Evolution of Circularly Polarized Luminescence in Atomically Precise Silver Nanoclusters with Intrinsic Chirality

Abstract

Circularly polarized luminescence (CPL) is a powerful tool to probe the chiroptical properties of molecular systems in the excited state. Here, we demonstrate the evolution of excited‐state chiroptical property in silver nanoclusters (NCs) by CPL measurements. The use of bidentate thiolate ligand, 1,3‐benzenedithiol (BDT), with the aid of triphenylphosphine (PPh₃), affords racemic [Ag₂₉(BDT)₁₂(PPh₃)₄]³⁻ NCs (Ag₂₉ NCs) with intrinsic chirality in the exterior cage composed of a silver(I)‐dithiolate coordination framework, that are separable by chiral column chromatography. We have recently reported that the post‐synthetic modification of negatively charged Ag₂₉ NCs with a cationic silver(I) complex on the exterior cage stabilizes a specific excited state with a triplet character, enhancing photoluminescence (PL) in the near‐infrared (NIR) region. While the same modification on the separated enantiomeric NCs does not change the circular dichroism (CD) profile so much, the CPL activity is dramatically enhanced with an increase in the anisotropy factor (|g _lum| value) from <1 × 10⁻³ to 6 × 10⁻³, demonstrating the excited‐state dependent CPL property of Ag₂₉ NCs. In combination with the picosecond PL decay study, the low‐temperature CPL measurement indicates the time development of excited‐state property, suggesting the possible evolution of CPL activity.

The rich excited‐state dynamics of intrinsically chiral silver nanoclusters tethering cationic species on the surface demonstrates the excited‐state dependent circularly polarized luminescence (CPL), suggesting the temporal evolution of exciton distribution.

Introduction

Emissive metal nanoclusters (NCs) with atomically precise compositions and structures have attracted significant attention due to their insightful structure–property relationships crucial for a wide range of potential applications.^{[ 1} , ² , ³ , ⁴ , ⁵ , ⁶ , ^{7 ]} Recent advancements have enabled the development of NCs showing remarkably high photoluminescence quantum yields (PLQY),^{[ 8} , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ^{15 ]} along with efforts to understand photoluminescence (PL) mechanism including the excited‐state dynamics through studies with transient absorption spectroscopy and variable‐temperature (VT)‐PL analysis.^{[ 16} , ¹⁷ , ¹⁸ , ¹⁹ , ^{20 ]} Nevertheless, compared to small organic molecules and complexes, the geometrical or structural changes of NCs in the excited state remain less understood.

Circularly polarized luminescence (CPL) is a luminescence phenomenon in which chiral systems emit dissymmetric right‐ and left‐circularly polarized light, providing information on the excited‐state chiroptical property of molecular systems. While CPL property has been extensively investigated in various materials, dual or responsive CPL systems in small organic molecules provide a valuable insight into excited‐state dependent chirality.^{[ 21} , ²² , ²³ , ²⁴ , ^{25 ]} For example, relaxations from a locally excited state to other low‐lying states such as a charge transfer state are often accompanied geometrical changes of molecules, bringing about the variation in CPL property.^{[ 24 ]} Conformational changes in the excited state can also afford responsive CPL signals depending on the environment such as solvent polarity.^{[ 22} , ^{25 ]} Meanwhile, a number of chiral NCs have also been reported to exhibit CPL^{[ 26} , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ^{41 ]} and some studies succeeded in modulating inherent CPL performance through stereo‐controlled synthesis,^{[ 36 ]} heterometal‐atom doping,^{[ 37 ]} and self‐assembly.^{[ 38} , ^{39 ]} However, the relationship between CPL property and the excited‐state dynamics including chiral structural relaxations has been yet to be discussed partly due to the lack of appropriate materials. To shed light on this unexplored aspect, it is essential to investigate a chiral NC system showing a rich excited‐state dynamics.

