A Terrylene–Anthraquinone Dyad as a Chromophore for Photothermal Therapy in the NIR-II Window

A terrylenedicarboximide–anthraquinone dyad, FTQ, with absorption in the second near-infrared region (NIR-II) is obtained as a high-performance chromophore for photothermal therapy (PTT). The synthetic route proceeds by C–N coupling of amino-substituted terrylenedicarboximide (TMI) and 1,4-dichloroanthraquinone followed by alkaline-promoted dehydrocyclization. FTQ with extended π-conjugation exhibits an optical absorption band peaking at 1140 nm and extending into the 1500 nm range. Moreover, as determined by dielectric spectroscopy in dilute solutions, FTQ achieves an ultrastrong dipole moment of 14.4 ± 0.4 Debye due to intense intramolecular charge transfer. After encapsulation in a biodegradable polyethylene glycol (DSPE-mPEG2000), FTQ nanoparticles (NPs) deliver a high photothermal conversion efficiency of 49% under 1064 nm laser irradiation combined with excellent biocompatibility, photostability, and photoacoustic imaging capability. In vitro and in vivo studies reveal the great potential of FTQ NPs in photoacoustic-imaging-guided photothermal therapy for orthotopic liver cancer treatment in the NIR-II window.


General Methods
Instrument: All reactions of air-or moisture-sensitive compounds were carried out under argon atmosphere using standard Schlenk line techniques.Nuclear Magnetic Resonance (NMR) spectra were recorded in deuterated solvents using Bruker AVANCE III 400, Bruker AVANCE III 500 or Bruker AVANCE III 700 MHz NMR spectrometers.The 1 H and 13 C chemical shifts (δ) were recorded in parts per million and the TMS signal was used as an internal standard.Coupling constants (J) were recorded in Hertz with multiplicities explained by the following abbreviations: s = singlet, d =doublet, t =triplet, dd =double of doublets, m =multiplet, br=broad.Melting points were determined on a Büchi hot stage apparatus.High-resolution mass spectra (HRMS) were recorded by atmospheric pressure chemical ionization (APCI) on a MicrOTOF-QII instrument and by matrix-assisted laser decomposition/ionization (MALDI) using 7,7,8,8-tetracyanoquinodimethane (TCNQ) as matrix on a Bruker Reflex II-TOF spectrometer.UV-vis-NIR absorption spectra were measured on a Perkin-Elmer Lambda 900 spectrophotometer at room temperature.Fluorescence spectra were recorded by two fluorescence spectrophotometers (Horiba Jobin Yvon FluoroMax-4 NIR and Edinburgh Instruments FLS980) at room temperature.Dielectric spectroscopy was measured using a Novocontrol Alpha frequency analyzer (with a frequency range from 10 -2 to 10 7 Hz).A 1064 nm laser (Stone-laser LTD, Beijing, China) was used in photothermal irradiation.The temperatures of samples were recorded by an IR-thermal camera (Ti400, Fluke, USA).Transmission electron microscopy (TEM, JEM-3010, JEOL, Japan) images of nanoparticles were obtained on air-dried carbon-coated copper grids.Dynamic light scattering (DLS) was measured with a Malvern Zetasizer Nano instrument with compatible disposable capillary cell (DTS 1070 from Malvern).EVOS™ FL Imaging System was used to acquire confocal microscopic images.In vivo fluorescence imaging was conducted utilizing IVIS® Spectrum in vivo imaging system (PerkinElmer Inc., Waltham, Massachusetts).Multispectral optoacoustic tomography (MSOT) INVISIO-256 system (iThera Medical) was used in a phantom and in vivo imaging.

