Research Article www.acsami.org
Reaction-Driven Self-Assembled Micellar Nanoprobes for Ratiometric Fluorescence Detection of CS2 with High Selectivity and Sensitivity Wei Lu, Peng Xiao, Zhenzhong Liu, Jincui Gu, Jiawei Zhang,* Youju Huang, Qing Huang, and Tao Chen* Key Laboratory of Marine Materials and Related Technologies, Ningbo Institute of Material Technology and Engineering, Chinese Academy of Science, Ningbo 315201, China S Supporting Information *
ABSTRACT: The detection of highly toxic CS2, which is known as a notorious occupational hazard in various industrial processes, is important from both environmental and public safety perspectives. We describe here a robust type of chemical-reaction-based supramolecular fluorescent nanoprobes for ratiometric determination of CS2 with high selectivity and sensitivity in water medium. The micellar nanoprobes self-assemble from amphiphilic pyrene-modified hyperbranched polyethylenimine (Py-HPEI) polymers with intense pyrene excimer emission. Selective sensing is based on a CS2-specific reaction with hydrophilic amino groups to produce hydrophobic dithiocarbamate moieties, which can strongly quench the pyrene excimer emission via a known photoinduced electron transfer (PET) mechanism. Therefore, the developed micellar nanoprobes are free of the H2S interference problem often encountered in the widely used colorimetric assays and proved to show high selectivity over many potentially competing chemical species. Importantly, the developed approach is capable of CS2 sensing even in complex tap and river water samples. In addition, in view of the modular design principle of these powerful micellar nanoprobes, the sensing strategy used here is expected to be applicable to the development of various sensory systems for other environmentally important guest species. KEYWORDS: carbon disulfide, dithiocarbamate, excimer emission, ratiometric fluorescent probe, self-assembled micellar nanoprobe
■
INTRODUCTION The detection of environmentally important volatile organic compounds (VOCs) has attracted tremendous attention over the decades. Among various VOCs, carbon disulfide (CS2) may be of particular interest because continuous exposure to even extremely low concentrations of CS2 can cause severe multiorgan disease, such as neurological, cardiovascular, ophthalmological, endocrinological, nephrological, gastrointestinal, and reproductive dysfunction.1 However, CS2 is nowadays widespread in environmental water where it is generated from both anthropogenic and natural processes.2 For example, CS2 is being heavily consumed in various industrial processes such as the production of resins, rubber, viscous rayon fibers, and carbon tetrachloride.3 It has also been widely used as an important insecticide in agriculture and food industry to protect fresh fruit and stored grain from insects and fungus.2 Additionally, CS2 is a primary degradation product of many extensively used sulfur-containing industrial chemicals, e.g., xanthate, dithiocarbamates, or NH4SCN, and also released naturally from marshland, food wastes, volcano eruptions, coal, and so on.4−6 Therefore, a maximum contaminant level for CS2 at 2 mg/L (26 μM) in drinking water is suggested in order to prevent potential environmental disasters.7 HPLC/GC-based instrumental analyses and colorimetric assay are currently used for CS2 sensing. Instrumental analyses,8 © 2016 American Chemical Society
albeit state-of-the-art, generally require bulky and expensive instruments and trained operators. The widely used colorimetric assay based on diethylamine-cupric acetate systems,9 however, usually suffers from relatively low sensitivity. The interference of H2S is also a thorny problem because it can lead to the precipitation of copper ions. Recently developed chemiluminescence-based method is promising for CS 2 detection but requires the fabrication of large and complex equipment and processes.2 Alternatively, fluorescence-based approaches could be attractive for CS2 sensing because of their high sensitivity, fast response, and technical simplicity.10−16 However, fluorometric methods for CS2 detection in water medium remain underdeveloped.17 The challenge in this field remains the creation of efficient fluorescent sensing materials for highly selective tracking of CS2 in complex environmental water samples with excellent sensitivity, given that CS2 concentration in real-world water is usually as low as tens of micromoles per liter, and other sulfur-containing compounds such as H2S, S2−, thioether, thioalcohol, and others usually coexist. Received: May 31, 2016 Accepted: July 15, 2016 Published: July 15, 2016 20100
DOI: 10.1021/acsami.6b06472 ACS Appl. Mater. Interfaces 2016, 8, 20100−20109
Research Article
ACS Applied Materials & Interfaces
Figure 1. Design principles of this type of chemical-reaction-based supramolecular fluorescent micellar nanoprobe. Amphiphilic Py-HPEI polymers self-assemble into polymer micelles. The micellar solutions emit intense excimer fluorescence of π-stacked pyrene dimers. When CS2 is present, it will quickly react with the amino groups of the hydrophilic micellar exteriors and spontaneously “migrate” into the hydrophobic micellar interiors in the form of dithiocarbamate (−NCSS) groups and bring the fluorophores (π-stacked pyrene dimers) and dithiocarbamate moieties (quencher) into close spatial proximity, which will result in the reduction of excimer emission.
