Introduction

Fullerenols was the most promising water-soluble derivatives of fullerenes, which was known as nanoassembly into clusters in the form of fullerenol nanoparticles (FNPs) (Djordjevic et al. 2015; Wang et al. 2015b). FNPs as model carbon-based nanomaterials (CNMs) exhibited excellent capabilities in the quenching of various physiologically relevant free radicals compared to conventional antioxidants. The antioxidant properties and potential application in nanomedicine of FNPs have extracted extensive attentions in the past decades, which were described in the comprehensive review papers (Chen et al. 2012; Djordjevic et al. 2015; Grebowski et al. 2013; Markovic and Trajkovic 2008; Nielsen et al. 2008; Wang et al. 2015b). The successful and tremendous applications of FNPs in nanomedicine have also raised considerable concerns in agriculture. For example, the first report on FNPs enhancing plant biomass and fruit yield published in 2013, using bitter melon as a modal to evaluate the effects of C60(OH)20 FNPs on seeds (Kole et al. 2013). Gao et al. reported that C60(OH)24–26 FNPs stimulated seedling growth of Arabdopsis thaliana (Gao et al. 2011). The results revealed that FNPs have possible promises for applications on regulating plant growth and boosting crop production and quality in agriculture. To be agriculturally useful, the major problem to overcome stems from the small-scale productivity and intolerably high costs in FNPs production. Therefore, facilitating the economical mass production and decreasing costs become necessary for agricultural applications of FNPs.

Various synthesis methods of FNPs have extensively been investigated since the early 1990s, including the use of nitronium chemistry (Chiang et al. 1992b), sulfuric and nitric acid (Chiang et al. 1992a; Chiang et al. 1993; Vileno et al. 2006), hydroboration (Schneider et al. 1994), oleum (Chiang et al. 1994), radical reaction (Chiang et al. 1996; Ratnikova et al. 2004), potassium (Arrais and Diana 2003), solid-state mechanochemical reaction (Zhang et al. 2003), polybrominated derivative (Mirkov et al. 2004), solvent-free reaction (Wang et al. 2005), hydrogen peroxide (Kokubo et al. 2008; Kokubo et al. 2011), sodium zincate (Wang et al. 2010), alkaline-oxidation (Alves et al. 2006; Husebo et al. 2004; Li et al. 1993, 2012a; Xing et al. 2004; Yao et al. 2010; Zhang et al. 2004), and so on. Among these methods, one of the predominated pathways to fabricate FNPs is O2/NaOH approach developed by Li et al. in 1993 (Li et al. 1993). This reaction was a few volume of extremely concentrated NaOH solution (e.g., 1.0 g/ml) with fullerene assisted by catalyst tetra-n-butylammonium hydroxide (TBAH) at room temperature under air atmosphere, followed by washing and precipitation with a huge amount of methanol and further purification. Henceforward, several research groups were dedicated to modify or improve this path, for example, using a large volume of extremely concentrated NaOH solution and prolonging reaction time from 10 h to 2 days (Husebo et al. 2004), adjusting the feed ratio of NaOH to fullerene (Xing et al. 2004), substituting of TBAH with PEG 400 (Alves et al. 2006; Zhang et al. 2004), replacing methanol precipitation by dialysis (Yao et al. 2010), as well as separation of the different components of FNPs with isoelectric focusing technology (Li et al. 2012a; Xiong et al. 2016). Furthermore, it was evidenced on the basis of experiments and calculations that oxygen (O2) played an important role in these procedures (Alves et al. 2006; Husebo et al. 2004; Wang et al. 2015a; Wang et al. 2015b; Xing et al. 2004). Recently, the reasonable mechanism of FNP formation under O2/NaOH condition has been studied by density functional theory calculations that the reaction experienced via three steps: hydroxylation, i.e., nucleophilic addition of OH−; oxidation, i.e., electrophilic addition of O2; and structural rearrangement (Wang et al. 2015a). Although being a currently popular and powerful method in laboratory from an academic point of view, facile and scalable fabrication of FNPs with O2/NaOH still has several disadvantages from an engineering point of view. First is the uncontrollability and insufficient of the oxidation in laboratory practice and consequently the generation of a mass of water-insoluble by-product. Secondly, the resulted products, such as C60(OH) x x = 8.5 (Zhang et al. 2004), x = 18–20 (Alves et al. 2006), x = 24–26 (Li et al. 1993), x = 27 (Zhang et al. 2004), x = 30, 32, 36, 42, 44 (Xing et al. 2004), or, Na n +[C60O x (OH) y ]n- n = 2–3, x = 7–9, y = 12–15 (Husebo et al. 2004), do not have good reproducibility due to their complicated structure and chemical composition, which have variable number and sites of hydroxyl groups (−OH) on the surface of the fullerene cages, and possibly, the simultaneous formation of impure groups such as ketone structures (Husebo et al. 2004; Xing et al. 2004). Thirdly, the solubility of fullerene in the reported solvents is very low at room temperature (i.e., the solubilities of C60 were 1.7 and 2.8 mg/mL in benzene and in toluene (Beeby et al. 1994), respectively) and also is a barrier to large-scale bulk production of FNPs. Finally, the consumption of extremely concentrated NaOH solution and a large amount of toluene (or benzene) and methanol is associated with strong corrosion and heavy pollution, and also is quite uneconomical.

