Sung Han Kim‡
a,
Shazid Md. Sharker‡b,
Haeshin Leeb,
Insik In*ac,
Kang Dae Lee*d and
Sung Young Park*ae
aDepartment of IT Convergence, Korea National University of Transportation, Chungju 380-702, Republic of Korea
bDepartment of Chemistry, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Republic of Korea
cDepartment of Polymer Science and Engineering, Korea National University of Transportation, Chungju 380-702, Republic of Korea. E-mail: in1@ut.ac.kr
dDepartment of Otolaryngology-Head and Neck Surgery, College of Medicine Kosin University, Busan, 602-702, Republic of Korea. E-mail: kdlee@ns.kosinmed.or.kr
eDepartment of Chemical and Biological Engineering, Korea National University of Transportation, Chungju 380-702, Republic of Korea. E-mail: parkchem@ut.ac.kr
First published on 22nd June 2016
Fluorescence and photothermal conversion mediated by near-infrared radiation (NIR) is reported for carbonized polydopamine nanoparticles. Carbonized polydopamine demonstrated excitation-dependent fluorescence emission, together with NIR-responsive photothermal conversion properties. The concentration-dependent photothermal heating from carbonized fluorescent carbon nanoparticles-polydopamine (FNP-pDA) induces hyperthermal killing of both cancer cell lines and bacteria in vitro. Although most of the dopamine moieties of polydopamine become dehydrated upon carbonization, the remaining dopamine-hydroxyl groups can confer adhesive properties. These fluorescent coatings are compatible with many substrates, and the surface passivation of FNP-pDA with polyethylene glycol improves quantum yield and extends fluorescence lifetimes. The novel infrared-responsive photothermal and fluorescent carbon nanoparticles reported here show promise for a range of potential biomedical and research applications.
Materials made from polydopamine (pDA), containing catechol and amine functional groups, have been widely studied in surface chemistry research where they have found use as adhesion layers. pDA is possessed of several distinctive materials-independent coating features; including biocompatibility, antibacterial surface properties, bio-mineralization potential and corrosion resistance.9 Recently, dopamine was used as a carbon source for the synthesis of yolk–shell carbon spheres. In this process, self-polymerized dopamine adhered followed carbonization, using a silica template that was eventually removed.10 Furthermore, carbonized pDA exhibit a graphite-like nanostructure giving distinctive D- and G-bands during Raman spectroscopy. It has been speculated that the 15 nm thicknesses of the 40–50 stacking layers in carbonized pDA offer substantial potential to the field of optoelectronics.11
Carbon-based materials, such as nanoscale reduced graphene oxide and carbon nano-tubes, have recently been applied in photothermal therapy (PTT), because of the ability of these materials to absorb light from the ultraviolet (UV) to near infrared (NIR) ends of the electromagnetic spectrum and convert it into heat through non-radiative decay.12,13 Although broad NIR absorption of carbon-based materials shows greater photothermal conversion efficiency, these carbon derivatives remain limited by poor colloidal stability, which requires surface modification/passivation using a stabilizing agent such as polyethylene glycol (PEG), or lipid. Furthermore carbon derived FNPs have similarly poor photoluminescence properties because of a surface ionic state that can become passivated following strong emission.14
In this report, we explored the multi-color photoluminescence properties of carbonized pDA both in solution, as well as in a coated surface in which carbonized pDA exhibits concentration-dependent photothermal heating in response to NIR irradiation. Passivation of the surface on carbonized pDA with PEG molecules terminated with amine groups, permitted distinguishable bright fluorescence. We also established that our material has promising antifouling properties that allow cultured cells to detach from coated substrate.
1H NMR spectra were recorded using a Bruker Advance 400 MHz spectrometer with deuterium oxide (D2O) and deuterium oxide (D2O) as the solvent. The UV-vis spectra were recorded using an Optizen 2020UV; Mecasys Co. XPS spectra were obtained using an Omicrometer ESCALAB (Omicrometer, Taunusstein, Germany) and photoluminescence (PL) spectra were obtained on a L550B luminescence spectrometer from Perkin Elmer. Static water contact angles were measured using a DO3210 (KRUSS Ltd., Germany), and the X-ray diffraction were recorded using an XRD Bruker AXS ADVANCE D-8. Using an infrared camera (NEC Avio, Thermo Tracer TH9100), particle size was measured with dynamic light scattering (DLS) (Zetasizer Nano, Malvern-Germany) and transmission electron microscopy (FEI, Netherlands). The NIR laser was 808 nm (PSU-III-LRD, CNI Optoelectronics Tech. Co. LTD, China). Raman spectra were investigated using a Laser Raman spectrophotometer (NRS-3200 Jasco, Japan).
