Article pubs.acs.org/est
Phototransformation-Induced Aggregation of Functionalized SingleWalled Carbon Nanotubes: The Importance of Amorphous Carbon Wen-Che Hou,*,† Chen-Jing He,† Yi-Sheng Wang,† David K. Wang,§ and Richard G. Zepp*,‡ †
Department of Environmental Engineering, National Cheng Kung University, Tainan City, Taiwan 70101 National Exposure Research Laboratory, Exposure Methods & Measurement Division, U.S. Environmental Protection Agency, Athens, Georgia 30605, United States § FIMLab-Films and Inorganic Membrane Laboratory, School of Chemical Engineering, The University of Queensland, Brisbane, Qld 4072, Australia ‡
S Supporting Information *
ABSTRACT: Single-walled carbon nanotubes (SWCNTs) with proper functionalization are desirable for applications that require dispersion in aqueous and biological environments, and functionalized SWCNTs also serve as building blocks for conjugation with specific molecules in these applications. In this study, we examined the phototransformation of carboxylated SWCNTs and associated amorphous carbon impurities in the presence or absence of H2O2 under simulated sunlight conditions. We found that while carboxylated SWCNTs were rather unreactive with respect to direct solar photolysis, they photoreacted in the presence of H2O2, forming CO2 and strongly aggregated SWCNT products that precipitated. Photoreaction caused SWCNTs to lose oxygen-containing functionalities, and interestingly, the resulting photoproducts had spectral characteristics similar to those of parent carboxylated SWCNTs whose amorphous carbon was removed by base washing. These results indicated that photoreaction of the amorphous carbon was likely involved. The removal of amorphous carbon after indirect photoreaction was confirmed with thermogravimetric analysis (TGA). Further studies using carboxylated SWCNTs with and without base washing indicate that amorphous carbon reduced the extent of aggregation caused by photoreaction. The second-order rate constant for carboxylated SWCNTs reacting with •OH was estimated to be in the range of 1.7−3.8 × 109 MC−1 s−1. The modeled phototransformation half-lives fall in the range of 2.8−280 days in typical sunlit freshwaters. Our study indicates that photosensitized reactions involving •OH may be a transformation and removal pathway of functionalized SWCNTs in the aquatic environment, and that the residual amorphous carbon associated with SWCNTs plays a role in SWCNT stabilization.
■
catalysts in addition to carbon nanotubes.14−16 While the postsynthesis purification can increase SWCNT content, varying amounts of impurities remain in the samples, depending on the synthesis and purification processes used.16 For example, amorphous carbon particles can adsorb onto the purified tubes.14−16 Functionalization using oxidative acid with heat to synthesize carboxylated SWCNTs also inadvertently creates additional carboxylated amorphous carbon particles that adsorb onto SWCNTs.15,17 However, the impact of these carbon impurities on the stability of carbon nanotubes in the aquatic environment has not previously been evaluated. Carbon nanotubes can be released into the aquatic environment through the life cycle of carbon nanotubeembedded products such as polymer nanocomposites.18−21 Photoreactions are likely to play a role in their fate and
INTRODUCTION Single-walled carbon nanotubes (SWCNTs) are desirable for nanoelectronic and nanophotonic applications, because of their unique metallic and semiconducting properties.1,2 SWCNTs have therefore been used to fabricate high-value electronic devices such as transistors and thin-film transistor displays and biosensors.3,4 Pristine SWCNTs consist of purely sp 2hybridized carbon atoms whose resulting hydrophobicity hinders applications such as bionanoelectronics and nanocomposites that require dispersions in hydrophilic media (e.g., biological fluids).5−7 While significant covalent modification can disrupt desired electronic properties (e.g., bandgap fluorescence), it has recently been shown that photoluminescence can be enhanced through controlled oxidation of SWCNTs.8−10 SWCNTs with specific functional groups such as carboxyl groups can be further conjugated with biomolecules such as proteins and other functionalities for intended applications.11−13 Most current processes for synthesis of carbon nanotubes invariably produce carbonaceous byproducts and residual metal © XXXX American Chemical Society
Received: September 27, 2015 Revised: February 24, 2016 Accepted: February 29, 2016
A
DOI: 10.1021/acs.est.5b04727 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 1. Photoreactions of carboxylated SWCNTs (25 mg/L) under simulated sunlight exposure at pH 7.0 (5 mM phosphate), showing the UV− visible−NIR absorbance spectra of samples in the (a) absence and (b) presence of H2O2, (c) the color changes of samples containing H2O2, SEM images of samples (e) before and (f) after irradiation in the presence of 100 mM H2O2, and (f) CO2 formation. The data shown in panels a, b, and f are based on the measurements of two separate samples. UV−visible−NIR spectra show the average of two measurements, and the DIC plot shows individual values.
reaction. The removal of amorphous carbon was directly quantified by TGA before and after photoreaction. To understand the role of amorphous carbon, we compared the photoreactivity and aqueous stability of carboxylated SWCNT samples with and without base washing. We monitored photoreactions by measuring the dissolved inorganic carbon (DIC) indicative of CO2 formation, as well as UV−visible−NIR absorbance to quantify the stability of SWCNTs. The samples before and after photoreaction were also characterized by a range of techniques, including Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and electron microscopies. The reactivity of carboxylated SWCNTs toward •OH was quantified by measuring the second-order rate constant that was used to model the half-life under environmental conditions.