The atomically precise [Ag₂₉(BDT)₁₂(PPh₃)₄]³⁻ (BDT: 1,3‐benzenedithiol, PPh₃: triphenylphosphine) NC (Ag₂₉ NC)^{[ 42 ]} is a possible candidate to discuss the excited‐state dependent chiroptical property. The structural chirality originates from an asymmetric arrangement of achiral BDT ligands in the surface cage surrounding the Ag₁₃ superatom core (Figure S1).^{[ 42} , ⁴³ , ^{44 ]} The intrinsic chirality of Ag₂₉ NCs enabled their enantioseparation through a chiral column.^{[ 43 ]} The Ag₂₉ NC exhibits distinct PL property attributed to the triplet (T₁) state‐based phosphorescence,^{[ 42} , ⁴⁵ , ^{46 ]} which could be modulated by the modification of NC surface.^{[ 40} , ⁴⁷ , ⁴⁸ , ^{49 ]} Recently, we demonstrated that alkali metal ions^{[ 47} , ^{49 ]} and silver(I) complexes^{[ 48 ]} could bind to the surface of trianionic Ag₂₉ NC in pyridine. The Ag(I)‐modified NCs (Ag₂₉‐IP NCs, IP: ion‐paired, Scheme 1) exhibited bright PL at 770 nm with quantum yield (QY) of 39%.^{[ 48 ]} The excited‐state dynamics of Ag₂₉ and Ag₂₉‐IP NCs were investigated through transient absorption spectroscopy and VT‐PL study. The original Ag₂₉ NCs exhibited the ultrafast (<0.5 ps) and quantitative intersystem crossing (ISC)^{[ 50 ]} to a long‐lived (>1 µs) T₁ state responsible for the phosphorescence at 680 nm. The Ag₂₉‐IP NCs showed an additional relaxation with a time constant of ca. 300 ps following the rapid ISC (Scheme 1).^{[ 48 ]} We assigned this decay to the relocation of cationic Ag(I)‐species on the surface of Ag₂₉ NC in response to the charge redistribution in the T₁‐state, lowering its energy level as T₁(IP) specific for ion‐paired (IP) NCs. The similar phenomenon was reported for ionic dyes with donor–π–acceptor structures showing the excited‐state intramolecular charge transfer (ESICT)‐induced counter ion migration.^{[ 51 ]} Given the relocation of cationic species (Ag⁺) on the surface induces the structural change or the migration of excited state (exciton) toward the chiral surface cage, the CPL property should be highly modulated. The Ag₂₉(‐IP) NCs thus serve as an ideal platform to discuss the relationship between the CPL and excited state properties for NC‐based materials. Here, we propose an unprecedented evolution of CPL property in Ag₂₉ NCs dependent on the excited‐state nature. The peripheral modification of enantioseparated Ag₂₉ NCs with Ag(I) complex to form Ag₂₉‐IP NCs enhanced the CPL performance alongside the PL peak shift to the near infrared (NIR) region. The PL decay measurement clearly suggests the excited‐state shift in the timescale of a few hundred ps. Complementary VT‐CPL study evokes the temporal evolution of excited‐state chirality in Ag₂₉‐IP NCs. Theoretical calculation based on the density functional theory (DFT) method demonstrated that the presence of cationic species on the surface stabilizes the T₁ energy and attracts the exciton to the chiral surface cage.

Scheme 1.

Simplified structures and PL pathways for a) original Ag₂₉ NC and b) Ag(I)‐complex modified Ag₂₉‐IP NCs. Color code: light blue = Ag in NC; purple = Ag in the peripheral complex; yellow = S; pink = P; blue = N; grey = C.

Results and Discussion

Enantioseparation and Ground‐State Chirality

Ag₂₉(BDT)₁₂(PPh₃)₄ (Ag₂₉) NCs were prepared according to the method reported by Bakr and coworkers.^{[ 42 ]} The obtained NCs dispersed in acetone were subjected to the enantioseparation process using a chiral column, DAICEL CHIRALPAK IA. Since the previous condition using ethanol as an eluent resulted in the incomplete enantioseparation,^{[ 43 ]} we searched for a better condition. The condition employing acetone as the eluent and a high NC concentration (4 mg mL⁻¹) for the injection solution led to the successful separation. It is of note that the low concentration (∼1 mg mL⁻¹) condition for the injection did not work. Part of NCs in the injected solution are presumed to coat or modify the surface of amylose‐based stationary phase stereoselectively and the chiral‐NCs‐modified surface further serves as an active layer for the enantioseparation through stereospecific NC–NC interactions. The eluate was fractionalized corresponding to the separated peaks at 16 and 19 min (Figure S2). Because of this less common separation behavior, the optical purity of the separated fractions could not be confirmed strictly using an analytical chiral column. The ESI‐MS measurement of the 1st fraction provided an m/z value corresponding to the [Ag₂₉(BDT)₁₂]³⁻ composition with loss of PPh₃ sub‐ligands, reproducing our previous report (Figure S3).^{[ 43 ]} Each separated fraction was evaporated to dryness and redispersed in 10 mM PPh₃ DMF solution to enhance the stability and PL intensity of Ag₂₉ NCs.^{[ 52} , ^{53 ]}