micelles in water. The micellar solutions emit intense excimer fluorescence of π-stacked pyrene dimers because the encounter probability of the excited-state pyrene monomer and groundstate one is greatly increased within the confined micellar inner cores. When CS2 is present, it will quickly react with the amino groups of the hydrophilic micellar exteriors and spontaneously “migrate” into the hydrophobic micellar interiors in the form of dithiocarbamate (−NCSS) groups and brings the fluorophores and dithiocarbamate moieties into close spatial proximity. It is recently reported that the thiocarbonyl-containing groups such as thioamide (−N−CS) usually act as strong quenching entities for many fluorophores, i.e., quinolone, fluorescein, and coumarin, through a photoinduced electron transfer (PET) mechanism,36,37 Hence, the close proximity of the fluorophore (π-stacked pyrene dimers) and quencher (dithiocarbamates) results in the reduction of excimer emission and generates a fluorescence signal proportional to the amount of CS2 present in water samples. Moreover, taking advantage of the different fluorescence response of the pyrene monomer and excimer emission, a desirable CS2-triggered ratiometric fluorescence response can be realized through special molecular design. This ratiometric approach, which can provide built-in correction of two emission bands for environmental effects, gives the potential of precise measurement of CS2 concentration in complex real-world water medium.38−41 To the best of our knowledge, this type of self-assembled micellar nanoprobes is the first reaction-based supramolecular fluorescent system enabling ratiometric determination of CS2 with high sensitivity and selectivity. It is characterized with simple preparation, fast response, and complete water solubility as well as a wide linear detection concentration range.
The recently developed reaction-based sensory strategy may be a promising solution to this difficult problem because it operates with superior detection specificity,18−20 which is derived from an analyte-specific chemical reaction. Therefore, the reaction-based method is particularly suitable for application in complex systems and shows promising results for several target species.21−30 In contrast, supramolecular fluorescent probes based on self-assembled polymeric assemblies have attracted increasing attention because of the improved detection sensitivity.31,32 Their signal amplification and enhanced binding efficiency for guest species usually come from functional cooperativity and adaptability of multiple recognition sites within specially tailored self-assembled polymer aggregates. Moreover, the supramolecular fluorescent nanoprobes could offer extra advantages such as excellent aqueous dispersibility, biocompatibility, and structural and luminescent stability compared with those of the smallmolecule counterparts.18,33 Therefore, it is highly desirable to combine the advantages of chemical-reaction-based approach and polymer-assembly-based supramolecular nanoprobes within one ensemble for both selective and sensitive detection of CS2, especially in complex real-world systems, i.e., domestic and river water. Herein, we report a reaction-based supramolecular fluorescent sensing strategy for CS2 utilizing the self-assembled micelles of pyrene-modified amphiphilic hyperbranched polyethylenimine polymers (Py-HPEI). The probe design relies on a CS2-specific chemical reaction to trigger the fluorescence response of pyrene excimer emission.34,35 As shown in Figure 1, amphiphilic Py-HPEI with specially designed hydrophilic− hydrophobic ratios can spontaneously assemble into polymer 20101
DOI: 10.1021/acsami.6b06472 ACS Appl. Mater. Interfaces 2016, 8, 20100−20109
Research Article
ACS Applied Materials & Interfaces
structures.42 HBPs are typically characterized with non-/lowmolecular chain entanglements, low solution viscosity, and a large number of highly reactive functional groups. Therefore, the chemical reactions between CS2 and amino groups could be significantly accelerated by the high density of functional amino groups linked at both the linear and terminal units of HPEI. The subsequent spontaneous “migration” process of the hydrophobic dithiocarbamate moieties into micellar interiors is also facilitated by the extremely low polymer chain entanglements and solution viscosity of HPEI. As a result, the equilibration response time of these micellar nanoprobes will be greatly shortened, and a fast fluorescence response is realized. In this study, HPEI with a weight-average molecular weight of 10 000 was chosen because larger or smaller molecular weight Py-HPEI cannot form very stable polymer micelles in