Fluorescence lifetimes were measured using a NanoLED laser light source (Horiba Jobin Yvon NanoLog spectrophotometer) at 375 nm for the excitation, and the data were fitted by a multi-exponential decay model. The samples for the fluorescence lifetimes measurements were prepared by dissolving FNP-pDA and surface passivation of FNP-pDA in an aqueous solution at very low concentrations (1.0 mg mL−1). Quinine sulphate (QY 55 (%) at 354 nm excitation) was used as a reference standard to measure quantum yield, as adapted from a published report elsewhere.15
The synthetic routes reported for producing FNPs from pDA employ controlled carbonization in a strongly acidic environment for a predetermined period.12 This synthetic strategy was fixed to retain dopamine moieties on the prepared FNPs. These approaches produce FNPs that display characteristic fluorescent properties, and they were improved by including a surface passivation step using PEG, which resulted in improved photo-luminescent properties. A detailed account of the experimental processes involved is provided in Scheme 1.
Scheme 1 Illustration of the preparation and application of FNP-polydopamine (FNP-pDA) obtained from polydopamine (pDA). |
Absorption analysis from the UV to the NIR provided information about the chemical nature of pDA, FNP-pDA and the effects of surface passivation. It has been shown that an absorption maximum at 280 nm is mainly attributed to pDA. Although FNP-pDA obtained through carbonization had the same absorption band at 280 nm, the intensity was moderately decreased, which indicated a functional decline of the dopamine component.9 However, the intensity of the characteristically broad absorption between the visible range and the NIR (600–1000 nm) was similar to that of other carbonized nanoparticles, including other photothermal and hyperthermal agents.12 FT-IR analysis showed two distinctive broad peaks of high intensity at around 3400 cm−1 and 3100 cm−1, which might result from surface active N–H/O–H stretching vibrations of pDA. At the same time, the amide bond (N–H) shearing band (1510 cm−1), the aromatic ring CC vibration band (1442 cm−1) and the C–N stretching band (1292 cm−1), which are all characteristic of pDA, appeared in the FT-IR spectrum (Fig. S1†).19 Moreover, the appearance of high intensity, broad signals from C–H bonds at around 2800 cm−1 is consistent with the presence of surface-passivated PEG. 1H-NMR spectroscopy revealed an aromatic proton peak (6.5–6.7 ppm), which is characteristic of pDA in FNP-pDA (Fig. S2†).9
Carbonized materials often contain a mixture of sp2 and sp3 hybridized carbon atoms. The opto-electronic properties of such carbonized materials are strongly influenced by the number and spatial distribution of double bonds, and sites of delocalized electrons. Since the optical bandgap depends in part on the size, shape, and fraction of the sp2-hybridized species; tunable fluorescence may be achieved by modulating the number and distribution of such species.5 Following the above principle, the excitation-dependent emission profiles of FNP-pDA and surface-passivated FNP-pDA were evaluated at a fixed concentration. As we can see in Fig. 1c and d, the expected excitation-dependent fluorescence emission was seen for FNP-pDA, whereas its source precursor (pDA) did not show these properties. Moreover, when compared to FNP-pDA (1.33% QY blue and 4.77% QY green), it is clear that surface-passivated FNPs have not only retained similar fluorescence profiles, but have even higher intensity peaks. As demonstrated by the excitation-dependent fluorescence profiles, upon changing the surface ionic state of FNP-pDA with amine capped PEG, an increased QY was measured. This feature is indeed attractive, since it establishes that the brightness of carbon derived FNPs can be significantly increased (2.90% QY blue and 8.80% QY green). The fluorescence lifetimes (τ) of FNP-pDA and surface-passivated FNP-pDA are 5.96 ns and 8.13 ns, respectively, a finding which also demonstrates the sustainability of these materials for a range of envisioned applications in opto-electronics (Fig. S3†).14
Our transmission electron microscopy (TEM) measurements revealed that the resulting particles exhibited a spherical shape with an average diameter of 15 ± 2 nm, and that the lattice separations (0.308 nm) were consistent with those of graphitic carbon structures (Fig. 2a).8 The formation of surface-passivated FNP-pDA results in almost same lattice separations (0.308 nm) for what appears to be surface coated dark carbon materials (Fig. 2b). The particle size distributions of both FNP-pDA and surface-passivated FNP-pDA in the aqueous phase were evaluated by DLS. The size dispersion of the FNP-pDA was between 12 and 19 nm with an average size of 15.69 nm. In contrast, PEG-passivated FNP-pDA diameter sizes were much larger, ranging between 26 and 42 nm, with an average size of 34.7 nm. It seems likely that the pronounced differences in size between the two types of FNP reflect the behaviour of the PEG chains on particles surface.