Synthesis details
Materials: All chemical reagents and solvents were purchased from Aldrich, Acros, ABCR, TCI and used as received without further purification unless otherwise noted.Thin layer chromatography (TLC) was performed on silica gel-coated aluminum sheets with F254 indicator and column chromatography separation was performed with silica gel (particle size 0.063-0.In a 25 mL Schlenk flask, compound 2 (980.0 mg, 1.0 mmol), 1,5diazabicyclo[4.3.0]non-5-ene(1.24 g, 10.0 mmol), and sodium tert-butoxide (480.0 mg, 5.0 mmol) were dissolved in diglyme (7 mL) under N2 atmosphere.The mixture was then heated at 70 °C for 2 hours.After cooling, the mixture was poured into 100 mL H2O and collected through filtration.The crude product was then purified using silica column chromatography with dichloromethane/hexane (2:1) as eluent.The product was obtained as a blue solid in a yield of 70%. 1

FTQ
Under N2 atmosphere, a 25 mL Schlenk tube was charged with sodium tert-butoxide (144.0 mg, 1.5 mmol), 1,5-diazabicyclo[4.3.0]non-5-ene(248.0 mg, 2.0 mmol), and TQ (50.0 mg, 0.25 mmol) in anhydrous diglyme (5 mL).The mixture was then heated at 130 °C for 16 h.After cooling to room temperature, the crude product was precipitated with 50 mL water and collected through filtration.The crude product was purified by a preparative GPC column with tetrahydrofuran as eluent.The final product was obtained as a grey solid in a yield of 60 % (30 mg).

Dipole moment measurement
Dielectric Spectroscopy (DS).The electric dipole moments were experimentally measured using dielectric spectroscopy.A Novocontrol Alpha frequency analyzer (with a frequency range from 10 -2 to 10 7 Hz) at 20ºC was employed.DS measurements were carried out in the usual parallel plate geometry of electrodes of 20 mm in diameter and a sample thickness of 100 μm (Teflon spacers).The complex dielectric permittivity  *  " (where ′ is the real and " is the imaginary part) was measured as a function of the solute concentration in chloroform.][5][6][7] In the case of both polar solute and solvent, according to the Böttcher equation, the dielectric permittivity is given as: where  refers to the dielectric permittivity of the solution,  is the number density of dipoles (  ; ρ is the mass density and M is the molar mass),  is the dipole moment,  is the refractive index and  is the molecular refraction in the limit of infinite wavelength ( . Indexes 1 and 2 stand for the solute and the solvent, respectively.The measured dielectric permittivity as a function of concentration is shown in Figure 1.Employing the slope at infinity solution, the derivative of  (eq.1), the dipole moments were calculated as  For the calculations, the refractive indexes were evaluated by measuring the dielectric permittivity of the samples in the bulk at very low temperatures (below 173 K).The corresponding values of the refractive indexes are  3.10 0.08 and  2.88 0.07, respectively.The density was evaluated from XRD data as explained in detail below.The uncertainty in dipole moments is strongly dependent on the uncertainty in density.

