Tailoring the Dynamics of Excited‐State Charge Transfer through Conformational Engineering to Improve Second Near‐Infrared Fluorescence for High‐Resolution Osteosarcoma Imaging

The dynamics of excited‐state charge transfer (CT) in second near‐infrared (NIR‐II) fluorophore proceeds with conformational change that govern fluorescence. Unveiling the relationship between CT dynamics and conformational change in excited state is of great fundamental significance in optimizing NIR‐II fluorescence but remains less explored. This study unveils the pivotal role of large conformational change in promoting the evolution of emissive CT state into nonemissive CT state in NIR‐II fluorophore (CA‐BBT). Spectroscopic and computational results reveal that large conformational rotation/twisting leads to a pronounced increase in the CT character in excited state. This heightened CT character in CA‐BBT enables a rapid evolution of emissive CT state into a nonemissive CT state within 1.4 ps, as observed by ultrafast spectroscopy. Subsequently, this nonemissive state dominates nonradiative decay, resulting in very low NIR‐II fluorescence. Preventing such detrimental evolution by constraining the conformational dynamics of CA‐BBT results in a 10‐fold enhancement of NIR‐II fluorescence, enabling high‐resolution dynamic visualization of vasculature within osteosarcoma. This study offers a profound understanding of the relationship between CT dynamics and conformational changes in NIR‐II fluorescence, presenting innovative perspectives to develop bright NIR‐II fluorophores.


Introduction
[3][4][5][6] Achieving superior imaging quality usually requires a NIR-II fluorophore with high photoluminescence quantum yield (PLQY).[9] This preference is primarily attributed to the ease of tailoring charge transfer (CT) transitions in D-A architecture, which showcase longer emission wavelengths compared to the locally excited (LE) state. [10,11]However, CT state features low radiative efficiency than the LE state, resulting in low PLQY. [12,13]Despite such fundamental hurdle, few examples of NIR-II fluorophores with CT transitions show high PLQY. [14]This apparent contradiction raises the intriguing question of how CT dynamics in excited state affects NIR-II fluorescence.
Conformational changes in excited state, proceeding with intramolecular CT, [15] affects the nature and dynamics of CT that governs NIR-II fluorescence.Moreover, conformational change (relaxation) is predictable to be very large, especially in NIR-II fluorophore due to its structural flexibility, leading to substantial nonradiative energy loss.This is partially responsible for the low NIR-II PLQY. [16,17]However, the exact role of CT state The dynamics of excited-state charge transfer (CT) in second near-infrared (NIR-II) fluorophore proceeds with conformational change that govern fluorescence.Unveiling the relationship between CT dynamics and conformational change in excited state is of great fundamental significance in optimizing NIR-II fluorescence but remains less explored.This study unveils the pivotal role of large conformational change in promoting the evolution of emissive CT state into nonemissive CT state in NIR-II fluorophore (CA-BBT).Spectroscopic and computational results reveal that large conformational rotation/twisting leads to a pronounced increase in the CT character in excited state.This heightened CT character in CA-BBT enables a rapid evolution of emissive CT state into a nonemissive CT state within 1.4 ps, as observed by ultrafast spectroscopy.Subsequently, this nonemissive state dominates nonradiative decay, resulting in very low NIR-II fluorescence.Preventing such detrimental evolution by constraining the conformational dynamics of CA-BBT results in a 10-fold enhancement of NIR-II fluorescence, enabling high-resolution dynamic visualization of vasculature within osteosarcoma.This study offers a profound understanding of the relationship between CT dynamics and conformational changes in NIR-II fluorescence, presenting innovative perspectives to develop bright NIR-II fluorophores.
in NIR-II fluorescence process remains under much debate.20] Therefore, unveiling the relationship between conformational change and excited-state CT dynamics holds both fundamental and practical significance, particularly for optimizing NIR-II fluorescence.However, this remains an avenue that requires further exploration.
Here, we identify the evolution from the initial CT state into an emissive or nonemissive CT state with conformational rotation/twisting in NIR-II fluorophore (CA-BBT), utilizing which produces ultrahigh NIR-II fluorescence brightness for highperformance osteosarcoma imaging (Figure 1).Femtosecond transient absorption (fs-TA) spectroscopy and quantum chemical calculations elucidate the evolution of the nonemissive CT state from the initial CT state within 1.4 ps through ultrafast conformational dynamics.Afterward, this nonemissive CT state decays to the ground state, resulting in almost diminished fluorescence.Based on this finding, we restricted conformational change by a nonengineering strategy to block the evolution into a nonemissive CT state, which produces a 10-fold enhanced NIR-II fluorescence for high-performance osteosarcoma imaging.

