Evaluation of Methods to Determineadsorption of Polycyclic Aromatichydrocarbons to Dispersed

SupplementaryInformation

Evaluation of methods to determineadsorption of polycyclic aromatichydrocarbons to dispersed carbonnanotubes

Berit Glomstad1, Lisbet Sørensen2, †, Jingfu Liu3, Mohai Shen3,§, Florian Zindler1,||, Bjørn M. Jenssen1 and Andy M. Booth2,*

1Department of Biology, Norwegian University of Science and Technology, Trondheim NO-7491, Norway

2SINTEF Ocean, Trondheim NO-7465, Norway

3State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, 100085, China

†Current address: Institute of Marine Research, Bergen NO-5817, Norway.

§Current address: School of Environment, Key Laboratory for Yellow River and Huai River Water Environment and Pollution Control, Ministry of Education, Henan Key Laboratory for Environmental Pollution Control, Henan Normal University,Xinxiang 453007, China.

||Current address: Aquatic Ecology and Toxicology Section, Centre for Organismal Studies (COS), University of Heidelberg, Im Neuenheimer Feld 504, D-69120 Heidelberg,

Germany.

* Corresponding author. E-Mail: ; Tel.: +4793089510.

Methods

Preparation of different media types

Where natural organic matter (NOM) was used in the study, Suwannee River NOM (SR-NOM; International Humic Substance Society) was added to the media to give a nominal concentration of 20 mg/L. The SR-NOM was dissolved by magnetic stirring for 24 h, with any undissolved SR-NOM remaining after this time removed by filtration (0.2 µm Nalgene® filter unit, Thermo Fisher Scientific, Inc.). Media containing SR-NOM is further referred to as media-NOM, or MHRW-NOM and TG201-NOM when discussing the specific types of media. In order to determine the final dissolved SR-NOM concentration in each of the specific media types, dissolved organic carbon (DOC) was determined (Sievers 900 TOC Analyzer, GE 118 Analytical Instruments). Dissolved SR-NOM concentrations were 8.13±0.31 mg/L and 8.77±0.03 mg/L for MHRW-NOM and TG201-NOM, respectively (n = 3) [1].

Dialysis tube protected negligible depletion solid phase microextraction (DT-nd-SPME)

For validation of the DT-nd-SPME method, the extraction units (PDMS-coated fibers placed inside DT) were inserted into 250 mLsolutions comprised of MHRW spiked with polycyclic aromatic hydrocarbons (PAHs). The sample solutions were contained in glass flasks with PTFE lined caps. The samples were placed in darkness on an orbital shaker (125 rpm, 25 °C) for 5 days (determined in an equilibrium time study) in order to extract the Cfree into the fibers. Following Cfree extraction, the extraction units were removed from the flasks, gently dried with a tissue and the PDMS-coated fibers removed. The PDMS-coated fibers were then placed in 2 mL GC-vials containing 950 µL of n-hexane and 50 µL of fluorene-d10 internal standard (10.28 µg/mL), and the target PAH allowed to desorb in darkness for 24 hours. The PAH concentration in the n-hexane was determined by gas chromatography mass spectroscopy (GC-MS).

Analytical chemistry

Gas chromatography mass spectroscopy (GC-MS)

GC-MS was performed using an Agilent 7890B GC equipped with an Agilent 5977A quadrupole MS and EI ion source operated in SIM mode. The GC was fitted with an Agilent J&W HP-5MS fused silica capillary column (30 m length, 0.25 mm id and film thickness 0.25 μm). The carrier gas was helium (grade 6.0) at a constant flow of 1.2 mL/min. 1 μL of sample was injected at 325 ˚C using pulsed splitless injection. The temperature was held at 40 ˚C for 1.1 min, then increased at 6 ˚C/min to 220 ˚C, and then at 4 ˚C/min to 320 ˚C and held for 15 min. Phenanthrene, pyrene and fluorene-d10 were quantified by their molecular ions (m/z 178, 202 and 176 respectively).

High-performance liquid chromatography with photodiode-array UV detection (HPLC-UV)

Direct injection HPLC-UV was performed using an Agilent 1200 HPLC fitted with an Agilent 1260 Diode Array Detector (DAD) (Agilent Technologies Inc.). The column was a SUPERCOSILTM LC-PAH column (Agilent Technologies Inc.; 10cm x 3mm, 3 µm). The mobile phase consisted of deionized water and acetonitrile (30:70, isocratic elution). The flow rate was 1.5 mL/min and the sample injection volume 30 µL. The total run time was 4 min and phenanthrene was quantified at 230 nm (retention time 1.25 min) using external calibration. Samples with a fixed phenanthrene concentration (0.5 µg/L) were repeatedly analysed throughout the experiments to ensure reproducibility of the method.

Tables

Table S1. CNT physicochemical properties supplied by the manufacturer

OD=Outer Diameter SSA=Special Surface Area EC=Electric conductivity

(Adapted from:

* Values provided by manufacturer

Figures

Figure S1: DT-nd-SPME extraction unit. PDMS coated nd-SPME fibers of 1 cm were placed inside 3 cm DTs before the DT ends were folded and closed with metal clam clips.

Figure S2. Extraction of phenanthrene and pyrene onto the PDMS-coated fibers at various time points. Error bars show standard deviation (n=3). No difference (p>0.114) was observed in extracted amount of phenanthrene and pyrene after two and four days, respectively.

Figure S3. Extraction of phenanthrene and pyrene onto the PDMS coated fibers with and without DT protection in MHRW and MHRW-NOM (Cnominal 100 µg/L). No influence from NOM on the extraction was observed. Error bars represent standard deviations (n=3).

Figure S4. Nd-SPME PDMS-coated fiber A) without CNTs, B and C) mixed with MWCNTs (50 mg/L) for seven days, D) washed three times using ultrasonication (100 W, 20 min). The CNTs adhered to the PDMS coating, and after cleaning the fibers the presence of CNTs on the coating was still clearly visible.

Figure S5. Cfree determined by nd-SPME of A) phenanthrene and B) pyrene in the absence and presence of CNTs. A lower Cfree in the presence of CNTs compared to MHRW (no CNTs) indicate PAH adsorption to CNTs.

Figure S6. Calibration curves for DT protected PDMS-coated fibers. A linear correlation between Cnominal and Cfiber was observed in the concentration range tested. Error bars represent standard deviation (n=3).

Figure S7: Extraction by DT-nd-SPME in individual solutions and mixed solutions of phenanthrene and pyrene. Error bars represent standard deviation (individual solutions n=2, mixed solutions n=3).

Figure S8. Absorbance measured at 800 nm in the supernatant of the CNT dispersions after ultracentrifugation. Limit of detection (LOD) and limit of quantification (LOQ) are shown by dotted lines.

Figure S9: Absorbance measured after repeated filtration and filtration of diluted (1:1) dispersions of MWCNT-COOH using selected filters.

Figure S10: Scanning electron microscopy (SEM) images of (A) SWCNT, (B) MWCNT-15, (C) MWCNT-30, (D) MWCNT-OH and (E) MWCNT-COOH.

References

[1]Glomstad B, Altin D, Sørensen L, Liu J, Jenssen BM, Booth AM. 2016. Carbon Nanotube Properties Influence Adsorption of Phenanthrene and Subsequent Bioavailability and Toxicity to Pseudokirchneriella subcapitata. Environmental Science & Technology 50:2660-2668.

1