Novel NIR-II organic fluorophores for bioimaging beyond 1550 nm

Novel NIR-II organic fluorophores were designed and synthesized using an AIE and highly twisted donor–acceptor distortion strategy for bio-imaging beyond 1550 nm.


Introduction
Optical uorescence imaging is a promising modality for realtime monitoring of disease progression, drug delivery and image-guided surgery with high spatial and temporal resolution. 1 However, traditional uorescence imaging techniques mostly focused on the visible and near-infrared region below 900 nm, where imaging resolution and penetration depths were largely limited due to the photo-scattering, auto-uorescence and absorption of biological tissues. Hence, developing novel uorophores with longer emission wavelengths to improve imaging resolution for in vivo deeper imaging is crucial and still a great challenge. 2 During the past few years, molecular imaging in the second near-infrared region (NIR-II, 1000-1700 nm) has emerged as a powerful tool for the delineation and treatment of cancers. [3][4][5][6][7][8][9][10][11][12] The rst organic small-molecule dye CH1055 with 90% renal excretion for NIR-II bio-imaging was reported in 2015, 5 and opened up a new era for small molecule imaging. Deeper imaging depths (up to 3 cm), excellent temporal and spatial resolutions (50 FPS and 1 mm), and a higher tumor-to-normal tissue (T/NT) ratio (up to 15) were achieved in the NIR-II region. 5, 12 Especially, the NIR-IIb sub-window (1500-1700 nm) showed tremendous advantages of near-zero auto-uorescence, negligible scattering, and unparalleled tissue-imaging depths, and turned out to be a hot spot for in vivo uorescence bioimaging. 13 Very few inorganic NIR-IIb uorophores such as single-walled carbon nanotubes (SWNTs), rare earth doped nanoparticles, and quantum dots have been investigated for biosensing and bioimaging beyond 1500 nm. 7g,8f, [13][14][15][16][17] It is worth noting that organic FD-1080 J-aggregates were rst accomplished with high resolution imaging of the cerebral and hindlimb vasculature with uorescence emission tailing into 1500 nm with a quantum yield (QY) of 0.0545%. 17 The signal-tobackground ratio (SBR) was 3.3-fold higher than that of NIR-IIa (1300-1400 nm) imaging. However, small-molecule NIR-IIb uorophores (beyond 1500 nm) are still in their infancy. Shiing small-molecule NIR-II uorophore emissions, specically into the NIR-IIb sub-window, is a great challenge but crucial for their expansion to in vivo biomedical applications. Here, we have rationally designed and synthesized new uorescent probes HL1-HL3 based on our previously reported NIR-II uorophores H1 7d and Q4. 6a The hexyloxy chain substituted at positions R 1 and R 2 of thiophene not only served as a strong donor, but also signicantly increased the dihedral angle up to 45.5 between BBTD and thiophene for the S 0 geometries (Fig. 1). Among them, HL3 showed remarkable brightness, excellent AIE features with uorescence emission stretching to 1550 nm and with a quantum yield of 0.05%. Furthermore, in vivo imaging beyond 1550 nm of the blood vessels, cerebral vasculature, and lymph nodes was achieved for the rst time.