The enantioseparated NCs afforded almost identical absorption spectra and mirror image CD spectra in 10 mM PPh₃ solution in DMF (Figure 1a,c). According to our previous report,^{[ 43 ]} the major species in the 1^st and 2^nd fractions showing large negative and positive Cotton effects at 460 nm could be assigned to the right and left‐handed NCs, respectively (R‐ and L‐Ag₂₉ NCs, respectively, Figure S1). The anisotropy factors in absorption g _abs, defined as g _abs = Δε/ε, were −2.6 × 10⁻³ and + 2.5 × 10⁻³ at 460 nm for R‐ and L‐Ag₂₉ NCs, respectively, under the assumption that the complete enantioseparation was achieved (Figure S4). In fact, the |g _abs| values were more than twice compared to those in our previous report (|g _abs| < 1.2 × 10⁻³ at 460 nm) in accordance with the better separated peaks in the HPLC elution profile (Figure S2).^{[ 43 ]}

Figure 1.

a) and b) Absorption and c) and d) CD spectra of a) and c) R/L‐Ag₂₉ NCs in 10 mM PPh₃ DMF solutions and b) and d) R/L‐Ag₂₉‐IP NCs in pyridine. Blue: R‐Ag₂₉, red: L‐Ag₂₉, purple: R‐Ag₂₉‐IP, orange: L‐Ag₂₉‐IP NCs.

The effect of cation binding on the ground‐state chirality of Ag₂₉ NCs was first investigated. The enantioseparated samples were redispersed in pyridine and subsequently treated with the Ag(I)‐ complex, Ag(PPh₃)₄NO₃ (Ag‐PPh₃), to form the ion‐paired Ag₂₉‐IP NCs (Scheme 1).^{[ 48 ]} Figure 1b,d shows the absorption and CD spectra, respectively, of R/L‐Ag₂₉‐IP NCs in pyridine. The absorption profile exhibited slight changes around 360, 400, and over 520 nm after the addition of Ag‐PPh₃, reproducing the previous result using racemic Ag₂₉ NCs (Figure S5).^{[ 48 ]} In the CD spectra, the presence of Ag‐PPh₃ had a negligible effect on the peak positions but slightly modified the peak amplitudes. Specifically, an increase in the CD peak at 405 nm and decreases at 450, 390, and 360 nm were observed. These changes suggest a small alteration in the electronic structure of Ag₂₉ NCs. The |g _abs| values slightly decreased to 2.2 × 10⁻³ at 460 nm (Figure S4). The relative atomic positions between Ag₂₉ and Ag₂₉‐IP NCs are almost identical to each other except for the presence of Ag(I) complexes for the latter in their X‐ray crystal structures (Figure S6),^{[ 48 ]} explaining their similar CD profiles.

