Article pubs.acs.org/est
Degradation and Microbial Uptake of C60 Fullerols in Contrasting Agricultural Soils Timothy D. Berry,† Timothy R. Filley,*,† Andrea P. Clavijo,‡ Marianne Bischoff Gray,‡ and Ronald Turco‡ †
Department of Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, Indiana 47907, United States Department of Agronomy, Purdue University, West Lafayette, Indiana 47907, United States
‡
S Supporting Information *
ABSTRACT: The environmental fate of functionalized carbon nanomaterials (CNM) remains poorly understood. Using 13Clabeled nanomaterial we present the results of a study investigating the mineralization and microbial uptake of surface-functionalized C60 (fullerols) in agricultural soils with contrasting properties. Soil microcosms rapidly mineralized fullerol C, as determined by 13Ccontent in the respired CO2, with higher fullerol mineralization in an organic, clay-rich soil versus a silty, low C soil (∼56.3% vs ∼30.9% fullerol C mineralized over 65 days). By tracking the enriched 13C from fullerol into microbial phospholipid fatty acids (PLFA) we also report, for the first time, the incorporation of nanomaterial-derived C into soil microbial biomass, primarily by fungi and Gram-negative bacteria. While more fullerol C was incorporated into PLFA in the organic C-rich soil (0.77% vs 0.19% of PLFA C), this soil incorporated fullerol C into biomass less efficiently than the silty, low C soil (0.13% and 0.84% of assimilated fullerol C, respectively). These results demonstrate that, in contrast to pristine C60, surface functionalized C60 are unlikely to accumulate in surface soils and are readily mineralized by a range of soil microorganisms.
■
exposure to sunlight and environmental oxidants.10−12 Both pristine and surface-functionalized CNM have the potential to enter the environment following accidental release from manufacturing facilities, weathering of CNM containing products in landfills, or via wastewater streams following consumer use of CNM products.13 Previous studies, for example, have noted that both C60 fullerenes and fullerols entering liquid-waste streams are concentrated into biosolids.14 CNM accumulated from such waste streams then have the potential to enter terrestrial ecosystems due to the wide-scale application of biosolids to intensively managed agricultural landscapes in the United States15,16 We present the results of a study investigating the environmental fate of 13C-labeled C60 fullerols in two agricultural soils with contrasting edaphic properties. Informed by pure-culture studies demonstrating the significant mineralization of fullerols6 and the enzymatic response of fungi toward functionalized CNM,5 we utilized 13C-enriched fullerols to investigate the hypothesis that, in contrast to pristine CNM, highly functionalized CNM such as fullerols are rapidly mineralized in soils. The use of a stable isotope label in this
INTRODUCTION Many previous studies on CNM have focused on the importance surface chemistry in controlling the environmental fate and reactivity of these potential emerging pollutants, particularly with respect to transport, toxicity, and environmental partitioning.1,2 As with other polyaromatic materials released into the environment, surface chemistry is expected to be a major influence on degradation of CNM. Mechanistic studies of the transformation of carbon nanotubes (CNT) and C60 by oxidative enzymes report increased degradation of the surface-functionalized analogs in comparison to pristine CNT,3,4 whereas pure-culture studies using saprotrophic fungi have found increased enzymatic activity and CNM bleaching when CNM are surface-functionalized.5,6 Recent works, however, have largely focused on the degradation of pristine CNM and report either negligible decay or particle half-lives on the order of 100s of years.7−9 We, however, have recently found that coupled photochemical and biotic degradative processes have the potential to significantly accelerate the degradation of C60 fullerenes, likely as a result of photochemical disruption of the stable condensed aromatic structure and the inclusion of oxidizable and hydrolyzable sp3 and sp2 oxygenated carbon.10 Fullerols are representative of both manufactured CNM designed for enhanced solubility (e.g., for drug delivery applications) and CNM that have been chemically altered through environmental processes such as the prolonged © 2016 American Chemical Society