Molecular Design and Characterization
The NIR-II fluorophore (CA-BBT) was synthesized by coupling rigid carbazole and planar benzobisthiadiazole as donor and acceptor units, respectively (Figure 1 and S1, Supporting Information).Meanwhile, alkylating thiophene was employed as both π-conjugation linker and electronic-bridge to proceed intramolecular CT.The structure of CA-BBT was systematically characterized in Figure S2-S4, Supporting Information.

Photophysical Properties and Computational Calculations
CA-BBT in organic solvents show structureless absorption around 700 nm and NIR-II emission (Figure 2a).Compared with the slight absorption blueshift as the increase in solvent polarity, a sizable emission redshift reveals a much stronger CT characteristic in the singlet excited state (S 1 ) than in the ground state (Figure 2a). [21]Furthermore, natural transition orbitals (NTO) analysis confirms the involvement of the CT character in emission (Figure 2b), which coexists with the LE character.The incremental weights of the CT-emission with the increase in solvent polarity contribute to the decreasing PLQY in high polar solvents (Figure 2c). [22]However, an interesting abnormality is noticed when high polar dimethylsulfoxide (DMSO) produces a higher PLQY of 1.54% compared to low polar dimethyl formamide (DMF) (0.86%).This highlights that other factors besides solvent polarity might play a significant role in controlling the NIR-II fluorescence of CA-BBT.
Given the marked structural flexibility of CA-BBT, the restricted conformational change in viscous DMSO probably leads to such enhanced PLQY. [23]This hypothesis gains support from the enhanced emission of CA-BBT in Me-THF at a low temperature of 77 K (Figure 2d,e), a condition that imposes more rigorous restrictions on conformational dynamics due to the heightened viscosity. [24]The calculated conformational change and reorganization energy further clarify the conformationtailoring NIR-II fluorescence.The optimized geometry of S 0 and S 1 with the largest RMSD of 0.517 Å in DMF reveals significant energy loss in DMF through conformational vibration (Figure 2f ), which accounts for the vast Stokes shift and reduced PLQY in DMF (Figure 2a,c).To gain a deeper insight into the relationship between conformation and emission, total reorganization energy from S 1 to S 0 was calculated across varying solvent polarity (Figure 2g).CA-BBT in DMF shows the highest total reorganization energy (4868 cm À1 ), consisting of low-frequency and high-frequency vibrations.A quantitative analysis discloses the predominant contributions from dihedral angles (Figure 2h), which corresponds to low-frequency conformational rotation/twisting. [18]These results confirm large conformational rotation/twisting in highly polar solvents, which leads to the decreasing PLQY of CA-BBT.

Tailoring NIR-II Fluorescence via Conformational Engineering
Building upon the above findings, restricting conformational dynamics through nanoengineering strategy is validated to improve NIR-II PLQY of CA-BBT.As shown in Figure 3a, we achieved this by tightly encapsulating CA-BBT within the organic matrix (F-127) to form water-soluble nanoparticles (CA-BBT NPs), which also exhibit additional merits for bioapplications.We also adjust the mass ratio of the CA-BBT to the F-127 to explore the confinement effects toward conformational dynamics since small-sized NPs should have a more stringent confinement environment.The transmission electron microscopy (TEM) and dynamic light scattering (DLS) demonstrate the uniform spherical nanostructures with decreasing hydrodynamic diameters from ≈45 to ≈33 nm (Figure 3b and S5, Supporting Information).Different-sized NPs were abbreviated as CA-BBT NPs ≈ 33 and CA-BBT NPs ≈ 45, respectively.As expected, CA-BBT NPs ≈ 45 in strong polar aqueous solution present remarkably enhanced PLQY (6.4% vs. 0.86%) compared to free CA-BBT in high polar DMF (Figure 3c,d).Moreover, the more significant suppression of conformational dynamics within small-sized CA-BBT NPs ≈ 45 produces a higher PLQY from 6.4% to 7.4% in aqueous solution, which outperforms most available NIR-II fluorophores. [7]These results provide a conformation-tailoring strategy to improve NIR-II PLQY.
As an additional merit, restricting conformational dynamics of CA-BBT improves molar extinction coefficient from 7.8 Â 10 3 to 1.1 Â 10 4 M À1 As an additional merit, restricting conformational dynamics of CA-BBT improves the molar extinction coefficient from 7.8 Â 10 3 to 1.1 Â 10 4 M À1 cm À1 as the decreasing NP size (Figure S6, Supporting Information).Enhanced absorption of CA-BBT NPs ≈ 33 is attributed to the more significant orbital overlap of the π-conjugated system arising from the increased planarity/rigidity of conformation in small-sized NPs.This hypothesis could be validated by the more pronounced vibronic peak in emission spectra of small-sized CA-BBT NPs (Figure 3c).In other words, confinement effects restrict the conformational dynamics of CA-BBT and flatten its geometry for intense absorption.The small-sized CA-BBT NPs in aqueous solution with both high extinction coefficients and PLQY produce a 10-fold enhanced NIR-II fluorescence brightness (344 vs. 26) compared to CA-BBT in DMF upon 808 nm excitation.Notably, the prominent superiority of CA-BBT NPs in NIR-II fluorescence brightness (344 vs. ≈100) over clinically approved ICG upon 808 excitation highlights the enormous potential in NIR-II fluorescence imaging. [7]