PL/CPL Study

The PL dynamics was investigated for Ag₂₉ NCs in DMF and Ag₂₉‐IP NCs in pyridine prior to the CPL study. The enantioseparated samples dispersed in the DMF solution of 10 mM PPh₃ exhibited a PL peak at 690 nm (Figure 2a). The excess amount of PPh₃ sub‐ligand stabilized the NCs^{[ 52} , ^{53 ]} affording moderate PLQY values of 7.6% and 5.3% for R‐ and L‐Ag₂₉ NCs, respectively (Figure S7). R‐Ag₂₉ NCs in DMF exhibited the apparent single exponential decay in the entire measurement range (∼1.4 µs) confirming the simple excited‐state relaxation pathway: the ultrafast and quantitative ISC (<0.5 ps)^{[ 48} , ^{50 ]} followed by the phosphorescence from T₁ state (∼1.4 µs, Scheme 1, Figure S8, Table S1). Meanwhile, the pyridine solutions of R‐ and L‐Ag₂₉‐IP NCs gave the identical PL peak at 770 nm with the enhanced PLQYs of 31% and 36%, respectively (Figures 2b and S7). The excitation spectrum of each solution almost reproduced the absorption profile (Figure S9). The PL decay curves in ∼100 µs range could apparently be fitted by single exponential functions, giving the PL lifetimes of 10.4 and 11.0 µs for R‐ and L‐Ag₂₉‐IP NCs, respectively (Figure S8). These observations are identical to what we observed for the racemic mixture of Ag₂₉‐IP NCs.^{[ 48 ]} However, a very rapid decay was observed for Ag₂₉‐IP NCs at 720 nm in the initial time window up to 10 ns, followed by the long‐lived component with the 10 µs lifetime appearing as an offset signal (Figure 3b). This rapid decay was modeled by a single exponential function with a time constant of 0.3 ns (fitted with a free parameter resulting in 0.51 ns, Figure S10), consistent with the timescale for the stabilization of T₁ toward T₁(IP) state (Scheme 1), which was observed in the transient absorption measurement.^{[ 48 ]} In contrast, such a rapid decay was absent in the PL decay (∼ a few ns) for R‐Ag₂₉ NCs in DMF (Figure 3a). These results verify that the 0.3 ns decay arises from the change in the excited‐state property in Ag₂₉‐IP NCs.

Figure 2.

a) and b) PL and c) and d) CPL spectra of a) and c) R/L‐Ag₂₉ NCs in DMF and b) and d) R/L‐ Ag₂₉‐IP NCs in pyridine. Blue: R‐Ag₂₉, red: L‐Ag₂₉, purple: R‐Ag₂₉‐IP, orange: L‐Ag₂₉‐IP NCs.

Figure 3.

Emission decay curves of a) R‐Ag₂₉ NCs in DMF (10 mM PPh₃) at 660 nm and b) R‐Ag₂₉‐IP NCs in pyridine at 720 nm, together with the instrument response function (IRF).

The chiroptical property change in the excited state was then investigated by the CPL measurement. The positive and negative CPL signals could be barely recognized for R‐ and L‐Ag₂₉ NCs in DMF (Figure 2c). The PL anisotropy factors defined as g _lum = 2(I _L‐I _R)/(I _L + I _R), in which I _L and I _R are the PL intensities of left‐ and right‐circularly polarized light, respectively, were estimated around ± 1 × 10⁻³ comparable with the detection limit of the home‐made CPL instrument.^{[ 54 ]} Interestingly, R/L‐Ag₂₉‐IP NCs displayed intense CPL signals at around 770 nm without changing their signs (Figure 2d). The |g _lum| values are as high as 6 × 10⁻³ at 770 nm for both enantiomeric NCs and among the highest ones for chiral NCs with CPL activity in solution.^{[ 33} , ³⁴ , ³⁵ , ³⁶ , ^{37 ]} In accord with the drastic change in the excited‐state dynamics, the surface modification of Ag₂₉ NCs thus had the greater impact on the excited‐state chirality rather than the ground‐state one.

Temperature‐Dependent CPL Study

According to the PL decay study, the excited state is assumed to evolve in the timescale of several hundred ps toward the T₁(IP) state for Ag₂₉‐IP NCs. However, the time‐resolved CPL spectroscopy operating at such a fast timescale with the acceptable precision (in terms of |g _lum| ∼ 10⁻³) has never been developed.^{[ 55 ]} We therefore performed VT‐CPL study as a complementary measurement. The VT‐PL measurement for Ag₂₉‐IP NCs suggested the existence of an activation barrier for the excited‐state stabilization of T₁ state, which could be overcome over 160 K (Figure S11).^{[ 48 ]} The relocation of Ag(I)‐species on the surface of NC may require the reorganizations of ligand conformations and solvating molecules, providing the causes of activation barrier in the excited state. The low‐temperature (T < 160 K) PL at 720 nm could be thus considered as the emission mainly from the original T₁ state before being stabilized to the T₁(IP) state. It is also considered to have the same origin as the rapid component (∼0.5 ns) observed in the PL decay study at 720 nm (Figure 3b). The VT‐CPL should therefore give an important insight into the evolution of excited‐state properties complementary to the time‐dependent behavior.