Results and discussion
A vast majority of small-molecule NIR-II contrast agents were synthesized using the donor-acceptor-donor (D-A-D) backbones, in which benzobisthiadiazole (BBTD) was used as the electron acceptor unit. 3-13 Nevertheless, the uorescence quantum yields of organic NIR-II chromophores in aqueous solution were relatively low. The rigid planar aromatic structures with tremendous intermolecular p-p stacking interactions and the dominant non-radiative decay may be attributed to the aggregation-caused quenching (ACQ) effect in lowbandgap materials. 18 A feasible solution is to fully use the brightness of the dihedral twisted NIR-II backbones with strong emission extending into the NIR-IIb region by the aggregation-induced emission (AIE) strategy. 18-22 Thus, three novel organic small-molecule NIR-II uorophores HL1-HL3 were designed (Fig. 1). 3,4-bis(hexyloxy)thiophene served as the rst donor (D1), and triphenylamine was utilized as the second donor (D2) and a building block of AIEgens. In addition, the inuence of the electron-withdrawing group nitrobenzene and the electron donating group aminobenzene on the whole backbone was also investigated. Density functional theory (DFT) was rst employed to calculate the electronic properties of HL1-HL3 using Gaussian 09 soware and the B3LYP/6-31G(d) method. For the optimized ground state (S 0 ) geometries, the twisted angles of HL1-HL3, H1 and Q4 were calculated. All dihedral angles of HL1-HL3 between BBTD and donor 3,4bis(hexyloxy)thiophene were $45 ( Fig. 1), exhibiting more distortion than that of Q4 ($1.9 ) and H1 ($0.3 ) (Fig. S1 †). The E gap of HL1 (1.78 eV) indicated short wavelength infrared characteristics (Table S1 †). Moreover, the E gap of HL2 and HL3 was 1.45 eV and 1.48 eV, respectively ( Fig. S2 †), lower than that of CH1055 (1.5 eV) with a typical NIR-II optical E gap , resulting in a hypsochromic shi compared to H1 (1.21 eV) and Q4 (1.12 eV) (Table S1 †).
The small-molecule uorophores HL1-HL3 were synthesized mainly through Stille coupling, Zn reduction, N-thionylanilineinduced ring closure and Suzuki coupling in 45-50% yield over 4 steps from compound 1 (Fig. 2 and ESI †). The structures were conrmed by 1 H NMR, 13 C NMR, MALDI-TOF-MS or ESI-MS ( Fig. S12-S26 †). The spectroscopic properties of HL1-HL3 in THF are shown in Fig. 3B and C, and it was found that their maximum emission wavelengths were $922 nm, 1062 nm and 1050 nm, respectively, which were consistent with the results of E gap calculated using the Gaussian 09 soware and 6-31G(d, p). HL3 exhibited a remarkable increase in uorescence intensity with a strong tail in the 1500 nm region (Fig. 3C). The molar extinction coefficients (3) of HL1-HL3 in THF were measured to be 8.3 Â 10 3 L mol À1 cm À1 , 4.1 Â 10 3 L mol À1 cm À1 and 7 Â 10 3 L mol À1 cm À1 , respectively. The QYs of HL1-HL3 in THF were measured to be 0.2%, 0.34% and 2% with IR-26 (QY: 0.5%) as a reference, respectively (Fig. S3 †). The AIE properties of HL1-HL3 were studied in the THF/water mixture solvents upon increasing the water volume fraction (f w ). As shown in Fig. 3A, HL3 exhibited extremely strong uorescence emission in THF/ water with 90% f w . To further conrm the AIE properties of HL1-HL3, the uorescence emission spectra with different f w s were subsequently obtained under 808 nm excitation (Fig. 3D, E and S4 †). The uorescence (FL) intensity of HL1-HL3 gradually decreased with the increase of f w from 0 to 40-50%, and increased sharply for HL3 from f w 50% to 90%, indicating a typical AIE characteristic. Meanwhile, no AIE characteristics were observed for H1 and Q4 under the similar conditions.
HL2 and HL3 dots with high monodispersity and homogeneity were prepared in amphiphilic DPPE-5KPEG (Fig. 4A, S5 and S6 †). HL3 dots were characterized for bioimaging applications by transmission electron microscopy (TEM), dynamic light scattering (DLS) and the zeta potential with an average size of $90 nm, a dynamic size of $120 nm (Fig. 4B) and $ À9.2 eV, respectively (Fig. S7 †). The encapsulation efficiency of HL3 dots was calculated to be $82% (Fig. S8 †). The maximum absorption wavelength was 750 nm (Fig. 4C). The maximum emission wavelength was centered at 1050 nm and tailed to 1600 nm Fig. 1 The chemical structure and optimized ground state geometries (S 0 ) of HL1-HL3 by using the Gaussian 09 software. PdCl 2 (dppf) 2 CH 2 Cl 2 and K 2 CO 3 . (Fig. 4D). The quantum yields were calculated to be $11.7% in the NIR-II region (1000-1700 nm) and $0.05% in the NIR-IIb region (beyond 1550 nm). The molar extinction coefficient (3) of HL3 dots in water was measured to be 9.3 Â 10 3 L mol À1 cm À1 . The reasonable quantum yields and NIR-IIb uorescence emission of HL3 dots encouraged us to explore their NIR-IIb imaging capabilities in vitro. HL3 dots exhibited excellent uorescence intensity beyond 1550 nm under 808 nm laser irradiation (1550 nm LP, Fig. 4D). As shown in Fig. 4E, the uorescence intensity of the HL3 dots showed no obvious changes in different media (FBS, PBS and water) under continuous 808 nm laser irradiation for 1 h (90 mW cm À2 ). L929 mouse broblast cells were applied to evaluate the potential cytotoxicity of HL3 dots using a standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay. High cell viability was observed even at a high concentration (100 mg mL À1 , Fig. 4F). The pharmacokinetics of HL3 dots were also investigated through the measurement of the blood circulation half-life. The blood half-life of HL3 dots was 114 min (Fig. S9 †). All these results have demonstrated that HL3 dots have superior stability and excellent biocompatibility, and are more applicable for NIR-IIb bioimaging (beyond 1550 nm) in vivo.