Received: Revised: Accepted: Published: 1387
September 12, 2016 December 19, 2016 December 26, 2016 December 26, 2016 DOI: 10.1021/acs.est.6b04637 Environ. Sci. Technol. 2017, 51, 1387−1394
Article
Environmental Science & Technology
were meant to reflect a CNM concentration resulting from accidental release of CNM from a point source. Microcosms treated to suppress microbial activity were prepared to help distinguish between abiotic and biotic mineralization of fullerol C. These treatments were prepared by autoclaving microcosms at 121 °C and 100 kPA for 30 min. Autoclaving was repeated 4 times with a 1-day recovery period between sterilization cycles. Fullerols were then added to the autoclaved microcosms as above, before being chloroform fumigated to further suppress the microbial community. Soilfree controls were also prepared to help assess the role of edaphic properties in controlling fullerol oxidation. Each soilfree control contained 4.5 g of ashed quartz sand and received 1.57 mg of fullerols (equivalent to the loading of the Clermont microcosms). All microcosms were incubated in darkness at 26 °C for 65 days with caps sealed to prevent moisture loss. Isotopic composition of headspace CO2 was measured using a Sercon (Sercon Ltd., Crewe, UK) TGA2 trace gas analyzer interfaced to a Sercon 20/20 isotopic ratio mass spectrometer (IRMS) and is reported in standard delta (δ) notation (‰) relative to Vienna Pee Dee Belemnite (vPDB), an international standard for 13C content. Since microcosm headspace was flushed during analysis, cumulative carbon respiration was calculated by integrating the measured respiration rates between time points. After incubation a portion of soil from each microcosm was analyzed to determine the isotopic composition of the remaining fullerol treated soil. Soil subsamples were freeze-dried and ground before being weighed (∼18 mg and ∼30 mg for Drummer and Clermont soils respectively) into tin capsules for analysis. The tin-wrapped soils were flash-combusted in a Sercon (Crewe, U.K.) elemental analyzer interfaced to the Sercon 20/22 IRMS. The relative contributions of soil organic carbon and fullerol C to head space CO2 and to the postincubation microcosm soil mixture was calculated using a simple two-member mixing model as described in the Supporting Information. Phospholipid Extraction and Analysis. PLFA were extracted from 3 g subsamples of the microcosm soil mixture as outlined by Bligh and Dyer and White and converted into fatty acid methyl esters (FAMES) to allow for analysis by GC.20,21 Briefly, lipids were extracted from soils by mixing with chloroform, methanol, and phosphate buffer prepared in a 1:2:0.8 ratio before being allowed to stand 24 h and centrifuged to separate soil particles from solvent layers. The chloroform layer was then extracted and concentrated by rotary evaporation and blown dry under N2. Resuspended total lipid extracts were transferred to silicic acid extraction columns.22,23 Neutral lipid fractions and glycolipids fractions were discarded prior to elution of PLFA by methanol into clean vials. PLFA were converted to the corresponding FAMES by mild alkaline methanolysis in methanolic KOH. 10% of the resulting FAMES were separated and quantified using a 60 m DB-5 capillary column (0.25 mm ID, 0.25 μm film thickness; J&W Scientific) inside of an Agilent 5975 gas chromatograph/mass spectrometer (Agilent, Santa Clara, USA). FAMES were identified by retention time and fragmentation pattern and quantified by peak area against a standard (Standard #1114, Matreya LLC, Pleasant Gap, PA). The 13C content of the remaining FAMES was measured using gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS). FAMES were separated on a Hewlett-Packard 6890A GC (Agilent, Santa Clara, CA) containing a 60 m RTX-5MX (0.25 mm ID, 0.25 μm film
study allowed for the determination of microbial mineralization of fullerols while compound specific isotope analysis (CSIA) of phospholipid fatty acids (PLFA) was used to assess the relative uptake of fullerol carbon by soil microbes and thus ascertain which microbial guilds might drive a soil’s response to fullerol exposure. Together, these values allow for an estimation of degradation rates and a determination of the growth efficiency of microbes using fullerols as a substrate.