The pyridine solution of R‐Ag₂₉‐IP NCs was placed in a cryostat and subjected to CPL measurement in the temperature range of 120–260 K (Figure S12). The g _lum value increased a little (8 × 10⁻³) at 220 K compared to room temperature. Since the melting point of pyridine is ca. 230 K, the sample forms either a transparent supercooled solution or an opaque frozen solid at 220 K. Both states gave similar CPL signals (Figure S13), demonstrating that the macroscopic appearance of the sample did not influence the CPL signals. The CPL signals then clearly weakened with a blue‐shift as the temperature decreased from 220 to 160 K. The average g _lum value exhibited a sharp decrease below 220 K, dropping from 8 × 10⁻³ to ∼1 × 10⁻³ (Figure 4). The temperature‐dependent g _lum value plot well accords with the temperature‐dependent PL peak position (Figure S11). Considering the above discussion on the temperature‐dependent PL origin, the PL from the original T₁ state involves the less CPL activity compared to that from the ion‐pair stabilized T₁(IP) state. Furthermore, taking the similar values of g _lum into account, the initially formed T₁ state for Ag₂₉‐IP NCs in pyridine should have a similar property to that of Ag₂₉ NCs in DMF without the Ag(I)‐complex modifications free from the stabilization by ion‐pairing. However, since the intensity fraction for the initially formed T₁‐based PL is negligible above 160 K, the entire CPL performance is dominated by the characteristics of the long‐lived T₁(IP) state.

Figure 4.

a) VT‐CPL spectra measured at 220 (red), 200 (green), 160 K (blue) and b) plot of g _lum values for R‐Ag₂₉‐IP NCs in pyridine as a function of temperature. The g _lum was obtained by averaging values within a 50 nm range centered around the peak. The error bars correspond to the standard deviation of the values within that 50 nm range.

Theoretical Calculations

The mechanism for the stabilization of T₁ state and the concomitant enhancement of CPL property by the peripheral modification were investigated with the aid of DFT and TD‐DFT calculations. Since the calculations on the full NC structure with four Ag(I)‐complex moieties with PPh₃ and pyridine ligands on the surface were too computationally expensive to complete, the modifiers were simplified as Ag⁺ ions. To investigate the effect of the number of surface bound‐Ag⁺ ions on the T₁ state properties, we prepared two models for the Ag₂₉‐IP NC binding a single and four Ag⁺ ions as Ag₂₉‐1Ag⁺ and Ag₂₉‐4Ag⁺ NCs, respectively (Figures 5 and S14). The T₁‐optimized structures of these models were obtained by the method described in the Supporting Information with the aid of the X‐ray crystal structures.

Figure 5.

a) and b) Optimized structures of Ag₂₉ NCs and c) and d) corresponding hole (yellow) and electron (blue) distributions in the T₁ state for a) and c) [Ag₂₉(BDT)₁₂]³⁻ (Ag₂₉ NC) and b) and d) [Ag₂₉(BDT)₁₂‐Ag⁺]²⁻ (Ag₂₉‐1Ag⁺ NC) models. Color code in a) and b): pink = Ag; yellow = S; grey = C.

The effect of Ag⁺‐binding on the PL property was first evaluated in terms of the phosphorescence energies (E _T) in the presence and absence of Ag⁺ ions on the surface of Ag₂₉ NC (Figures 5 and S14). The E _T value was estimated to be 1.61 eV (769 nm) for the unmodified Ag₂₉ NC using the B3LYP functional,^{[ 56 ]} which lowers in energy than the experimentally observed phosphorescence peak position at 680 nm (Figure 2a). Regardless of the deviation from the experimental value, the binding of a single Ag⁺ ion (Ag₂₉‐1Ag⁺ NC) on the surface effectively reduced the calculated E _T value to 1.28 eV (967 nm). The presence of four Ag⁺ ions (Ag₂₉‐4Ag⁺ NC) also decreases E _T to 1.36 eV (Figure S14), demonstrating that the T₁ stabilization is almost independent of bound‐Ag⁺ ion numbers. The substitution of Ag⁺ with Na⁺ ions in the calculation models also yielded similar result regarding T₁‐stabilization (Figure S15). These findings are in line with experimental results showing that the modifications using either alkali metal ions (Na⁺, K⁺, Cs⁺)^{[ 47} , ^{49 ]} or Ag(I)‐complex^{[ 48 ]} commonly resulted in a PL red‐shift together with an increase in PLQY. This observation suggests the importance of the nearby positive charge. The use of the range‐separated hybrid ωB97XD functional^{[ 57 ]} also estimated the decreases in E _T values by 0.32 and 0.26 eV for Ag₂₉‐1Ag⁺ and Ag₂₉‐4Ag⁺ NC models, respectively. As the different types of functionals afforded the similar changes in E _T values, the stabilization of T₁ state by the binding of Ag⁺ ions was reliably demonstrated, supporting the experimental observation of PL redshift of 0.21 eV from 1.82 to 1.61 eV.^{[ 47} , ⁴⁸ , ^{49 ]}