water medium. Another important probe design is the employment of πstacked pyrene dimers as signal reporters in order to achieve satisfying detection sensitivity. Besides a long-lived excited state, high quantum yields, and chemical stability, the spatially sensitive fluorescent dyes, e.g., pyrene and BODIPY FL, are well-known to exhibit enhanced excimer fluorescence output at a substantially longer wavelength than monomer emission when located close enough.43,44 It has also been shown that for pyrene-containing polymers the intensity of the excimer emission can be further amplified by both inter- and intramolecular aggregation of pyrene monomers due to a greater probability of dimerization than small-molecule counterparts.45 These advantages make these bis-pyrene labels versatile for sensitive fluorescence analysis of various guest molecules.43,44 More importantly, the presence of both pyrene monomer and excimer emission bands as well as their different fluorescence response features endow our micellar nanoprobes the potential to work as ratiometric sensors for CS2. This is very desirable for quantitative fluorescence analysis of CS2 in complex real-world water samples because this ratiometric approach can effectively eliminate most or all interferences from environment by built-in correction of two emission bands.38 Therefore, on the basis of the possible modulation of the luminescent and sensing features via changes in the content of functional pyrene moieties, a family of Py-HPEI polymers were synthesized. The entire synthesis was a continuous two-step reaction over 24 h at room temperature in ethanol (Scheme S1). The reaction mixtures were dialyzed against deionized water for 3 days and then concentrated under vacuum to give the pure products in high yield (>80%). As shown in Table S1, the content of pyrene fluorogens in the obtained polymers is in direct proportion to the HPEI/Py feed ratios, demonstrating that the conversion of 1-pyrenecarboxaldehyde is quite high. All of these polymers are readily soluble in deionized water and common organic solvents such as ethanol, DMSO, DMF, CHCl3, and THF. Their chemical structures are characterized by 1H NMR spectroscopy in CDCl3. As summarized in Figure S1, the typical resonance peaks ascribed to aromatic protons of pyrene groups appear at 7.95−8.45 ppm, whereas the signals of methylene protons of polyethylenimine are noticed at 2.1−2.9 ppm. Moreover, typical signal peak of pyrenylmethyl protons (Ph−CH2−) is also evident around 4.5 ppm for all these four Py-HPEI polymers, suggesting the covalent grafting of pyrene chromophores onto HPEI polymers. Thin layer chromatography (TLC) analysis shows no indication of 1-pyrenemethanol (the reduction product of 1-pyrenecarboxaldehyde) in the
Remarkably, it is proved to be highly tolerant to the usually coexisting H2S species. These promising advantages promoted us to further explore the feasibility and practicability of its use in complex real-world systems, including both tap and river water.
■
EXPERIMENTAL SECTION
Instruments. 1H NMR spectra were measured on a Bruker Advance AMX-400 spectrometer in CDCl3 using tetramethylsilane as the internal reference. Steady-state fluorescence spectra were recorded on a Hitachi F-4600 spectrofluorometer under controlled conditions equipped with a Xenon (Xe) lamp (150 W) with an excitation wavelength of 341 nm. The UV−vis absorption spectra were conducted on a PerkinElmer Lambda 950 UV−vis−NIR spectrometer in a 10 mm path length cell. Transmission electron microscopy (TEM) images were recorded with a JEOL JEM-2100F microscope at an accelerating voltage of 200 kV. TEM samples were prepared by dropping the diluted pure water solutions on a carbon-coated copper grid and dried under ambient temperature before testing. Dynamic light scattering (DLS) measurements were conducted at 25 °C using a Malvern Zetasizer Nano ZS particle size distribution analyzer. Materials. 1-Pyrenecarboxaldehyde (98%) and allyl methyl sulfide (98%) were obtained from Energy Chemical Co. Hyperbranched polyethylenimine (HPEI, Mw ≈ 10 000, 99%), 1-mercaptododecane (98%), and carbon disulfide (99%) were purchased from Aladdin Shanghai Reagent Co. Sodium borohydride (98%), analytical pure methanol, ethanol, ethylene glycol, diethyl ether, tetrahydrofuran, 1,4dioxane, anisole, trimethylamine, pyridine, hexane, dichloromethane, chloroform, benzene, toluene, xylene, N,N-dimethylformamide, N,Ndimethylacetamide, ethyl acetate, acetone, N-methyl-2-pyrrolidone, acetonitrile, and dimethyl sulfoxide were purchased from Shanghai Sinopharm Chemical Reagent Co. 2-Hydroxyethyl disulfide (98%) and thiophene (98%) were obtained from J&K Chemical Co. and TCI (Shanghai) Chemical Co., respectively. Carbon disulfide was distilled before use. The other chemicals and reagents were used without further purification. Synthesis of the Pyrene-Modified Polyethylenimine (PyHPEI) Polymers. A typical synthetic procedure is described as follows. Into the solution of 1.89 g of polyethylenimine (PEI, Mw ≈ 10 000) in anhydrous ethanol (50 mL), 1-pyrenecarboxaldehyde (210 mg, 0.9 mmol) was added. The mixture was stirred at room temperature for 12 h. Excess sodium borohydride (0.6 g, 16.2 mmol) was then added in batches. After being stirred at room temperature overnight, the mixture was dialyzed (with a molecular weight cut off of 3500 g/mol) against deionized water for 3 days. The solution was then concentrated under vacuum to afford Py-HPEI as a pale yellow semisolid in 89% yield based on polyethylenimine. Other Py-HPEI polymers were obtained using the same synthetic procedure. 1H NMR (CDCl3, 400 MHz, ppm) 1.75−2.98 (N−CH2, multiple strong peaks), 7.99−8.51 (ArH, multiple weak peaks). Fluorescence Titration Studies. In a typical experiment, the stock solutions of Py-HPEI-1 were prepared by dissolving 10 mg of Py-HPEI-1 in 10 mL of phosphate-buffered solution (0.01 M, pH 7.4). A 2 mL aliquot of the sensory solution was transferred to the fluorescence cuvette with a cover, and the initial fluorescence was measured. The fluorescence titration experiments were performed by successively adding the CS2 solution to the cuvette and recording the spectra 5 min after each aliquot addition. The fluorescence titration spectra of other Py-HPEI polymers toward increasing concentration of CS2 were measured using a similar method.
■
RESULTS AND DISCUSSION Design and Synthesis of the Sensory Py-HPEI Polymers. The key design in our self-assembled micellar nanoprobes lies in the choice of hyperbranched polyethylenimine (HPEI) to achieve fast fluorescence response. It is wellknown that hyperbranched polymer (HBPs) represent a unique type of chain topology possessing highly branched molecular 20102
DOI: 10.1021/acsami.6b06472 ACS Appl. Mater. Interfaces 2016, 8, 20100−20109
Research Article
ACS Applied Materials & Interfaces
Figure 2. (a) Time-dependent transmittance change of Py-HPEI-1 in pure water in the presence of CS2 (0.1 vol %); inset: photos showing CS2driven phase transition of Py-HPEI-1 in pure water. (b) DLS data for self-assembled Py-HPEI-1 aggregates before and 30 min after the addition of CS2; CS2-triggered time-dependent morphological evolution details visually tracked by transmission electron microscopy (TEM) at 0 min (c), 5 min (d), and 30 min (e). (f) Schematic illustrations of the time-dependent changes of the micellar radius and internal structures in response to CS2.
bands of the S0 to S2 transition of pyrene chromophores.46 Its fluorescence spectrum consists of three emission peaks when excited at λ = 341 nm: The first two weak bands at 381 and 397 nm (labeled M1 and M2 in Figure S3b, respectively) are attributed to the emission of monomeric pyrene from an excited singlet state; the third broad and intense one centered at 488 nm (labeled E in Figure S3b) corresponds to the emission from the relaxation of pyrene excimers. The presence of both weak monomeric and strong excimer emissions suggests that pyrene moieties mainly exist as π-stacked dimers rather than the isolated ones. This result is easily understood because the hydrophobic interior of Py-HPEI-1 micelles is composed of pyrene moieties only, in which the close proximity of pyrene groups leads to the formation of π-stacked dimers at a significantly high propability.45 With the addition of CS2 (0.1 vol %), the intense fluorescence emission of the colloidal solutions turns weak immediately under a UV lamp at 365 nm and becomes nearly nonluminescent after 2 min (Figure S3b). Meanwhile, the nearly transparent polymer solutions become increasingly opaque accompanying with a drastic change in the UV−vis spectra (Figures S3a and 2a). We next explored whether the Py-HPEI-1 aggregates can deform upon exposure to CS2. The size of a colloidal particle has been reported to be in proportion to solution turbidity, so a time-dependent transmittance experiment was then conducted