transformation, including the formation of reactive oxygen species (ROS).18,22,23 For example, functionalized SWCNTs have been shown to photogenerate 1O2 and •OH under sunlight exposure, while pristine SWCNTs were somewhat inert.24−26 Similar results have been reported for carboxylated multiwall carbon nanotubes (MWCNTs).27,28 Photoreactions can also result in the transformation of carbon nanotubes. For example, Qu et al. demonstrated that carboxylated MWCNTs photodecarboxylate in the presence of added H2O2 under UVA light irradiation and phototransformed MWCNTs exhibited reduced stability caused by adding 10 mM NaCl.27 H2O2 was added as an external source of •OH generated by H2O2 photolysis in this study. Bitter et al. used harsh UVC 254 nm irradiation in the absence of H2O2 (i.e., direct photolysis of MWCNT) and reported that oxidized MWCNTs lost carboxyl groups after irradiation and that photoreacted tubes exhibited reduced aqueous stability and formed aggregates.29 Studies using UVC 254 nm irradiation represent MWCNTs’ fate in engineered systems (e.g., water treatment) because UVC light is not present in sunlight reaching natural surface waters. Our earlier work indicated that unfunctionalized SWCNTs were unreactive under direct sunlight exposure but photooxidized in the presence of added H2O2.30 Pristine SWCNTs are more photoreactive under UVC light conditions, as Alvarez et al. demonstrated that pristine SWCNTs could be photohydroxylated under 254 nm light irradiation.31 However, to the best of our knowledge, there has been no prior study of the phototransformation of functionalized SWCNTs under sunlight conditions. In this study, we examined the photoreactivity of carboxylated SWCNTs with and without added H2O2, under simulated sunlight conditions. One focus was to evaluate the presence of amorphous carbon on carboxylated the aqueous stability of SWCNTs against aggregation driven by photo-
■
MATERIALS AND METHODS
Materials. Carboxylated SWCNT (catalog no. P3-SWNT) and unfunctionalized SWCNT (catalog no. P2-SWNT) samples were purchased from Carbon Solutions, Inc. (Riverside, CA). According to the manufacturer, the samples are made using the electric arc discharge technique and then purified and functionalized (for carboxylated SWCNT) to contain 1.0−3.0 atom % carboxyl groups and >90% (w/w) SWCNTs. The base-washed carboxylated SWCNT sample (catalog no. P33-SWNT) was also obtained from Carbon Solutions, Inc. The washing procedure involved repeated washing of carboxylated SWCNT (i.e., P3-SWNT) in 0.01 M NaOH and collection on filter membranes.17 Other chemicals were of the highest available purity from Sigma-Aldrich (St. Louis, MO) and used without further purification. All aqueous samples were prepared using water purified with an Aqua Solutions 2121BL system (≥18.0 MΩ). B
DOI: 10.1021/acs.est.5b04727 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology Aqueous Dispersions of SWCNTs. In a typical preparation of a stable stock solution, 10 mg of carboxylated SWCNTs (P3 or P33 sample) was added to 100 mL of water and the resulting mixture was sonicated in a water bath (1210R-MT, Branson Ultrasonics) for 3 h. Because of the presence of carboxyl groups, the samples were readily dispersible in water. To prevent some large aggregates from settling in long-term experiments (i.e., 300 h), we centrifuged the stock SWCNT dispersion at a low speed (∼1600g) to remove these large aggregates and irradiated the collected supernatants after dilution to target concentrations. The supernatant concentration was standardized by a calibration curve established using an as-dispersed (no centrifugation) stock suspension where all mass remained in suspension. The procedure and standard curve (Figure S1) are presented in the Supporting Information. Irradiation. We conducted irradiation experiments in an Atlas SunTest CPS+ solar simulator. The solar simulator was equipped with a 1 kW xenon arc lamp. The emission spectrum of the light source is presented in Figure S2a of the Supporting Information. In typical experiments, an aqueous sample containing 25 mg/L carboxylated SWCNTs, 100 mM H2O2, and 5 mM phosphate buffer (pH 7.0) was added to each 24 mL quartz tube and sealed with septum-lined caps. A high H2O2 concentration was used to ensure that H2O2 would not be depleted during the lengthy irradiation and that a steady concentration of •OH, the key reactive species, would be formed from H2O2 photolysis. A series of sample tubes were prepared and completely submerged in a thermostated water bath (25 °C) during irradiation. At predetermined times during irradiation, we removed the tubes from the reactor and sacrificed them for material analysis. Dark control tubes were wrapped with aluminum foil and irradiated concurrently. All experiments were performed at least twice. Samples for photoproduct characterization were prepared in the same manner, except that the pH buffer was not used to avoid potential interference during analysis. Irradiated samples were freeze-dried before photoproduct characterization. To measure DIC indicative of CO2 formation during photoreactions, aqueous samples were sparged with CO2-free air for 1 h to eliminate DIC derived from the atmosphere. The samples were immediately loaded into 24 mL quartz tubes, leaving no headspace, and sealed with open-top caps lined with gastight septa. After irradiation under simulated sunlight, the aqueous samples were taken by using a syringe and directly injected into a Shimadzu TOC-VCPH carbon analyzer operating in IC mode. Sodium bicarbonate was the standard for DIC measurements. This procedure has been used in our earlier work to quantitate DIC in irradiated aqueous fullerenol samples and allows us to distinguish the difference in the DIC level in the irradiated and dark control samples.32 As will be shown later (Figure 1f), the DIC in the dark control samples (
