) Kambhampati, P. Acc. Chem. Res. 2011, 44, 1. (c) Hu, J.; Li, L.-s.; Yang, W.; Manna, L.; Wang, L.-w.; Alivisatos, A. P. Science 2001, 292, 2060. (2) (a) Wood, V.; Panzer, M. J.; Halpert, J. E.; Caruge, J. M.; Bawendi, M. G.; Bulović, V. ACS Nano 2009, 3, 3581. (b) Coe, S.; Woo, W.-K.; Bawendi, M.; Bulovic, V. Nature 2002, 420, 800. (3) (a) Kramer, I. J.; Sargent, E. H. Chem. Rev. 2014, 114, 863. (b) Nozik, A. J.; Beard, M. C.; Luther, J. M.; Law, M.; Ellingson, R. J.; Johnson, J. C. Chem. Rev. 2010, 110, 6873. (4) Dang, C.; Lee, J.; Breen, C.; Steckel, J. S.; Coe-Sullivan, S.; Nurmikko, A. Nat. Nanotechnol. 2012, 7, 335. (5) (a) Bruchez, M., Jr Science 1998, 281, 2013. (b) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538. (6) (a) Tohgha, U.; Deol, K. K.; Porter, A. G.; Bartko, S. G.; Choi, J. K.; Leonard, B. M.; Varga, K.; Kubelka, J.; Muller, G.; Balaz, M. ACS Nano 2013, 7, 11094. (b) Tohgha, U.; Deol, K. K.; Porter, A. G.; Bartko, S. G.; Choi, J. K.; Leonard, B. M.; Varga, K.; Kubelka, J.; Muller, G.; Balaz, M. ACS Nano 2013, 7, 11094. (c) Han, B.; Zhu, Z.; Li, Z.; Zhang, W.; Tang, Z. J. Am. Chem. Soc. 2014, 136, 16104. (d) Moloney, M. P.; Gun’ko, Y. K.; Kelly, J. M. Chem. Commun. 2007, 3900. (e) Elliott, S. D.; Moloney, M. c. l. P.; Gun’ko, Y. K. Nano Lett. 2008, 8, 2452. (f) Han, C.; Li, H. Small 2008, 4, 1344. (g) CarrilloCarrión, C.; Cárdenas, S.; Simonet, B. M.; Valcárcel, M. Anal. Chem. 2009, 81, 4730. (h) Nakashima, T.; Kobayashi, Y.; Kawai, T. J. Am. Chem. Soc. 2009, 131, 10342. (i) Zhou, Y.; Yang, M.; Sun, K.; Tang, Z.; Kotov, N. A. J. Am. Chem. Soc. 2010, 132, 6006. (j) Ben Moshe, A.; Szwarcman, D.; Markovich, G. ACS Nano 2011, 5, 9034. (k) Li, Y.; Zhou, Y.; Wang, H. Y.; Perrett, S.; Zhao, Y.; Tang, Z.; Nie, G. Angew. Chem., Int. Ed. 2011, 50, 5860. (l) Mukhina, M. V.; Maslov, V. G.; Baranov, A. V.; Fedorov, A. V.; Orlova, A. O.; Purcell-Milton, F.; Govan, J.; Gun’ko, Y. K. Nano Lett. 2015, 15, 2844. (7) Tohgha, U.; Varga, K.; Balaz, M. Chem. Commun. 2013, 49, 1844. (8) Zhou, Y.; Zhu, Z.; Huang, W.; Liu, W.; Wu, S.; Liu, X.; Gao, Y.; Zhang, W.; Tang, Z. Angew. Chem., Int. Ed. 2011, 50, 11456. (9) (a) Peng, X.; Manna, L.; Yang, W.; Wickham, J.; Scher, E.; Kadavanich, A.; Alivisatos, A. P. Nature 2000, 404, 59. (b) Peng, Z. A.; Peng, X. J. Am. Chem. Soc. 2002, 124, 3343. (c) Qu, L.; Peng, X. J. Am. Chem. Soc. 2002, 124, 2049.