obtained Py-HPEI polymers. The aqueous solutions of these Py-HPEI polymers were also extracted with chloroform (a good solvent for unreacted pyrene derivatives). UV−vis measurements showed that there are nearly no pyrene derivatives in chloroform phase compared with those in aqueous solution, implying that the postmodification reaction is nearly quantitative. These results clearly verify that pyrene moieties are covalently grafted onto HPEI. CS2-Driven Polymer Self-Assembly and Excimer Fluorescence Quenching. To clearly explain the CS2-driven polymer self-assembly and fluorescence response of Py-HPEI polymers, the amphiphilic Py-HPEI polymer containing 0.91 wt % of pyrene (Py-HPEI-1) was chosen as an example and subject to systematic studies in deionized water solutions. When it is dissolved in water (1 mg/mL), a nearly transparent colloidal solution is formed spontaneously. A noticeable Tyndall effect implies the formation of micellar particles. 1H NMR spectroscopy indicates complete disappearance of pyrene protons (Figure S2), further suggesting the self-assembly of PyHPEI-1 into polymer micelles with hydrophobic pyrene cores stabilized by well-solvated polyethylenimine coronas. Its CS2sensitivity was first investigated by UV−vis and fluorescence spectroscopies. In the absence of CS2, the Py-HPEI-1 micellar solutions exhibit two prominent absorption peaks centered at 329 and 341 nm (Figure S3a), assignable to the vibrational 20103
DOI: 10.1021/acsami.6b06472 ACS Appl. Mater. Interfaces 2016, 8, 20100−20109
Research Article
ACS Applied Materials & Interfaces to detect the particle size change.47 The work profile (Figure 2a) indicates that the transmittance deceases rapidly from 96 to 74% within 5 min and remains nearly constant within at least 1 h. These results suggest that the particle dimensions undergo a rapid change with time and finally reach a plateau, corresponding to a certain stable phase after 5 min. Similar phenomena can be found for other Py-HPEI polymers, and many reported stimuli-responsive polymeric systems.48−51 DLS measurements (Figure 2b) further supported this hypothesis. In the absence of CS2 stimulus, the average hydrodynamic radius (Rh) of Py-HPEI-1 assemblies can reach up to 249 nm. When exposed to CS2 (0.1 vol %) for 30 min, these aggregates exhibit an extensive size reduction to 184 nm. Such a series of significant size changes reflect a time-resolved morphological differentiation of Py-HPEI-1 in water medium. To visualize the differences of Py-HPEI-1 polymeric morphologies in deionized water upon CS2 stimulus, TEM was used. Without any stimulus, Py-HPEI-1 can spontaneously self-assemble into gray and small spherical micelles with the diameter range (DTEM) from about 20−50 nm (Figure 2c). They are stable in aqueous solutions for at least 2 months without forming aggregates if kept at 0.99) between the fluorescence intensity ratio (FE/FM1) and CS2 concentration in the concentration range of 5−400 μM is obtained. All these results suggest that the micellar nanoprobe based on Py-HPEI-4 containing 6.98 wt % of pyrene is the best combination of selectivity, sensitivity, and ratiometric determination of CS2. Furthermore, the different detection ranges of CS2 found for these four chemical-reaction-based supramolecular fluorescent probes make it possible to design systems for qualitative or quantitative determination of CS2 in water medium at different concentration ranges. For instance, it is possible to design a promising sensing array consisted of both Py-HPEI-4 and Py-HPEI-1, where the PyHPEI-1 sensor is used for aqueous samples containing high concentrations of CS2 (400 μM< [CS2] < 1600 μM), whereas the Py-HPEI-4 sensor is used for slightly polluted water samples ([CS2] < 400 μM). These self-assembled micellar nanoprobes, characterized with good selectivity, sensitivity, and photostability, were then
attributed to the specific structure of these sensory polymeric micelles, where a large number of highly reactive functional amino groups are located as the terminating groups of the polymer and thus are exposed outward and easily accessible for CS2. Moreover, the spontaneous movement of the newly produced dithiocarbamate moieties into the hydrophobic interior is also significantly facilitated by both the extremely low polymer chain entanglement and solution viscosity of the hyperbranched sensory polymers. As a result, the equilibration response time of these micellar nanoprobes is greatly shortened. Fluorescence titration experiments were further conducted upon addition of varying amount of CS2 into these micellar nanoprobes, in which the excimer emission intensities were measured 5 min after CS2 added. As shown in Figure 5a, the excimer fluorescence intensity of Py-HPEI-1 micellar solutions gradually decreases with increasing CS2 concentration. Note that a semi-log linear plot is obtained between the excimer fluorescence intensities of Py-HPEI-1 versus the concentration of CS2 in the range of 40−1600 μM (Figure 5c), implying great potential of Py-HPEI-1 for reporting CS2 over a wide concentration range. However, negligible changes of emission intensity at 488 nm are observed at lower CS2 concentration ranges (