To figure out the origin of the T₁ stabilization by the binding of Ag⁺ ions, the interaction energies between the NC body ([Ag₂₉(BDT)₁₂]³⁻) and Ag⁺ ions were analyzed in the S₀ and T₁ states based on the T₁‐optimized structures (Table S2). Since the NC‐body possesses –3 charge, the attractive interaction with cationic Ag⁺ ions was suggested both in the S₀ and T₁ states. Nevertheless, the interaction energies are larger in the T₁ state (−5.34 and −1.80 eV) than those in the S₀ state (−5.08 and −1.46 eV) for Ag₂₉‐4Ag⁺ and Ag₂₉‐1Ag⁺ NC models, respectively. This result also suggests that the PL redshift originates from the T₁ stabilization rather than the S₀ destabilization. The natural population analysis showed the increase in the partial atomic charge on the binding site sulfur atoms from −0.3 e for Ag₂₉ NC to −0.4 e for Ag₂₉‐(1 or 4)Ag⁺ NC models in the T₁ state. These results indicate the attractive interaction between the excited electrons and Ag⁺ ions on the Ag₂₉ NC surface in the excited T₁ state. It is noted that the partial atomic charges of Ag⁺ ions are estimated to be less than + 1 e (0.89 and 0.72 e for Ag₂₉‐4Ag⁺ and Ag₂₉‐1Ag⁺ NC models, respectively), suggesting the partial contribution of charge transfer interaction to the T₁ stabilization.

While the vertical transition from the S₀ state to the Franck–Condon state proceeds with a ligand‐to‐metal charge transfer character considering the spatial distributions of HOMO and LUMO in the ground state,^{[ 45} , ^{46 ]} the eventual relaxation to the T₁ state separately localizes the hole and electron on the core and surface cage, respectively, regardless of the presence of Ag⁺ ions (Figures 5c,d, S16 and Table S3). Such a charge transfer character was also reported for the T₁ state of [PtAg₂₈(BDT)₁₂(PPh₃)₄]⁴⁻ NC.^{[ 58 ]} The presence of Ag⁺ ions seems to lead to the further attraction of excited electrons toward the surface as demonstrated by the more surface‐localized nature of electron distribution in the Ag₂₉‐1Ag⁺ NC model (Figures 5d and S16). The slight assignment of electron distribution on the Ag⁺ ion also supports the partial contribution of charge transfer interaction to the T₁ stabilization as suggested in the natural population analysis. For the quantitative discussion, the relative distributions of hole and electron in the Ag₁₃ core and Ag₁₆(BDT)₁₂ (+Ag ions) exterior cage are analyzed in the T₁ state (Table S3). In the absence of Ag⁺ ions (Ag₂₉ NC), the Ag₁₃ core has the larger occupancy ratios both for the hole (63%) and electron (54%) than the surface cage. Meanwhile, the electron distribution in the surface increased from 45% to 67% for the Ag₂₉‐1Ag⁺ NC model, while the hole distribution also increased from 37% to 42%. This result suggests that the surface‐bound Ag⁺ ions attract the excited electron which concomitantly attracts the hole, leading to the migration of exciton toward the surface exterior cage.