Herein, we further improved the method to fabricate FNPs under O2/NaOH condition up to super 10 g in laboratory, as a facile, labor-saving, economical, and efficient one-step technique using mildly concentrated NaOH solution (i.e., 0.1 g/mL) as the hydroxylation agent and the solvent with high-density and high dissolving capacity, followed by dialysis. That is, to our knowledge by far, the simplest one-step strategy of facile and large-scale fabrication of FNPs. The novel FNPs were characterized by various spectroscopic methods, the thermal, nanoassemble and electronic properties were also discussed. Meanwhile, the effect of the water-soluble FNPs on scavenging of free radicals was investigated in vitro as well. Furthermore, maybe as a nanoantioxidant, it enhanced the drought tolerance of maize. In particular, it was preliminarily found that the maize germination was promoted by the newly synthesized FNPs.

Experimental procedure

Materials and equipment

All chemicals, unless otherwise stated, were purchased from commercial sources and used without further purification. C60 (purity >99.9%) was purchased from Henan Puyang Yongxin Reagents Company (China). 1,1-Diphenyl-2-picrylhydrazyl radical (DPPH·), 2,2′-azino-bis(3-ethylbenzo-thiazoline-6-sulfonic acid diammonium salt) (ABTS) and potassium persulfate (K2S2O8) were purchased from Sigma-Aldrich (Shanghai) Trading Company (China). All of the other reagents and solvents were at least analytical grade.

The product was analyzed with matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS, Autoflex, Bruker, Germany). The high-resolution 1H nuclear magnetic resonance spectra were recorded in Fourier transform mode at 22 °C on a Bruker AVANCE 500 spectrometer (1H–NMR, Bruker, Germany). Fourier transform infrared spectra were performed on a Thermo Scientific Nicolet iN10 MX spectrometer (FTIR, Madison, USA). UV-Visible absorption spectra under ambient temperature were recorded on a Shimadzu UV-2550 spectrophotometer (UV-Vis, Shimadzu, Japan). The absorbances were recorded on a Unico 2100 spectrophotometer (Shanghai Instruments, China). Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) were performed on a NETZSCH STA 449C Thermogravimetric Analyzer (TA Instruments, Germany) attached to a model 2910 DSC cell, at a heating rate of 10 °C/min under nitrogen (N2) atmosphere, in a scanning range from 45 to 800 °C. The average hydrodynamic diameter (HD), particle size distribution, and polydispersity index of the product were measured using a NanoBrook Omni particle size and zeta potential analyzer (Brookhaven Instruments, USA). The scanning electron microscopy images were photographed by a Hitachi S-4800 field-emission microscope (SEM, Hitachi, Japan). All electron spin resonance (ESR) experiments were carried out by using a JES-FA200 X-band ESR spectrometer (JEOL, Japan). The powder and aqueous solution of the sample were placed in the thin sealed glass capillary (0.5 mm).