An apparent G-band at 1590 cm−1 and D-band at 1360 cm−1 were observed in the Raman spectrum of both FNP-pDA and surface-passivated FNP-pDA (Fig. 3a), implying the presence of both sp2 and sp3 hybridized carbons in these nanoparticles.9 However such a Raman active bands are not seen for the native pDA precursor. The intensity ratio [ID/IG], which is often used to correlate the structural constituency of carbonized nanoparticles, also indicated different ratios of the sp2 and sp3 forms of carbon. Therefore, we concluded from the intensity of the D/G bands that pDA derived FNP-pDA and modified surface-passivated FNP-pDA were mainly mixture of sp2/sp3 form of carbons.13 Additionally, the characteristic X-ray powder diffraction (XRD) patterns (2θ) of surface-passivated FNP-pDA were observed at around 19°, 23° and 27°, due to the size increase wrought by PEG.19 The XRD pattern of FNP-pDA had decreased crystallinity relative to pDA, which may be accounted for by the increased variation in oriented carbon–carbon bonds arising from the mixture of sp2 and sp3 hybridized carbon atoms in these nanoparticles (Fig. 3b).
Owing to intrinsic absorption in the NIR region, carbon nanomaterials such as carbon nanotubes (CNTs) and graphene oxide (GO) have recently received much attention as efficient materials for photothermal cancer therapy (PTT).13,20 Carbonization of pDA broadens the NIR absorption bands. Furthermore XRD and Raman spectroscopy demonstrate the presence of sp2 hybridized carbon, similar to graphene oxide. Since NIR irradiation of FNP-pDA results in rapid photothermal heating (5 mg generates 45 °C), whereas native pDA remains insensitive to NIR radiation. Surface-passivated FNP-pDA also exhibits photothermal heating, but does so to a slightly lower extent than FNP-pDA (Fig. 4 and S4†). Temperature generation in these materials is als1o concentration-dependent. These findings suggest that the NIR absorption-dependent photothermal heating seen in our materials shows promise for several photothermal applications.
Compositionally, the carbonized materials contained of a mixture of sp2 and sp3 bonds and, photoelectric properties of such materials largely depend on the p states of the sp2 fraction. The strong localization of p and p* electronic levels of the sp2 domain lie within s and s* states, that essentially works as luminescence centers or chromophores.5 Those localized confined sp2 π-electron have ascribed photoluminescence, which cover the near infrared (NIR), visible and blue spectral ranges.6 Since the NIR light has long penetration ability with minimum tissue absorption, the infrared responsive carbon nanoparticles should allow monitoring of theragnosis from deep tissue.7 Moreover, the lattice of carbonized FNPs might also dissipating absorbed light (fluorescence) energy by non-radiative pathways such as vibrational relaxation that can be converted light into local heat influencing photothermal conversion (Fig. 5).5,8
PTT has been used to locally generate the temperatures required for vascular damage in the tumor environment. The ability to generate heat specifically in tumor tissues, avoiding surrounding compartments, is required for effective PTT.21 From a materials perspective, the characteristic absorption in the NIR window enables FNP-pDA to generate photothermal heating, which resulted in concentration-dependent in vitro killing of KB and MDCK cells. Cells viability assays using MTT showed a concentration-dependent decrease in the growth of cells. Treatment of cells with 1 mg mL−1 FNP-pDA resulted in 35% killing in response to 5 min NIR (808 nm) irradiation. The cell killing efficiency decreased moderately in the PEG-passivated FNP-pDA treated groups, due to lower absorption in NIR window. However, the viability of cell growth appeared unchanged in the control groups (unexposed to NIR radiation), suggesting that the toxicity of these nanoparticles stems predominantly from photothermal heating and not from direct chemical toxicity. To establish photothermal cytotoxicity further, FNP-pDA and surface-passivated FNP-pDA treated cells were studied by examining stained cells under confocal microscopy. Both NIR-irradiated FNP-pDA and surface-passivated FNP-pDA killed cells efficiently, as reflected in the number of dead cells (red color). In contrast, an increased number of live cells (green color) were observed in the control group, which were not exposed to NIR irradiation. Since the NIR absorption intensity and photothermal conversion efficiency of FNP-pDA increase in direct proportion to the concentration used, it is conceivable that the usefulness of FNP-pDA for hyperthermal tumor ablation could be optimized by adjusting the dosing and irradiation protocols during treatment.