Differential Scanning Calorimetry (DSC).
The thermal properties of the samples were examined using differential scanning calorimetry (DSC) with a Q2000 (TA Instruments) equipped with a liquid nitrogen cooling system (LNCS) at a temperature range of 200 to 350 K.The temperature protocols involved a cooling scan and a subsequent heating scan at a rate of 10 Kꞏmin -1 .Samples were sealed in a Tzero aluminum low-mass pan and an empty pan was used as the reference.The instrument was calibrated in the specific temperature range for the baseline using a sapphire standard, and for the enthalpy and transition temperature employing an indium standard (ΔH = 28.71J/g, Tm = 428.8K, at a heating rate of 10 Kꞏmin -1 ).The heat flow traces of the two samples are presented in Figure S1.The TQ (precursor) is amorphous with a liquid-to-glass temperature at  241 5 K. FTQ has a Tg about 10 K higher than its precursor at  251 5 K associated with the more rigid structure and the suppressed rotational freedom.In addition, a weak crystal-tomelt transition, with an enthalpy of ΔH = 2.82 J/g, resembling liquid crystal transitions, is evident at a melting point  317 3 K.Small-Angle X-ray Scattering (SAXS).SAXS measurements were made with the N8 Horizon vertical setup (Bruker), using a 50W CuKα radiation (IμS micro-focus source with integrated MONTEL optics).The diffraction patterns were recorded on the VÅNTEC-500 2D detector (Bruker) at a sample-detector distance of 660 mm.The samples were placed in the form of powder within borosilicate glass capillaries with a diameter of 1 mm.Intensity distributions as a function of the modulus of the total scattering vector, q = (4π/λ) sin(2θ/2), where 2θ is the scattering angle and λ = 0.154 nm is the wavelength, were obtained by radial averaging of the 2D datasets.In the case of FTQ, three amorphous broad regions were evident at d = (2π/q) = 0.22 nm, d = 0.38 nm and d = 1.09 nm.The first two regions were also visible in the WAXS pattern of the precursor.The peak at d = 0.22 nm, corresponds to intramolecular (bond) distances whereas the peak at d = 0.38 nm corresponds to intermolecular (van der Waals) distances.Due to the large size of FTQ, the first Bragg reflection was detected within the low-q (SAXS) region (q ~ 2.4 nm -1 ).The pattern of the sharper peaks conformed to a simple orthorhombic unit cell with lattice parameters a = 2.68 nm, b = 0.35 nm and c = 2.72 nm.A few Miller indexes of characteristic peaks are shown in Figure S2a.The density was evaluated from the aforementioned lattice by assuming 100% crystalline material, as  1.37 0.05 g/cm .Because of the more flexible structure of TQ, the first peak was found at higher q (shorter distances, d) in comparison to FTQ.Considering that the two molecules have equivalent molecular mass, the difference suggests an increase in density.To obtain a broad estimate, we have further assumed the same orthorhombic lattice (with parameters a = 2.14 nm, b = 0.37 nm and c = 2.55 nm) and calculated the density, as  1.7 0.1 g/cm .It has to be noted that there is a large uncertainty in the density estimation, especially for the precursor, TQ.

Preparation of nanoparticles
The nanoparticles of FTQ were prepared via a matrix-encapsulation method.In brief, a mixture of FTQ (1 mg) and DSPE-PEG2000 (5 mg) in DMF (2 mL) was dropped slowly into 2 mL water using a micro syringe pump.After stirring for 12 h, the micelle solution was dialyzed against deionized water via a cellulose membrane (cutoff MW: 3.5 KDa) for 48 h.The product (FTQ NPs) was then collected through freeze drying at 0℃.The drug loading ratio (DLR) of FTQ NPs was calculated to be 12.9 % according to the equation (2) as shown below: DLR 100% (2)

Photothermal conversion efficiency (PCE)
The photothermal conversion capability of FTQ NPs was evaluated in water under irradiation with a 1064 nm laser (1.0 W cm -2 ) for 10 min.The temperature changes during the experiment were monitored every 10 s by an IR thermal camera.The photothermal conversion efficiency (η) was calculated according to the following equation ( 3): where h represents the heat transfer coefficient, A is the surface area of the container, ∆TMax represents the difference between the maximum steady-state temperature with the ambient temperature, Qs is the heat dissipation of solvent (water) which has been measured by a power meter (407A, Spectra-Physics), I represents the incident laser power (1.0 W cm -2 ), and A1064 represents the absorbance of FTQ NPs at 1064 nm.Herein, hA was calculated by the following equation ( 4): where mi and Ci are the mass (1.0 g) and heat capacity (4.2 J/g) of solvent (pure water), respectively.τs represents the sample system time constant which was calculated by the following equation ( 5): where T represents time.θ is the dimensionless driving force defined as (T-Tsur)/( Tmax-Tsur).