To explore the feasibility of HL3 dots as a novel AIE NIR-IIb probe for reliable imaging of the whole body and cerebral vasculature system in KM mice and C57BL/6 mice (n ¼ 3 per group), HL3 dots (200 mL, 1 mg mL À1 ) were then injected into KM mice via the tail vein, and NIR-II and NIR-IIb images of blood vessels were recorded using an InGaAs camera with different LP lters (1000 nm, 1250 nm, and 1550 nm) and different exposure times under 808 nm laser irradiation (90 mW cm À2 ). Aer 5 min post-injection, whole blood vessels were clearly visualized, and the hind limb vasculature was chosen for analysis via the Gaussian-tted full width at half maximum (FWHM) (Fig. 5). It was found that the imaging of HL3 dots in the NIR-II window (>1250 nm, 1250 nm LP) with a 50 ms exposure time exhibited highly superior resolution (Fig. 5B). An extended exposure time (500 ms) was needed for higher resolution NIR-IIb imaging beyond 1550 nm (>1550 nm, 1550 nm LP) (Fig. 5C). The FWHM values of the hind limb vasculature were 777 mm (1000 nm LP), 768 mm (1250 nm LP) and 719 mm (1550 nm LP), respectively. The signal-to-background ratio (SBR) of NIR-IIb imaging (SBR ¼ 2.5, 1550 nm LP) was much higher than that of NIR-II imaging (SBR ¼ 1.4, 1000 nm LP and SBR ¼ 1.8, 1250 nm LP) (Fig. 5). The cerebral vasculature system was also imaged using the NIR-IIb probe HL3 dots. The superior resolution of tiny vessels was obtained (Fig. 6). The corresponding SBR of NIR-II 1000 nm, NIR-II 1250 nm and NIR-IIb 1550 nm imaging was 1.4, 1.8 and 3.4, respectively ( Fig. 6D and E). The SBR of NIR-IIb imaging (1550 nm LP) was 2.4-fold higher than that of NIR-II imaging (1000 nm LP). The FWHM values of the vessels at the same position (red dashed line) for different LP lters were calculated to be  701 mm (1000 nm LP), 687 mm (1250 nm LP), and 562 mm (1550 nm LP), respectively. The NIR-II uorescence images obtained using 1000 nm, 1250 nm and 1550 nm LP were also evaluated at the same exposure time (200 ms) at a concentration of 0.8 mg mL À1 . It was found that HL3 saturated the detector with 1000 nm and 1250 nm long-pass lters under imaging conditions suitable for HL3 with a 1550 nm long-pass lter (Fig. S10 †). All these results indicated that HL3 dots have great potential for in vivo NIR-IIb imaging (beyond 1500 nm) at an extended exposure time.
The lymph node drainage plays a vital role in tumor metastasis. We next demonstrated the application of HL3 dots for lymph node NIR-IIb imaging (beyond 1550 nm). HL3 dots (15 mL, 1 mg mL À1 ) were injected intra-dermally at the le forefoot pad of KM mice (n ¼ 3 per group) (Fig. 7). The process of lymphatic drainage was monitored under 808 nm laser irradiation (90 mW cm À2 ) using 1550 nm LP. Lymphatic vessels were notably identied in 1 min aer injection. The popliteal lymph node was gradually lighted up, and both the popliteal lymph node and the subiliac lymph node were clearly visualized in 2 h. The diameter of the lymphatic vessel between the popliteal lymph node and the subiliac lymph node was calculated to be 533 mm via FWHM, and the SBR reached 4 (Fig. 7F). The SBR values of the popliteal lymph node and the subiliac lymph node were 2.5 and 3.2 (1000 nm LP), 2.9 and 3.4 (1250 nm LP), and 5.1 and 4.8 (1550 nm LP), respectively (Fig. S11 †). These results reveal that NIR-IIb imaging beyond 1550 nm of lymph node drainage can be achieved with a higher SBR.

Experimental section
All animal experiments were performed according to the Chinese Regulations for the Administration of Affairs Concerning Experimental Animals and approved by the Institutional Animal Care and Use Committee (IACUC) of Wuhan University. And all the experimental details are provided in the ESI. †

Conclusions
In summary, we have successfully synthesized a series of small molecule uorophores HL1-HL3 by introducing different donors and distortion groups. Among them, HL3 showed extremely stronger AIE characteristics and highly twisted donor-acceptor distortion. HL3 dots exhibited excellent water solubility, photo-stability and biocompatibility with a remarkable increase in NIR-II uorescence intensity with QYs of 11.7%   and 0.05% in the NIR-II (>1000 nm) and NIR-IIb region (>1550 nm), respectively, in water. Superior quality NIR-IIb imaging beyond 1550 nm of the whole body, cerebral vasculature and the lymphatic drainage system was demonstrated for the rst time with a higher SBR. It is hoped that this novel NIR-II uorophore HL3 obtained using an integrated AIE and D-A distortion strategy may become a practical strategy to develop smallmolecule NIR-IIb uorophores with the maximum emission wavelength beyond 1500 nm with a deeper penetration depth and higher resolution.

Conflicts of interest
There are no conicts to declare.