The change in the chiral geometry of Ag₂₉ NC ([Ag₂₉(BDT)₁₂]³⁻ model) in the ground and excited states was then calculated in the presence and absence of Ag⁺ ions. For the quantitative evaluation, the continuous chirality measure (CCM) values^{[ 59} , ^{60 ]} were introduced. All the equilibrium structures in the S₀ and T₁ states afforded the CCM values of 32–33 (Table S4). The binding of Ag⁺ ions has little effect on the chiral geometry of Ag₂₉ NC in the ground state, corresponding to the similar X‐ray crystal structures and CD profiles observed in the presence and absence of Ag(I)‐complexes (Figures 1 and S6). Regardless of the presence of Ag⁺ ions, the CCM values of the T₁‐optimized structures were also similar. Considering the above results, the enhanced CPL observed in the cation‐modified Ag₂₉‐IP NCs can be attributed to the attraction of excitons to the surface wherein the chiral exterior cage is involved. Given the static binding of cationic species on the surface as demonstrated by the X‐ray crystal structures,^{[ 47} , ⁴⁸ , ^{49 ]} the immediate (< a few ps) formation of stabilized T₁(IP) state could happen without any energy barrier. However, the low‐temperature PL^{[ 47} , ^{48 ]} study clearly suggests the presence of an energy barrier for the T₁ stabilization. The dynamic nature for the binding‐dissociation of cationic species was also suggested by the ESI‐MS study.^{[ 47} , ^{48 ]} Taking those experimental results into consideration, a plausible explanation for the evolution of CPL is as follows (Scheme 2). The PL peak position (720 nm) at low temperature (<160 K)^{[ 47} , ^{48 ]} should be attributed to the initially formed T₁ state with lower energy than that for the unmodified Ag₂₉ NC (680 nm) by 0.1 eV, suggesting the moderate stabilization of T₁ state by the Ag(I)‐complexes. The inherent nature of the T₁ state with a relative localization of the excited electron on the surface should attract cationic species toward the NC body through cation–electron interaction. In practical systems, Ag⁺ ions come along with coordinating and solvating molecules, such as PPh₃ and pyridine. Considering the reorientation of polar solvent molecules around excited molecules usually completes within 50 ps,^{[ 61 ]} the relocation or translocation of modifiers, together with the reorganization of accompanying ligands, may take <0.5 ns and require thermal energy.^{[ 51 ]} This event serves as the origin of the energy barrier. To obtain further stabilization of 0.1 eV, the proximation of Ag⁺ ion by only less than 0.5 Å is needed (Figure S17). The proximate Ag⁺ ions then strongly interact with the exciton in the T₁ state and localize the emission process in the chiral exterior cage. The similar mechanism including the stabilization and manipulation of emissive charge transfer excited state upon the binding of Mg²⁺ cations was reported for a ruthenium polypyridyl complex tethering carboxyl groups as binding sites.^{[ 62 ]} Then, in the Ag₂₉ NC, as the T₁ state is stabilized through the attraction of Ag⁺ ions, the exciton or excited state migrates toward the chiral exterior cage, temporally developing the CPL property. The significance of excited‐state confinement to a specific site has also been demonstrated in the CPL performance of various helicene compounds.^{[ 63} , ⁶⁴ , ^{65 ]} Consequently, the confinement of excited state within the chiral exterior cage resulted in the enhancement of CPL properties of the Ag₂₉ NCs.

Scheme 2.

Schematic diagram of the evolution of CPL along with T₁‐state stabilization by the exciton (hole and electron)–cation interaction in Ag₂₉‐IP NC.

Conclusion

We have demonstrated that the CPL performance of intrinsically chiral Ag₂₉ NCs highly depends on the excited‐state property. In our previous studies, the interaction between the Ag₂₉ NC body and cationic species in the excited state was suggested, stabilizing the T₁ state with a charge transfer character. The experimental and theoretical investigations in the present study indicate that the presence of positive charges proximate to the surface of NC attracts the exciton toward the chiral exterior cage, leading to the eventual enhancement of CPL property. The present study sheds light on the importance of spatial control of excited‐state localization in the design of CPL active materials.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supplementary Information

Ishii W., Shiraogawa T., Ehara M., Sotome H., Miyasaka H., Kawai T., Nakashima T., Angew. Chem. Int. Ed.. 2025, 64, e202513118. 10.1002/anie.202513118

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.