In vitro cytotoxicity assay
The cytotoxicity of FTQ NPs was tested by the CCK-8 assay.Hepa 1-6, MCF-7, and A549 cells were cultured in 96-well plates with 6×10 3 cells per well for 24 h.The FTQ NPs at different concentrations (0, 5, 10, 20, 30, 60 μg/mL) were added to the cell culture medium.After another 24 h incubation, the medium was removed and washed with PBS.Subsequently, CCK-8 (10 μL) was added into each well and the cells were incubated for 4 h.The absorbance at 450 nm was then measured using a microplate reader.

Staining of dead and living cells
The Calcein-AM and PI dyes assay kit was used to stain the Hepa 1-6 cells to verify the in vitro photothermal effect.Cells were evenly seeded in 96-well plates at a density of 6×10 3 cells per well and were incubated for 12 h at 37 °C in 5% CO2 atmosphere.
After different treatments (incubated with/without FTQ NPs and irradiated), all cells were filled up with a PBS buffer (1 mL per well) containing calcein-AM (2 μM) and PI (5 μM) for 20 min in the cell-cultured container.Then, all cells in 6-well plates were washed with PBS three times and observed by fluorescence microscopy.

Photoacoustic (PA) imaging in phantoms
The multispectral optical tomography system (MSOT in Vision 256, iThera8 Medical, Germany) was used to measure PA signal.PA imaging in phantoms was measured with agar as carrier.The samples were prepared by loading the mixture of control group (DI water) or FTQ NPs aqueous solution at different concentrations (2, 5, 10, 20, 30 and 60 μg/mL) with agar into the plantation.The distribution of photoacoustic signal intensity was tested at different excitation wavelengths (680, 685, 690, 695, 700, 710, 730, 750, 780, 800, 850 and 900 nm) in order to obtain the optimal excitation wavelength for photoacoustic imaging.

Tumor-bearing mouse model
All animal studies were conducted under the guidelines set by the Ethical Committee Peking Union Medical College and performed under legal protocols.Six-week-old female BALB/c mice were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd.The tumor-bearing mouse model was established by xenotransplantation of Hepa1-6 cells.In brief, Hepa1-6 cells (1×10 7 ) suspended in 50 μL of PBS were injected subcutaneously into the liver part of the mouse.After 10 days, the mice were used in the subsequent experiments.

Photoacoustic imaging in vivo
The tumor-bearing mice were injected with FTQ NPs (60 μg/mL, 100 μL) through the tail vein and placed in a dark box containing a 37 ℃ water tank after being anesthetized with 2% isoflurane in oxygen.The PA signals at the liver tumor site were collected using a multispectral PA tomography instrument as a function of post-injection times under the maximum excitation wavelength (680 nm, laser power: 5.02 m J).The mice injected with 100 μL 1×PBS were used as blank control.

In vivo photothermal therapy
Female BALB/c mice (6 weeks old) were chosen as the Hepa 1-6 cells orthotopic liver cancer model and the bioluminescence intensity is 10 8 .The mice were randomly divided into three groups (n = 10 per group), named as G1-G3.The mice injected only with PBS were selected as control group (G1).The other two groups were injected with FTQ NPs and treated without (G2)/with (G3) 1064 nm laser irradiation (1.0 W cm -2 ) for 10 min.For G1 and G3, after intravenous injection of pure PBS or FTQ NPs for 4h, the liver tumor site of each mice was continuously irradiated with 1064 nm laser (1.0 W/cm 2 ) for 10 min.In the meantime, the temperature changes of the tumors were recorded via a Fluke IR thermal camera.After different treatment for 20 days, the mice in G1-G3 were dissected.The liver organs were excised and paraffin-embedded, sectioned and stained by hematoxylin and eosin (H&E) for histopathological evaluation.

Figure S1 .
Figure S1.DSC traces of (a) FTQ and (b) TQ obtained during cooling (blue) and subsequent heating (red) at a rate of 10 Kꞏmin -1 .A weak crystal-melt transition is evident in TQ.The vertical blue and red arrows indicate the crystallization and melting points, respectively.The gray arrows indicate the liquid-to-glass transition.

