Article pubs.acs.org/est
Slow Biotransformation of Carbon Nanotubes by Horseradish Peroxidase D. Xanat Flores-Cervantes,†,‡ Hanna M. Maes,§ Andreas Schaff̈ er,§ Juliane Hollender,‡ and Hans-Peter E. Kohler*,† Department of Environmental Microbiology, Eawag, Ü berlandstrasse 133, 8600 Dübendorf, Switzerland Department of Environmental Chemistry, Eawag, Ü berlandstrasse 133, 8600 Dübendorf, Switzerland § Institute for Environmental Research, RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany † ‡
S Supporting Information *
ABSTRACT: Due to steady increase in use and mass production carbon nanotubes (CNTs) will inevitably end up in the environment. Because of their chemical nature CNTs are expected to be recalcitrant and biotransform only at very slow rates. Degradation of CNTs within days has recently been reported, but excluding one study, conclusions relied solely on qualitative results. We incubated 13 different types of CNTs and subjected them to enzymatic oxidation with horseradish peroxidase and concluded that the analytical methods commonly employed for studying degradation of CNTs did not have the sensitivity to unequivocally demonstrate degradation of these materials. To obtain unambiguous results with regard to the biotransformability of CNTs in the horseradish peroxidase system we incubated: (a) 14C-labeled multiwalled CNTs, homologous to Baytubes CNTs; and (b) 13C-depleted single-walled CNTs, used in previous studies. Our results show that 14C−CO2 evolved linearly at a rate of about 0.02‰ per day, and at the end of the 30-day incubations the CO2 evolved amounted to about 0.5‰ of both initial substrates, the 14C-labeled multiwalled and 13C-depleted single-walled CNTs. These results clearly show that CNT material is oxidized in the horseradish peroxidase system but with half-lives of about 80 years and not a few days as has been reported before. Adequately addressing biotransformation rates of CNTs is key toward a better understanding of the fate of these materials in the environment.
■
sp2 bonds between individual carbon atoms within aromatic rings.14−16 Based on these characteristics, we expect CNTs to be persistent in the environment and biotransform at very slow rates. However, it is known that defects, such as vacancies, pentagon-heptagon pairs (Stone−Wales defects), add-atom defects, and combinations of the above, are often present in graphitic structures.17−21 These defects affect the physical17 and chemical22 properties of CNTs and could act as reaction sites20 for chemical or biological mediated transformation reactions. There is an ongoing debate regarding the persistence of these materials in the environment. In recent studies it has been postulated that biodegradation of single- (SWNT)23−26 and multiwalled CNTs (MWNTs)26−28 by phagolysosomal simulant fluid, exogenous or endogenous enzymes, and bacteria proceeds within 10 days of incubation. However, the extent of degradation between studies varies substantially (Supporting Information (SI) Table S1). In addition, results from other
INTRODUCTION Due to their unique properties, potential uses, and wide range of applications (e.g., optical, electronic, biomedical), carbon nanotubes (CNTs) have seen a steady increase in production during the last two decades with an estimated production of more than 300 t/yr.1,2 As a response to the increase in public awareness and concern regarding CNTs in the environment, recent research has focused largely on the uptake3 and toxicity,4−6 and to a lesser extent on modeling the fate of CNTs in the environment.7−9 However, there is scarce information regarding possible biotransformation reactions that CNTs might undergo once released into humans or natural environments.10,11 As a result, assessing the possible human and ecological health effects of CNTs remains a challenge. Furthermore, recent environmental and biomedical applications, such as targeted delivery of drugs12 or remediation agents,13 where CNTs are purposefully introduced, raise questions regarding the risk and persistence of these materials in humans and environmental settings. Apart from their excellent thermal and electrical conductivity properties, CNTs are well-known for their very strong thermal, chemical, and mechanical durability that is associated with the © 2014 American Chemical Society