To test the effect of photothermal treatment on bacterial cells, Staphylococcus aureus (Gram-positive) and Escherichia coli (Gram-negative) were premixed with FNP-pDA and surface-passivated FNP-pDA. NIR radiation (808 nm) was then used to irradiate the S. aureus and E. coli culture plates. Photothermal killing of bacteria was confirmed after 24 h by counting colonies. We found that bacterial killing efficiency depends on nanoparticle concentration. Fig. 6 shows the percentage of bacteria killed after 5 min NIR exposure; and illustrates that killing efficiency increased with sample concentration.22 However, the FNP-pDA treated agent exhibits more prominent antibacterial efficiency compared with surface-passivated FNP-pDA. After 5 min of NIR treatment with nanoparticles at a concentration of 0.5 mg mL−1, almost 80% of bacteria were killed in the FNP-pDA treated group; while in the group treated with surface-passivated FNP-pDA, only 60% were killed under identical experimental conditions. The weak photothermal conversion of surface-passivated FNP-pDA resulted in only moderate toxicity against the bacteria tested. However, FNP-pDA generated and released heat more efficiently, inducing irreversible bacterial damage. Moreover, when 1 mg of FNP-pDA was used against bacteria, 99.9% of bacteria were killed in a very short period, for both species (Fig. S5 and S6†). We conclude that we have demonstrated that our materials can also kill certain bacteria effectively, via a photothermal mechanism.
pDA and its derivatives can confer benefits on composite materials when used as independent coating agents. The adhesive properties of dopamine-scaffolds are widely used in the fabrication of a wide variety of coating agents on various substrates.18,23 To examine surface coating efficiency, the carbonized FNP-pDA and surface-passivated FNP-pDA composites were applied to polypropylene (PP), polyethylene terephthalate (PET), polyvinyl chloride (PVC), polystyrene (PS) and silicon (Si) substrates. As can be seen in Fig. 7a, the application of FNP-pDA and surface-passivated FNP-pDA coating on PP, PET, PVC, and PS substrate resulted in decreased hydrophobicity, which was confirmed by the decrease of the water contact angle.24,25 In contrast to those substrates, Si substrates showed increased water contact angles between 36° and 43° arising from the bare Si surface, which were more hydrophilic compare to FNP-pDA and PEG passivated FNP-pDA coating. However, all of the FNP-pDA coated substrates maintained 60–70° static contact, revealing the uniformity of the surface modification with FNP-pDA coating. In addition to contact angle, the fluorescence properties were evaluated from coated surface by using confocal laser scan microscope (CLSM) dependent on excited laser (nm). The fluorescence emission shows blue, green and red color.26 The observed light emission was also noticeable increased from surface-passivated FNP-pDA coating substrate in related with its precursor FNP-pDA coating (Fig. 7b). These results have revelled promising potentiality to draw fluorescence image on versatile substrate.
Having established that NIR excitation of FNP-pDA results in photothermal conversion, it was then necessary to examine the photothermal sensitivity of this surface.27,28 Irradiation of the FNP-pDA and surface-passivated FNP-pDA surface with the NIR laser demonstrated striking thermal elevation (54 °C) of the coated substrate (PET), which demonstrated the strong photothermal conversion ability of these materials (Fig. 7c).
The XPS C1s spectra of FNP-pDA and surface-passivated FNP-pDA are shown in Fig. 7d and e. The carbon moieties in the spectrum at 283 eV were assigned to CC; the binding energies at 284, 285.5, 286.2, 288 and 289 eV were assigned to C–C, C–N, C–O, CO and O–CO, respectively.12 The intensity of binding energy peaks for FNP-pDA, at about 283 eV for CC and 284 eV for C–C; decreased compared with surface-passivated FNP-pDA. The signals at 288 eV (CO) and 289 eV (O–CO) for surface-passivated FNP-pDA were decreased relative to FNP-pDA. These results demonstrate that both FNP-pDA and surface-passivated FNP-pDA are rich in sp2 and sp3 carbon sites (CC, C–C), which matches the expected constitution of carbon dots (CDs).12,14
Bio-fouling limits the performance of certain materials during biomedical application.29,30 Self-polymerized dopamine has been shown to automatically adhere to a wide range of substrates which could efficiently overcome antifouling properties, however carbonized FNP-pDA had not been examined. Our prepared FNP-pDA coated surface was evaluated by the quantification of nonspecific cell-adhesion and electron microscopic imaging. HeLa cells were cultured according to a previously reported method24 and imaged by optical microscopy. A detailed morphology report of the cell adhesion is presented in Fig. 8. Moderate (60–70%) adhesion of cells onto FNP-pDA coated PP and Si substrates, respectively. Unlike FNP-pDA coating, the surface-passivated FNP-pDA totally prevented cells adhesion owing to natural antifouling properties of PEG. The percentage of cells quantification shows less than 5% present on the PEG passivated FNP-pDA coated surface demonstrated antifouling strategy.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra08196g |
‡ Sung Han Kim and Shazid Md. Sharker contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2016 |