■
EXPERIMENTAL SECTION Fullerol Synthesis. The 13C fullerols were synthesized as described in Schreiner et al.;6 briefly, 13C60 fullerenes (MER Corp.) were oxidized in CH2Cl2 by 18-crown-6 stabilized KO2. The precipitant produced in this reaction was collected, washed in CH2Cl2, suspended in H2O and purified by filtration through a 0.45 μM PTFE filter and repeated methanol/water precipitations. Purity of the resulting product was verified by HPLC in which no evidence of amorphous carbon was found.6 As originally reported in Schreiner et al., the resulting fullerols were characterized by solid state 13CNMR and were found to be functionalized with between 19 and 27 hydroxyl groups (13C60(OH)19−27).6 A 5 atom % 13C fullerol suspension was prepared by dissolving 64.5 mg of non-13C-enriched fullerols C 60 (OH) 19−27 (1.09 atom % 13 C) and 15.3 mg of 13 C60(OH)19−27 (22 atom % 13C, prepared as above) together in ultrapure water to produce a suspension with a concentration of 4 mg/mL and an isotopic composition of 5 atom % 13C. Microcosm Incubation Study. The two surface soils (0− 10 cm) used in this study are characteristic of soils commonly used in agriculture in the upper-Midwest U.S. and have been used in previous studies on the impacts of CNM by our research group.10,17,18 The Clermont series silt loam soil, obtained from the Southeast Purdue Agricultural Center (Butlerville, IN), is a fine-silty, mixed, superactive, mesic Typic Glossaqualf with 0.77 wt % C, 16% clay, and a pH of 4.8. The Drummer series loam soil, is a fine-silty, mixed, superactive, mesic Typic Endoaquoll obtained from the Purdue Agronomy Center for Research and Education (West Lafayette, IN) characterized by a higher pH, 7.8, carbon, 2.41 wt %, and clay, 26%, content. Soils were sieved to 4 mm at field moisture following collection, air-dried, and then stored in the dark at room temperature before use. A detailed characterization of each soil is available in Table S.1. Microcosms to evaluate microbial decomposition of fullerols were constructed as outlined in our previous studies.10,19 Briefly, 3 g of air-dried soil, sieved to remove large plant fragments and stones (0.5 cm) and 1.5 g of ashed quartz sand (53−250 μm) were gently mixed and added to sterile 12 mL borosilicate glass vials with screw-caps fitted with piercable rubber septa (LabCo, UK). Sand was added to both soils in order to prevent caking of soil and promote aeration. The aqueous fullerol suspension detailed above was then added to soils such that fullerol carbon accounted for 1% of total microcosm organic matter (∼3% soil carbon). Thus, a fullerol loading of 1.57 mg (0.69 mg carbon) was used for Clermont soil microcosms and 5.33 mg (2.34 mg carbon) for Drummer soil microcosms, corresponding to 2.9 and 3.1% of total microcosm carbon, respectively. Additional sterile ultrapure water was added to the microcosms in order to bring the soil in each to field moisture (see Table S.1.) Control microcosms were brought to field moisture using only sterile ultrapure water. Application rates of the fullerols used in this study are comparable to those used previously in similar studies10,18 and 1388
DOI: 10.1021/acs.est.6b04637 Environ. Sci. Technol. 2017, 51, 1387−1394
Article
Environmental Science & Technology
composition of CO2 at each time-point were performed using one-way ANOVA with a Tukey HSD posthoc test, as the two soils used had fundamentally different properties, significance between treatments was assessed for each soil separately. The significance of differences in microbial abundance, oxidative enzyme activity (see SI), PLFA isotopic composition, and fullerol/NSC mineralization in fullerol containing and control microcosms was assessed using unpaired student’s t tests. Statistical analysis was performed with JMP 8 (v8.0.2.0 SAS Institute, Cary, NC, USA) with significance set for α = 0.05 in all tests.