Received: Revised: Accepted: Published: 4826
November 29, 2013 March 25, 2014 March 28, 2014 March 28, 2014 dx.doi.org/10.1021/es4053279 | Environ. Sci. Technol. 2014, 48, 4826−4834
Environmental Science & Technology
Article
Table 1. Characteristics of CNTs Useda CNT name b
CT1 CT6b CT8b CT10b CT11b CT14b BTd CNT - PEGc CT6-COOHb,g HiPCOe P2f HiPCO−COOHe,g P2-COOHf,g
length (μm)
OD (nm)
5−30 0.5−2.0 0.5−2.0 10−30 10−30 10−20 1−10 0.5−0.6 0.5−2.0 0.5−2.0 0.5−2.0 0.5−2.0 0.5−2.0
1.1 95 94.11j/>95 72.00j/>80 87.17j 77.19j/>85 87.08j/>90
SW/long/thin MW/short/thin MW/short/thin MW/long/thin MW/long/thick MW/long/thick MW/short/thin SW/short/thin MW/short/thin SW/short/thin SW/short/thin SW/short/thin SW/short/thin
a
SW: single walled; MW: multiwalled. bCheap Tubes (Cheap Tubes Inc., VT). cSigma-Aldrich (Sigma-Aldrich Chemie GmbH, Buchs, Switzerland). Baytubes (Bayer AG, Leverkusen, Germany). eNanointegris (Nanointegris, IL). fCarbon Solutions (Carbon Solutions Inc., CA). gfurther carboxylated. hSobek and Bucheli (2009). iProvided by manufacturer as %C; impurities include metal catalysts (e.g., Fe, Co, Mn, Mg, Al, Ni, and Y). j measured using elemental analysis. OD, ID are outer and inner diameter, respectively. d
of 250 μL 800 μM (or 125 μL 1.6 mM) H2O2 were injected into the HRP-CNT suspension to maintain a constant H2O2 concentration in the tested system. All the material and suspensions used were autoclaved, with exception of H2O2 and HRP, which were filter-sterilized. Throughout the incubation time, the vials were kept in a incubator shaker (Adolf Kuhner AG, Switzerland) at 25 °C ± 1 °C and 180 rpm. CNT Carboxylation. Some of the tested CNTs were additionally carboxylated based on previous protocols,32,33 by adding approximately 10 mg of CNTs to 20 mL of H2SO4:HNO3 at a ratio of 3:1, and placing the suspension in a sonication bath (Bandelin electronic, Germany, 35 kHz) for 8 h at 60 °C, vigorously shaking the suspension every 15 min. The time and temperature chosen provided the highest degree of carboxylation and defects in the CNT walls, while keeping the integrity of the CNTs (based on inspection of characteristic Raman spectra bands; RBM (150−300 cm−1) and D band (∼1300 cm−1); SI Figure S11). The suspension was then diluted (1:20) and filter-washed with excess water until the pH of the filtrate was close to neutral. The resulting carboxylated CNTs were dried overnight at 60 °C, and resuspended in PBS as described above. FTIR spectra of the carboxylated 13Cdepleted CNTs are presented in SI A.3. CNT Characterization. CNT suspensions were characterized at the beginning and at the end of each incubation experiment. The characterization methods and techniques used include the following: Raman Spectroscopy. About 3 mL of the CNT suspensions were filtered onto precombusted 0.45 μm fiberglass filters. To increase the intensity of the signal, the suspensions were filtered through pipet tips. Raman spectra of the filtered suspensions were obtained with a Bunker Senterra Raman microscope (Bunker Optic GmbH, Ettlingen, Germany) of 5 cm−1 spectral resolution using a 532 nm laser, a spectral range from 110 to 3690 cm−1, a 30 s integration time, and a laser power of 10 or 20 mW for SWNTs and MWNTs, respectively. Average Agglomerate Size Determination by Dynamic Light Scattering (DLS). Changes in the average agglomerate size of the CNT suspensions, in terms of the hydrodynamic diameter in solution, were monitored using a quasi-elastic light scattering spectrometer (dynamic light scattering, DLS, Malvern Zetasizer, Herrenberg, Germany). For each replicate