■

CONCLUSION In summary, the AR-dependent optical activity of L- or Dcysteine-stabilized CdSe QRs is studied in detail. A nondegenerate coupled-oscillator (NDCO) model, in which the amino acid, Cd−S bond, and QRs are simplified as three chromophores, is developed to elucidate the origin and change of the complex CD response of CdSe QRs in both the UV and the visible regions. Both experimental and theoretical results reveal that the transition of linearly polarized exciton caused by the increased shape asymmetry is the reason why QRs exhibit much higher optical activity than spherical QDs. It deserves to be stressed that the previously reported techniques to determine AR-dependent linear polarization optics of CdSe NCs are time and cost consuming, for example, detecting a single nanoparticle,1c adding an electric field,19c,23 or arraying the samples.9a For comparison, application of CD spectroscopy will provide a new, general, and simple method to explore the transition polarization of size- and shape-dependent nanomaterials without requirement of tedious sample preparation and processing. In summary, this work lays the foundation for understanding and application of excitonic CD response of chiral semiconductor nanomaterials, which will be an important additive to chiral chemistry, nanoscience, and optical physics.

■

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b04224. Experimental results, theoretical calculations, and additional figures (PDF) 8738

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739

Article

Journal of the American Chemical Society (10) Murray, C. B.; Norris, D. J.; Bawendi, M. G. J. Am. Chem. Soc. 1993, 115, 8706. (11) Yu, W. W.; Qu, L.; Guo, W.; Peng, X. Chem. Mater. 2003, 15, 2854. (12) (a) Hu, J.; Wang, L.; Li, L.-s.; Yang, W.; Alivisatos, A. P. J. Phys. Chem. B 2002, 106, 2447. (b) Li, J.; Wang. Nano Lett. 2003, 3, 101. (13) Berova, N.; Bari, L. D.; Pescitelli, G. Chem. Soc. Rev. 2007, 36, 914. (14) (a) Osted, A.; Kongsted, J.; Christiansen, O. J. Phys. Chem. A 2005, 109, 1430. (b) Maul, R.; Preuss, M.; Ortmann, F.; Hannewald, K.; Bechstedt, F. J. Phys. Chem. A 2007, 111, 4370. (c) Nishino, H.; Kosaka, A.; Hembury, G. A.; Aoki, F.; Miyauchi, K.; Shitomi, H.; Onuki, H.; Inoue, Y. J. Am. Chem. Soc. 2002, 124, 11618. (15) Willner, H.; Vasak, M.; Kaegi, J. H. R. Biochemistry 1987, 26, 6287. (16) Nordén, B.; Rodger, A.; Dafforn, T. Linear Dichroism and Circular Dichroism: A Textbook on Polarized-Light Spectroscopy; Royal Society of Chemistry: Cambridge, U.K., 2010. (17) Schellman, J. A. Chem. Rev. 1975, 75, 323. (18) Efros, A. L. Phys. Rev. B: Condens. Matter Mater. Phys. 1992, 46, 7448. (19) (a) Li, J.; Wang. Nano Lett. 2003, 3, 1357. (b) Zhao, Q.; Graf, P. A.; Jones, W. B.; Franceschetti, A.; Li, J.; Wang; Kim, K. Nano Lett. 2007, 7, 3274. (c) Kamal, J. S.; Gomes, R.; Hens, Z.; Karvar, M.; Neyts, K.; Compernolle, S.; Vanhaecke, F. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 035126. (20) (a) Schellman, J. A. Acc. Chem. Res. 1968, 1, 144. (b) Schellman, J. A. J. Chem. Phys. 1966, 44, 55. (c) Applequist, J. J. Chem. Phys. 1973, 58, 4251. (21) Li, L.-s.; Hu, J.; Yang, W.; Alivisatos, A. P. Nano Lett. 2001, 1, 349. (22) Chen, X.; Nazzal, A.; Goorskey, D.; Xiao, M.; Peng, Z. A.; Peng, X. Phys. Rev. B: Condens. Matter Mater. Phys. 2001, 64, 245304. (23) Mukhina, M. V.; Baimuratov, A. S.; Rukhlenko, I. D.; Maslov, V. G.; Purcell Milton, F.; Gun’ko, Y. K.; Baranov, A. V.; Fedorov, A. V. ACS Nano 2016, 10, 8904.

8739

DOI: 10.1021/jacs.7b04224 J. Am. Chem. Soc. 2017, 139, 8734−8739