Figure S2 .
Figure S2.SAXS (purple) and WAXS (black) patterns of (a) FTQ and (b) TQ at 293.15 K.The intensities were normalized to the maximum intensity of the most pronounced peak.In the insets are presented the 2-D intensity distribution in the detector from SAXS.The blue, red and green lines indicate the deconvolution of the amorphous regions (WAXS).(a) The sample presents a low degree of crystallinity with the first Bragg reflection in the low-q (SAXS) region.(b) The sample is largely amorphous.The peak in the low-q region associates with the average packing distance.
were performed with a D8 Advance Bruker diffractometer, CuKα (40 kV, 40 mA) radiation, equipped with a secondary beam graphite monochromator.The system employed a Bragg-Brentano geometry in a θ−θ configuration.Patterns were obtained over the range of 2θ from 2 deg in steps of 0.01 deg, and the rate was 32 s per step for all samples.The recorded intensity distributions are presented as a function of the modulus of the scattering vector (λ = 1.54184 nm).Scattering curves were taken at a temperature of 293 K.The WAXS patterns together with the SAXS patterns of the two samples are illustrated in FigureS2.
Figure S8.a) Average diameters of FTQ NP in PBS solutions (pH = 7.4); b) Average zeta potentials of FTQ NP in PBS solutions (pH = 7.4).Error bars, mean ± SD; c) Optical photographs of FTQ NP in PBS solutions (pH = 7.4); d) Absorption spectrum of FTQ NP in water.

Figure
Figure S10.a) Photothermal heating curves of FTQ NPs (60 μg/mL) under 1064 nm irradiation (1 W cm -2 ) for 10 min followed by cooling to room temperature.b) Linear correlation of the cooling times versus negative natural logarithm of θ. θ is the dimensionless driving force defined as (T-T sur )/( T max -T sur ); c) Photothermal conversion of FTQ NPs under 1064 nm laser irradiation at different concentrations (0-60 μg/mL); d) Photothermal conversion of FTQ NPs (60 μg/mL) under 1064 nm laser irradiation with different intensity (0.1-2 W cm -2 ); e) Average PA signal intensity as a function of nanoparticle concentration (coefficient of determination, R 2 =0.99).

Figure S11 .
Figure S11.The pathological examination by H&E staining of major organs collected from healthy mice after injection of FTQ NPs via tail.Scar bar: 150 μm.

Figure S12 .
Figure S12.The serum biochemistry and complete blood panels analysis (WBC represents white blood cell, RBC represents red blood cell count, HGB represents hemoglobin test, HCT represents hematocrit test, MCV represents mean corpuscular volume, MCH represents mean corpuscular hemoglobin, MCHC represents mean corpuscular hemoglobin concentration, RDW represents red cell distribution width, MPV represents mean platelet volume, PDW represents platelet distribution width).Healthy female Balb/c mice intravenous injected with aqueous solutions of FTQ NPs (dose = 45 mg/kg) were sacrificed after 7 and 14 days for blood collection.Untreated healthy mice were used as control.Error bars, mean ± SD.

Figure S13 .
Figure S13.Photothermal stability of ICG upon 808 nm laser irradiation of 1 W cm −2 for five on/off cycles, serving as a comparison of FTQ NPs.

Table S2 .
Summary of electrochemical data HOMO a LUMO a E gap a HOMO b LUMO b E gap b E gap c a) DFT calculation result (B3LYP/6-31 G*).b) Data from cyclovoltammetry using Fc/Fc+ standard (carried out in DCM containing 0.1 M n-Bu4NPF6 as supporting electrolyte at room temperature.A glassy carbon electrode was S12 used as a working electrode, a platinum wire as a counter electrode, and a silver wire as a reference electrode).c) Optical bandgap calculated from the absorption onset.

Table S3 .
Photo-indexes values /  ;  represents the sum of positive response rate;  represents the largest positive response rate;  represents the smallest positive response rate.