Excited-State Charge Transfer Dynamics
Femtosecond transient absorption (fs-TA) spectra identify the rapid formation of nonemissive CT intermediate in highly polar DMF as the underlying reason for its ultralow PLQY.In contrast, this intermediate was not observed in the moderately polar THF or CA-BBT NPs.The following CA-BBT NPs refer to the CA-BBT NPs ≈ 33.In DMF, the rapid evolution from the negative SE signal around 960 nm into positive ESA in 1 ps reveals the short-lived feature of emissive S 1 .Notably, the isosbestic point appearing at 1036 nm with a non-zero value is a distinctive indication of the formation of novel excited species, [25,26] identified as nonemissive species due to the absence of SE signal at extended delay times.However, this nonemissive species is unobservable for CA-BBT NPs or CA-BBT in lower polar THF (Figure S7, Supporting Information).These phenomena roughly correlate the formation of nonemissive species to solvent polarity or aggregation.Combined with the calculation results (Figure 2h), the conformational dynamics caused by solvation or aggregation should be the underlying reason for the formation of a nonemissive intermediate.
Quantitative analysis of kinetic trace within the SE and ground state bleaching (GSB) region further establishes a correlation between the formation of the nonemissive intermediate and conformational dynamics. [27]Specifically, the representative kinetics trace around 975 nm within the SE region for CA-BBT in DMF gives a rise component (1.4 ps) and two decay components (Figure 4c).The rise time reflects the conversion from a singlet excited state to a nonemissive intermediate.The prolonged rise time observed in highly viscous DMSO correlates the formation of nonemissive intermediate with conformational dynamics (Figure S8, Supporting Information).This correlation gains additional support from the absence of a discernible rise time in CA-BBT NPs (Figure 4c), where conformational change is significantly suppressed.
The two decay times around 975 nm for CA-BBT in DMF match well with the recovery time for the GSB around 720 nm (0.2, 14.5 ps).This alignment allows their assignment to S 1 !S 0 fluorescence (≈0.2 ps) and the nonradiative decay from the nonemissive intermediate to S 0 (≈14.5 ps), respectively.Analyzing the kinetics trace for CA-BBT NPs within the SE region around 1002 nm reveals a rise and decay component. [28]The decay component corresponds to S 1 !S 0 fluorescence.The rise component is likely attributed to the vibrational relaxation from its higher vibrational level to the lower vibrational level of S 1 , as the excitation at 800 nm could populate CA-BBT to its higher vibrational energy level in S 1 .Based on the above investigations, Figure 4d depicts a proposed conformational motion-controlling NIR-II emission mechanism.The higher polar solvent stabilizes the emissive CT state at the low energy level, as evident from the redshifted emission with increasing solvent polarity (Figure 2a).Meanwhile, the large conformational twisting/rotation in strong polar solvent facilitates the formation of low-energy nonemissive CT state, which opens a facile nonradiative channel.In contrast, CA-BBT within NPs cannot evolve into a nonemissive CT state because the confinement spaces are too small to proceed a large conformational change in the excited state.Therefore, conformational dynamics in low polar solvent or NP is significantly inhibited, thus blocking access to the nonemissive intermediate.