thickness, Restek) column before undergoing combustion in a ceramic microcombustion reactor containing two 25 cm lengths each of pure nickel, copper, and platinum wires, maintained at 1000 °C and interfaced to a Sercon 20/22 isotope ratio mass spectrometer (IRMS, Sercon, Crewe, UK). A continuous stream of He with 3% O2 was used as a makeup gas. Peak separation was achieved using a ramped temperature program; 50 °C for 5 min, 50−120 at 10 °C/min, 120−250 at 4 °C/min, 250 °C for 12 min. An internal standard of nC15, isotopically verified at a δ13C value of −30.2 ‰, was added to samples at a concentration of 10.0 ng/μL to correct isotopic values when necessary. Standard calibration curves were also included in each batch of analyses to ensure the consistency of isotopic data. PLFA-FAMEs are referred to using short-hand fatty acid nomenclature following the format “X:YnZ”, with X indicates the length of the carbon chain, Y indicates the number of double bonds and Z indicates the position of the double bonds relative to the methyl end of the fatty acid. Double bound geometry is denoted by the suffices c and t for cis and trans bonds while the prefixes i and a indicate iso- and anteisobranching of saturated FAMEs. Cyclopropyl groups are denoted with the prefix cy-. PLFA-FAMEs were combined into taxonomic guilds as follows; general (14:0, 15:0, 16:0, 17:0, 18:0), Gram-positive bacteria (i15:0, a15:0, i16:0, i17:0, a17:0), Gram-negative bacteria (cy17:0, cy19:0, monounsaturated 16C), and fungi (monounsaturated 18C), a classification scheme consistent with a number of previous studies.22,23 FAME isotopic composition was corrected to remove the contribution of the methyl group added in methyl esterification of the extracted PLFA. A two-member mixing model using PLFA extracted for fullerol-free controls and 13C60(OH)19−27 as end-members was used to calculate the relative contribution of fullerol derived carbon to PLFA carbon, this fraction was then used to calculate total 13C uptake by the soil microbial community. These calculations are fully described in the Supporting Information. Soil DNA Extraction and Bacterial Abundance. DNA was extracted from ∼250 mg of soil using the PowerSoil DNA isolation kit (MOBIO, Carlsbad, CA) as per manufacturer instructions. Agarose gel electrophoresis and spectrometry using a NanoDrop 1000 spectrophotometer (Thermo Scientific, Wilmington, DE) was used to assess DNA quality before fluorometric quantification using the Qubit 2.0 fluorometer and dsDNA BR assay kit (Life Technologies, Grand Island, NY) as per manufacturer instructions. Bacterial abundance was determined by qPCR of the 16S rRNA using primers 534R and 388F-GC to amplify the V3 region of the 16S rRNA.24 The Sso Advanced Universal Inhibitor-tolerant SYBR green supermix (Biorad, Hercules, CA) at 1X was used in the amplification reaction with 8 pmol of each primer and ∼10 ng DNA, in a final volume of 20 μL. The temperature profile consisted of an initial denaturation at 98° C for 3 min, followed by 40 cycles of denaturation at 98° C for 30 s and annealing and extension at 60° C for 1 min. Melting curve consisted of an initial denaturation at 95° C for 15 s followed by renaturation at 60° C for 15 s, and 35 subsequent 1° C temperature steps. Standard DNA for absolute quantification was constructed using pCR2.1 TOPO TA vector (Life Technologies, Grand Island, NY) containing a full-length copy of the 16S rRNA gene from Bacillus subtilis. Statistical Analyses. Comparisons to assess the significance of differences in microcosm respiration rates and isotopic
■
RESULTS AND DISCUSSION Fullerol Carbon Is Readily Mineralized in Soils. Incubation of 13C-enriched fullerols in both soils resulted in significant enrichment of 13C−CO2 in the headspace, indicating that the fullerols were mineralized to CO2 (Figure S.1). The isotopic composition of headspace CO2 was used in conjunction with respiration rate (Figure S.2) to calculate total mineralization of fullerols by each soil (Figure 1). Over
Figure 1. Cumulative mineralization of added fullerols during the incubation. Plotted values are means ± standard error, n = 5. Suppressed soils mineralized significantly less fullerol than the corresponding nonsuppressed soils over the course of the incubation (p < 0.001 for both soils).