studies indicate much slower biotransformations and, therefore, higher biopersistence (more than 60 days) of SWNTs and MWCNTs.29,30,25,26 Although in some cases the same or a very similar set up has been used, the lack of detailed descriptions, standard procedures or standard CNT testing materials creates a challenge for reproducing or comparing results between different studies (SI Table S1). Furthermore, to date conclusions regarding the biodegradability of CNTs are mostly based on qualitative results and not on mass balances and reliable biotransformation rates. To overcome the disparity in the literature regarding the persistence of CNTs and to test the concordant hypothesis that the persistence of these materials depend on their surface functionalization,25−27 in this study we test the oxidation with horseradish peroxidase (HRP) of various CNTs with different sizes, shapes, and degrees of functionalization. HRP was chosen because of previous reports on CNT degradation with this enzyme and because of its ability to catalyze the oxidation of aromatic compounds, including a wide variety of phenols.31
■
MATERIALS AND METHODS Incubation Experiments. Thirteen different types of CNTs (Table 1) were subjected to oxidative degradation through enzymatic catalysis with horseradish peroxidase (HRP, Type VI, Sigma-Aldrich) with slight modifications to the procedure reported by Allen and collaborators.24 In brief, CNT suspensions (between 3 and 12 mg/L) were created by adding ∼1 mg CNT to 40 mL of 0.01 M phosphate buffer saline (PBS, adjusted to pH 6.0), and sonicating the suspension for 2.0 h in 15 min intervals in an ultrasonication bath (Bandelin electronic, Germany, 35 kHz). Every 15 min interval the suspension was vigorously shaken. The suspensions were then either centrifuged for 20 min at 3500 rpm or, when concentrations were below 3 mg/L after centrifugation, the samples were allowed to settle for 24 h after the sonication treatment. The later was the case for CT6, CT8, CT11, CT14, and BT. Four milliliters of the supernatant were added to 4 mL of solubilized HRP (0.385 mg/mL in PBS) and incubated for 24 h, followed by the addition of H2O2 to a final concentration of 400 μM to activate the HRP. The vials were then sealed with screw-in caps with septa and wrapped with parafilm to create an airtight seal. Throughout the incubation time (10 or 30 days) daily additions 4827
dx.doi.org/10.1021/es4053279 | Environ. Sci. Technol. 2014, 48, 4826−4834
Environmental Science & Technology
Article
initial value (measured by the UV-vis intensity signal of resorufin ∼570 nm; see SI Figure S5) at the end of the incubation period. 14 C-radiollabeled CNTs. 14C−CNTs with a specific radioactivity of 1.3 MBq/mg were produced via catalytic chemical vapor deposition of 14C-benzene. The product was purified with 12.5% HCl to remove excess metal catalysts present (material remaining 95% C purity). The produced material was not further functionalized (FTIR spectra in SI A.3) but contains a certain degree of surface defects as observed by their D/G ratio (SI Table S3). Eight milliliters of stable phosphate buffer saline (0.01 M at pH 6.0) suspensions of 14 C−CNT (approximately 7 mg/L) and 8 mL of horseradish peroxidase (Type VI, Sigma-Aldrich, at 0.385 mg/mL) were added to a 100 mL Schott flask and allowed to incubate for 24 h at 25 °C. After the initial 24 h incubation period 16 mL H2O2 (800 μM) were added to the suspension to activate the enzyme, and 250 μL H2O2 (1.6 mM) were added on a daily basis to replenish H2O2 losses. For measuring evolved 14C−CO2, two methods were tested: (a) Glass vials containing 2 mL 1.0 M NaOH were placed inside the incubation vials (SI Figure S6). To measure 14 C−CO2, on a chosen time interval the flasks were flushed for 30 min with N2 gas, which was then bubbled into parallel flasks containing 10 mL 1.0 M NaOH (see SI Figure S7 and S8). The flasks were then opened, the 2 mL 1.0 M NaOH present inside the incubation flasks were transferred to a scintillation vial, and 15 mL of Hionic Fluor (National Diagnostics, Atlanta, GA) scintillation cocktail were added before LSC (Tricarb 2200 CA, Canberra Packard, Zurich, Switzerland) quantification. LSC measurements of 2 mL of each of the flasks containing 1.0 M NaOH used to flush the sample headspace were also performed; however, 14C concentrations were below the limit of detection of our system (