Whole Body NIR-II Fluorescence Imaging
The high fluorescence brightness and photostability afford CA-BBT NPs great potential for noninvasive NIR-II fluorescence imaging in vivo (Figure S9, Supporting Information).A PBS solution of CA-BBT NPs was injected into female BALB/c mice through the tail vein and imaged through different longpass (LP) filters (980 LP, 1064 LP, 1300 LP, 1500 LP).After injection, clear visualization of vascular structures spanning the entire body became immediately apparent (Figure 5a).Moreover, a shift in the imaging window to longer wavelengths notably heightened the clarity of the targeted capillaries (red line), as substantiated by the smallest full width at half-maxima (FWHM) and the highest SBR in the 1500 LP region.These results prove the superior NIR-II fluorescence imaging ability of CA-BBT NPs in vivo.

NIR-II Fluorescence Microscopic Imaging
The NIR-II fluorescence imaging of CA-BBT NPs in the deepseated osteosarcoma model was further evaluated and compared with the clinically approved contrast ICG.A PBS solution of CA-BBT NPs or ICG (same mass dose) is injected through the tail vein.Fluorescence imaging results demonstrate that both the fluorescence brightness and the selective accumulation of CA-BBT NPs within the osteosarcoma region outperform ICG (Figure 5b).Moreover, CA-BBT NPs reach maximal accumulation after 60 h of injection and almost hold on at least 72 h (Figure 5c).This extended time window enables dynamic analysis and real-time monitoring of osteosarcoma formation and progression.Utilizing these advantageous properties, CA-BBT NPs enable clear visualization of the vessel network deep within the osteosarcoma over an extended monitoring period (Figure 5d).Notably, a remarkable FWHM of 0.41 mm and a high SBR of 1.59 of the tiny capillaries enable clear differentiation (Figure 5e), which is not achievable with ICG.Moreover, H&E staining examination of key organs reveals no discernible damage (Figure S10, Supporting Information).Additionally, there were no significant changes in blood cell counts (Figure S11, Supporting Information), demonstrating excellent biosafety of CA-BBT NPs for bioapplications.These results affirm the potential of CA-BBT NPs for high-resolution NIR-II fluorescence imaging in deep-seated tissues.

Conclusion
In conclusion, we have elucidated the relationships between excited-state conformational dynamics and the evolution of CT.We have demonstrated that the rapid evolution from the initial CT state to a nonemissive CT state, driven by large conformational twisting/rotation, is the key factor behind the significant reduction in PLQY in high polar solvent.Building on this finding, the nanoengineering strategy restricts the conformational dynamics, resulting in remarkably enhanced NIR-II PLQY.As an additional merit, this nanoengineering strategy flattens the CA-BBT conformation, improving the absorption coefficient.The high absorption coefficient and PLQY produce ultrabright NIR-II fluorescence brightness, which enables high-resolution dynamic visualization of vessel network within the osteosarcoma that is not achievable by clinically approved contrast ICG.Our results deepen the understanding of the intricate relationships between conformational dynamics and NIR-II fluorescence, which may produce a conformation-tailoring strategy to develop bright NIR-II fluorophores.

