Carbon Nanotubes Act as Contaminant Carriers and Translocate within Plants

Nanotechnology permits broad advances in agriculture. However, as it is still at a relatively early stage of development, the potential risks remain unclear. Herein, for the first time, we reveal the following: 1) the impact of multi-walled carbon nanotubes (MWCNTs) on the accumulation/depuration behaviors of contaminants in crop, mustard (Brassica juncea), and 2) the permeability and transportability of MWCNTs in intact mature mustard plants. Using an in vivo sampling technique, the kinetic accumulation/depuration processes of several contaminants in mustard plans exposed to MWCNTs were traced, and an enhancement of contaminant accumulation in living plants was observed. Meanwhile, we observed that the MWCNTs permeated into the roots of intact living plants (three months old) and were then transported to the upper organs under the force of transpiration steam. This study demonstrated that MWCNTs can act as contaminant carriers and be transported to the edible parts of crops.


Supplementary Text
Plant Growth Conditions and Preparation of the Exposure Mediums. The mustard plants used in this study were cultivated from seeding in a plant incubator (Conviron A1000, Canada) under 14 h light (25 °C ) and 10 h dark (23 °C )，40% humidity (the soil used for plant growth was characterized in Supplementary Table 4, and each pot contained equivalent soil). The mature plants (three-month-old) were used to exposure experiment. MWCNTs suspensions were prepared using an ultrasonic processor (Sonics 04711-35, USA) operated at 500 W. Probe sonication was conducted by using a 20 s on/40 s off pulse sequence with a 13 mm diameter probe tip. Three types of exposure mediums were used for experiment: 200 µg/L contaminants spiked water (HCB and p-p' DDT were 50 µg/L, the spiked water described as CW), CW supplemented with 1 µg/mL MWCNTs suspensions (MWCNTs-1) and CW supplemented with 10 µg/mL MWCNTs suspensions (MWCNTs-10).
Irrigation Method. All of the plants were watered once at day with tap water. The plants from control group were watered with contaminants spiked water (without MWCNTs added) additionally twice a week, and the plants from and MWCNT groups were additionally watered with suspensions containing 1 g/mL of MWCNTs (group MWCNTs-1) or suspensions containing 10 g/mL of MWCNTs (group MWCNTs-10) twice a week. To achieve this task, the contaminants spiked water or MWCNTs suspensions (50 mL for each used concentration) was added inside of soil into each experimental pot In vivo Sampling. In this study, three leaves of each plant were conducted to sampling.
The in vivo sampling process was as follow: The petiole of mustard plants was pierced with a 26 gauge hypodermic needle to a depth of approximately 1.4 cm. Subsequently, the needle was removed, and the custom-made PDMS fiber was deployed in the punched hole ( Supplementary Fig. 10). Two parallel samplings in both sides of each petiole were conducted at each sampling point for mutual reference. After 20 min extraction duration, the fiber was removed, rinsed with deionized water and dried with a Kimwipe tissue, and then assembled to a recycled SPME fiber assembly for being directly introduced to GC-MS for analysis.

Determination of in vivo Sampling Rates.
In the Sample Rate-SPME model (1), it assumes that the sampling rate Rs remains constant when the extracted amount is less than 50% of the equilibrium amount. The concentration of target analyte in the sample matrixes (C0) can be expressed with the following equation: where n is the amount of the extracted analyte in fiber, and t is the sampling time. It assumes that the variance of Rs between individuals is acceptable.
Eq. 1 showed that if the parameters of fiber extraction amount n, and initial sample concentration C0 were measured, the in vivo sampling rate Rs could be obtained in a certain extraction duration. Six mustard plants were irrigated with 50 mL 1000 μg·L −1 spiked water (HCB and p-p' DDT were 250 μg·L −1 ) once a day. After 3 d exposure, the leaves of mustard plants was extracted by in vivo SPME and quantified by GC-MS.
Since the non-exhaustive extraction nature of in vivo SPME, the concentrations of target analyte in the sample matrixes were insignificantly changed after extraction, so the initial sample concentration C0 could be measured using liquid extraction after fiber extraction (as seen below). The in vivo sampling rates for contaminants in leaves of mustard plants were displayed in Supplementary Table 2

Adsorbed Contents of the Contaminants on CNTs.
To study the adsorbed capacity of contaminants on MWCNTs, we used equilibrium SPME (immersion model) with custom-made PDMS fiber to trace the free concentrations of contaminants in the MWCNTs-contaminants suspension system for 7 d. The extraction temperature was 25 °C and the stirring rate was set at 1000 rpm, extraction time was optimized and 80 min was selected ( Supplementary Fig. 11). Raman Spectroscopy Analysis. For MWCNTs detection, the freshly slices of root and leaf were immobilized on the glass slides and then analyzed by Laser-Micro-Raman spectroscopy at room temperature. Raman scattering spectra were recorded using a Renishaw in Via equipped with a charge-couple-detector, and a spectrometer with grating of 600 lines/mm. a Ar + laser (514 nm) was used as the excitation source. The laser beam intensity measured at the sample was 20mW, and Raman shifts were calibrated with a silicon wafer at a peak of 521 cm −1 .