the course of the 65-day incubation microcosms containing the clay and organic C-rich Drummer series soil mineralized 1.38 ± 0.32 mg of C60 fullerol C (equivalent to 56.26 ± 5.37% of the added C) while microcosms containing the siltier Clermont soil mineralized 0.18 ± 0.05 of fullerol C (30.87 ± 5.95% of added C). Drummer and Clermont microcosms in which microbial activity had been suppressed mineralized significantly less of the added fullerol at 0.09 ± 0.06 mg (4.1%, p < 0.001) and 0.05 ± 0.02 mg (6.7%, p < 0.001), respectively. The majority of mineralization in these suppressed controls occurred later in the incubation as the soil microbial communities had begun to recover and respire (see Figure S.2 and discussion in SI). Soilfree, sterile controls were also included in the incubation study, and mineralized 0.04 ± 0.01 mg (6.1%) of the added fullerol C. The measurable mineralization in these soil-free controls indicates that a small portion of the 13C-enriched CO2 evolved from the microcosms was the result of nonmicrobial degradation. Photo-oxidation and autoxidation of fullerols are two possible avenues for the mineralization observed in these controls; the role of light exposure in mineralizing CNM has been previously observed11 and dissolved, condensed aromatic substances such as naturally occurring humic acids have been shown to undergo auto-oxidation resulting in mineralization.25 Regardless of the source, mineralization in these controls is 1389
DOI: 10.1021/acs.est.6b04637 Environ. Sci. Technol. 2017, 51, 1387−1394
Article
Environmental Science & Technology
recently highlighted by Navarro et al. which reported mineralization of ∼3% of added fullerols in a biosolid and soil mixture after 55 days.30 Although Navarro et al. reported more rapid mineralization than Schreiner et al., mineralization still occurred an order of magnitude more slowly than in the current study. Crucially, Navarro et al. added fullerols along with biosolids, which have been previously demonstrated to retain CNM,30,31 rather than as an aqueous suspension. Thus, the lower levels of mineralization observed in that study likely reflect decreased microbial accessibility of CNM following sorption to biosolids, rather than the best case scenario for soil microbial mineralization presented herein. Fullerol Addition Has Minor Effects on Net Soil Respiration and Bacterial Abundance. The addition of fullerols had minor effects on the net soil microbial respiration in both soils used in this study, resulting in measurable but not statistically significant increases in the amount of native soil carbon (NSC) respired with respect to the water-only controls (Figure 2). Drummer microcosms containing fullerol respired
minor in comparison to that observed in either Drummer or Clermont microcosms (p < 0.001). A postincubation isotopic analysis of microcosm soil was conducted to confirm that the loss of fullerol 13C could be reconciled with the mineralization calculated by headspace 13 CO2 analysis. The Drummer and Clermont soils (initially at a δ13C = −18.01 ± 0.33‰ and −20.53 ± 0.11 ‰ vPDB, respectively) were significantly enriched in 13C following fullerol addition but prior to incubation (δ13C ≈ 98‰ and 87‰ vPDB, respectively). Following incubation, however, both the Drummer and Clermont soils were depleted in 13C relative to initial fullerol treated soils (δ13C = 41.40 ± 6.19‰ and 39.15 ± 4.38‰, respectively), a finding that is consistent with removal of 13C-enriched fullerols from the soils as a result of mineralization. This shift in isotopic composition corresponds to a mineralization of 1.15 ± 0.13 mg of fullerol C (49.1% of added fullerol) in the Drummer soil and a mineralization of 0.31 ± 0.03 mg (44.9% of added fullerol) in the Clermont soil. These estimates of mineralization are comparable with those calculated from the headspace calculation (56.3% and 30.9%, respectively) but suggest that headspace calculations may slightly overestimate mineralization in Drummer soils while underestimating mineralization in the Clermont series soil. The decay rates observed for fullerols in this study are substantially higher than those previously reported for CNM with little or no surface functionalization. Hartmann et al. found no measurable mineralization of fullerenes following prolonged exposure to indirect lighting, whereas a recent study by Parks et al. reported a degradation rate for single-walled CNT of 30% to total PLFA derived just 0.27% of PLFA carbon from fullerols in Drummer soils and 0.09% in Clermont soils. Gram-negative bacteria, while contributing 2.25% and 1.82% to total PLFA of the Drummer and Clermont soil, respectively, derived 0.69% and 1.17% of their PLFA carbon from the fullerols. The affinity of Gramnegative bacteria for easily metabolizable soil and plant carbon or the labile fractions of other highly condensed aromatic carbon like pyrogenic organic matter (PYOM)37,38 has been documented previously using compound specific stable isotope 1391
DOI: 10.1021/acs.est.6b04637 Environ. Sci. Technol. 2017, 51, 1387−1394
Article
Environmental Science & Technology
Table 1. Total Microbial Guild PLFA Carbon, Fractional Contribution of Fullerol Derived Carbon to PLFA, and Estimated Total Fullerol Carbon Incorporated into Soil Biomass Drummer Series Soil
characteristic PLFA gram positive bacteria gram negative bacteria fungi general whole soil
i14:0a, i15:0, a 15:0, i 16:0, i17:0,a17:0 cy17, cy19, 16:1n7b, 16:1n9b 18:1n9,18:1n11 16:0,18:0c all of above
Clermont Series Soil
total PLFA C (ng/microcosm)
C60 C in PLFA (%)
estimated C60 C in microbial biomass (ug)
total PLFA C (ng/microcosm)
C60 C in PLFA (%)
C60 C in total microbial biomass (ug)
1700.0 ± 151.0
0.27
2.31 × 10−1
1914.0 ± 91.4
0.09
3.02 × 10−1
777.5 ± 116.7
0.69
2.52 × 10−1
105.9 ± 5.9
1.17
1.75 × 10−1
971.2 ± 138.8 1364.7 ± 221.2 4813.5 ± 323.5
1.67 0.79 0.77
6.89 × 10−1 5.14 × 10−1 1.75
1020.1 ± 73.8 2020.1 ± 89.3 5060.0 ± 147.7
0.24 0.21 0.19
3.45 × 10−1 6.69 × 10−1 1.53
a
i14:0 could not be separated from Clermont microcosms. b16:1n7 and 16:1n9 c/t could not be separated from Clermont microcosms. c18:0 Could not be separated from Drummer microcosms.