Experimental Section
General: All the chemicals and reagents utilized in this study were purchased from commercial suppliers (such as Aldrich and Adamas-beta) and used without further purification unless otherwise noted.Nuclear magnetic resonance (NMR) spectra were acquired using a Bruker Ultra Shield Plus 400 MHz.TEM images were obtained utilizing a transmission electron microscope (HT7700) with an operational voltage of 100 kV.DLS was performed on a particle size analyzer (NanoBrook 90Plus, Brookhaven Instruments Corporation).The steady-state near-infrared absorption spectra were collected on a SHIMADZU UV-3600 PLUS ultraviolet-visible-nearinfrared spectrophotometer.The NIR-II fluorescence spectra were acquired on a Fluorolog 3 spectrophotometer (Horiba) equipped with an 808 nm diode laser and an InGaAs NIR detector.The temperaturedependent PL decay measurements were conducted on a Fluorolog 3 spectrophotometer equipped with a cryostat (Oxford Instruments, Optistat DN2, UK).
Computational Details: The ground states and excited states of compounds were fully optimized with the density functional theory (DFT) method in Gaussian 09 by using B3LYP-D3(BJ) density functional and 6-31G(d,p) basis set.The long-range interactions were considered through single-point energy calculation with the CAM-B3LYP functional.The solvent environments were simulated by using the polarizable continuum model (PCM), [29] in which the equilibrium solvation method was applied for the geometry optimization and the nonequilibrium solvation one was adopted for the single-point energy calculations at the equilibrium geometry. [30]The state-specific method was chosen for analyzing the excited-state property. [31]The reorganization energy was analyzed by Dushin software. [32]The natural transition orbits of excited electrons were analyzed by Multiwfn* [33] and RMSDs were calculated by VMD programs. [34]s-TA Spectra: The fs-TA spectra were obtained using an amplified Ti: sapphire laser system that produces a fundamental 800 nm output with a pulse width of 120 fs and a 1 kHz repetition rate (Solstice Ace, Spectra-Physics).This laser system provides an average power of 7 W divided into two beams (7:3) for the pump and probe, respectively.The pump beam, ranging from 240 to 2600 nm, was generated using an optical parametric amplifier (TOPAS, Light Conversion).The other beam was focused in a 3 mm sapphire plate and filtered by a short pass filter (800 nm) to cut off the fundamental 800 nm laser pulse to produce a white light continuum (WLC), which was used as a probe pulse ranging from 420 to 850 nm (VIS) and from 840 to1600 nm (NIR).A computercontrolled high-precision translation stage controlled the pump-probe delay with a 14-fs delay precision and an 8 ns time window.The fs-TA spectra were measured using a quartz cuvette with a 2 mm path length.Before passing through the sample, the probe beam is split into signal and reference beams.The signal beam passed through the sample, while the reference beam was sent straight to the reference fiber-coupled spectrographs that monitored the fluctuations in the probe beam intensity.The pump-induced changes of transmission (ΔOD), a sequence of probe pulses with and without a pump, are generated using a chopper wheel on the pump beam and monitored using fiber-coupled spectrographs with linear array detectors.The entire setup was controlled by a PC using home-built software.A multi-exponential function was employed to fit the kinetic curves.

Figure 1 .
Figure 1.Schematic diagram illustrating the rational manipulation of conformational dynamics to produce bright nanoparticles for NIR-II osteosarcoma fluorescence imaging.

Figure 2 .
Figure 2. a) Normalized absorption and PL spectra in different solvents.b) Natural transition orbitals (NTO) describing the emission characters of the S 1 state in CA-BBT; the weights of the CT characters to the emission are included.c) PLQY of CA-BBT in different solvents.d) Temperature-dependent PL spectra of CA-BBT in Me-THF.e) Temperature-dependent PL intensity.f ) The optimized geometries at S 0 and S 1 states of CA-BBT with root mean squared deviations (RMSDs).g) Plots of reorganization energy versus normal mode wavenumber of CA-BBT in different solvents.h) Proportions of bond length, bond angle, and dihedral angle contributed to total reorganization energy (λ).

Figure 3 .
Figure 3. a) Schematic illustration of the restricted conformational dynamics of CA-BBT through nanoengineering strategy to produce bright CA-BBT NPs.b) DLS and TEM results of small-sized CA-BBT NPs.c) Size-dependent absorption and PL spectra of CA-BBT NPs compared CA-BBT in THF.d) PLQY of different-sized CA-BBT NPs.

Figure 4 .
Figure 4. a,b) Femtosecond transient absorption mapping and plot at different delay times.The inserted arrow indicates the signal change along the delay time.c) Representative kinetic traces within SE and GSB region.r and d represent rise and decay, respectively.d) Schematic illustration of the conformational motion-controlling NIR-II fluorescence.

Figure 5 .
Figure 5. a) In vivo NIR-II whole-body fluorescence imaging of C57BL/6 mice after injection of CA-BBT NPs.Cross-sectional fluorescence intensity profile measured along the red line and Gaussian fits.b) Time-dependent whole-body NIR-II fluorescence imaging of xenografted osteosarcoma mouse after intravenous injection of CA-BBT NPs and ICG.c) Time-dependent NIR-II fluorescence intensity within osteosarcoma region.d) NIR-II microstructural fluorescence imaging of osteosarcoma.e) Cross-sectional fluorescence intensity profile measured along the red line in (d) and Gaussian fits.