biomass, similar techniques have been used to approximate MGE; for example, Lashermes et al. used the abundance of ergosterol, a fungal membrane component, in conjunction with respiration data in order to determine the MGE of saprotrophic fungi growing on maize litter.49 Using this approach, microbial biomass in the Drummer series soil was estimated to retain 1.75 μg of fullerol carbon, whereas Clermont series soils incorporated 1.53 μg. As the Drummer soil mineralized 1.38 mg of fullerol, the assimilation of only 1.75 μg of fullerol carbon indicates a strong preference for catabolic metabolism of the substrate, resulting in a MGE of 1.27 × 10−03. Having mineralized only 0.18 mg of fullerol but incorporating 1.53 μg into biomass, the microbial community in the Clermont soil utilized fullerol carbon for growth with a higher efficiency, 8.43 × 10−03. The MGE reported herein is small in comparison to more conventional microbial substrates and other aromatic xenobiotic compounds. Although MGE varies with growth conditions, it is generally highest for compounds that can be directly incorporated into biomass, such as glucose (MGE ≈ 0.8)50 and decreases for more complex substrates that require specialized degradative processes and extensive modification in order to produce necessary biochemical precursors.45 Thus, while simple aromatic compounds such as toluene and phenol (MGE ≈ 0.28 and 0.38)50,51 are less efficiently utilized than simple carbohydrates, but are still much more efficiently incorporated into biomass than carbon from more condensed aromatics, including PAHs such as phenanthrene (MGE ≈ 0.01− 0.02).52,53 In contrast to PAHs, the fullerols in the current study are highly oxidized. Nevertheless, the efficiency of microbial incorporation of fullerol carbon into biomass herein is lower than that seen in these more condensed compounds, and may instead be the result of the high molecular weight of fullerols and their breakdown products, a characteristic previously associated with decreased degradation of aromatics.54 Low MGE has been reported previously with large aromatic rich particulates; for example, microbial utilization of carbon from PYOM can be as low as 1.5 × 10−03 (equivalent to the MGE of fullerol in the Drummer soil).55 The low MGE of this substrate suggests that the increase in bacterial abundance in the Clermont microcosms following fullerol addition was not strictly the result of fullerols representing a viable carbon source for soil microorganisms, as purely substrate induced growth should result in a much greater MGE.56 Implications for the Environmental Fate of Fullerols. The extensive mineralization observed in this study stands in contrast to the negligible mineralization reported in previous
analysis (CSIA) of soil phospholipids. Gram-positive bacteria, conversely, are frequently associated with more slowly degraded pools of carbon in soils.37,39 These observations may help explain the difference in microbial uptake of CNM carbon by these groups in the present study; the highly soluble fullerol are degraded more readily by Gram-negative bacteria, which tend to quickly utilize dissolved carbon sources rather than expend energy on extracellular enzymes to break down complex organic matter. Direct microbial uptake of fullerols may partially explain the extensive mineralization observed herein, despite the lack of a corresponding increase in extracellular oxidative enzyme activity (see SI text). Indeed, previous studies have observed the uptake of C60 into Gram-negative bacteria with minimal losses to cell viability.40 The only previous study to assess the uptake of fullerolderived carbon into microbial biomass reported a CNM contribution to PLFA C of