Large-scale preparation of fullerenols

The water-soluble fullerenols was synthesized under O2/NaOH condition with some modification (Husebo et al. 2004; Xing et al. 2004). Briefly, 400 mL 0.1 g/mL (equal to 2.5 mol/L) NaOH was mixed with 400 mL o-dichlorobenzene (ODCB) solution containing 10.8 g pristine fullerene C60 in a 2000 mL volume of beaker, and then 1.0 mL 40% TBAH were added as phase-transfer catalyst. The mixture was vigorously stirred by a mechanical agitator at room temperature under air. For 24 h, when the deep purple ODCB layer on the understratum of the reaction system became colorless, while the aqueous layer on the superstratum was changed from the originally colorless to deep brown, the ODCB layer was separated and washed twice with deionized water. The combined aqueous solution was filtered, followed by dialysis against deionized water using a dialysis membrane with molecular weight cutoff of 3500 Da until the pH of effluent was consistent. The dialysate was filtered through a 0.45-μm filter membrane, concentrated and dried on a rotary evaporator at 60 °C. Finally, the purified powder was obtained and marked as Fol, whereafter identified as [C60(OH)22·8H2O]n; yield 17.47 g, 94.1%.

ABTS+· scavenging assay

The scavenging activity to ABTS+· radicals were measured, following the method in the literatures with slight modification (Li et al. 2012b; Wang et al. 2011). Briefly, ABTS+· was produced by mixing 20 mmol/L phosphate buffer solution of 7.4 mmol/L 4.0 mL ABTS and 2.6 mmol/L 4.0 mL K2S2O8. The mixture was kept in the dark at room temperature for 12 h to allow completion of radical generation, followed by dilution with phosphate buffer solution so that the absorbance of the stock ABTS+· at 734 nm was 1.32 ± 0.02. Then, 1.5 mL diluted ABTS+· solution was mixed into 3 mL by diluting the phosphate buffer solution of Fol, which the concentrations varied between 12.5 and 225 μg/mL. The UV-Vis spectra and the absorbances at 734 nm were recorded at 6 min after addition of Fol. The inhibition percentage of the samples was calculated using Eq. (1):

$$ \mathrm{Inhibition}\%=\frac{A_0-A}{A_0}\times 100\% $$
(1)

where A 0 and A are the absorbances of ABTS+· at 734 nm before and after infusion of Fol for each treatment, respectively.

DPPH· scavenging assay

The scavenging activity to DPPH· was determined, following the method in the literature (Li et al. 2012b). Briefly, 11 samples were prepared that 1.5 mL of the stock ethanol-water (v/v = 20:80) binary solution of DPPH· (110 μg/mL) were mixed into 3 mL using the binary solution of Fol with the concentrations between 10 and 383 μg/mL. The absorbances at 519 nm were recorded at 30 min after addition of Fol. The inhibition percentage of the samples was calculated with Eq. (1). The absorbance of Fol at 519 nm must be deducted.

Maize treatments

Mature maize (Zea mays L. cv. Zhengdan 958) seeds were selected and sterilized as described in our previous work (Wang et al. 2014). All seeds were washed thoroughly and soaked with water, 5 and 50 μg/mL of Fol, respectively. The seeds were placed in culture dish with water, 5 and 50 μg/mL of Fol incubated at 25 °C for germination 20 h after inhibition, respectively. The germination of seed samples was analyzed as follows: seed germination energy (%) = the number of germinated seeds on 2 days × 100 / the number of original seeds; seed germination rate (%) = the number of germinated seeds on 4 days × 100 / the number of original seeds; seed germination index (GI) = ∑(Gt/Dt), where Gt is the germination rate at different days and Dt is the time in days. The germinated seeds were taken photos at 4 days after germination.

After germination, the maize seedlings were cultured with water, 5 and 50 μg/mL of Fol in a light chamber (day 28 °C/night 22 °C, relative humidity 75%) with a 16/8 h day/night cycle, respectively. When the second leaves were fully expanded, the seedlings were exposed to 15% PEG (MW6000) solution. The leaves were sampled for physiological parameters analysis at 2, 10, 24, or 48 h after treatment.

Contents of MDA and H2O2 and antioxidant enzyme assay

The activity of catalase (CAT) and superoxide dismutase (SOD) and the content of malondialdehyde (MDA) were measured with the corresponding kids according to the manufacturer’s specifications (Jiangcheng, China). Leaf samples (0.1 g) were homogenized with a mortar and pestle at 4 °C in 0.9 ml normal saline to examine the activity of CAT and the content of MDA and H2O2. The homogenates were centrifuged according to the instructions. For SOD activity, leaf samples (0.25 g) were homogenized with a mortar and pestle at 4 °C in 1.0 ml normal saline. The homogenates were centrifuged at 3500 rpm for 10 min. The supernatants were measured by a spectrophotometer. Finally, the corresponding contents of MDA and H2O2 and antioxidant enzyme activities were analyzed according to the instructions.

Statistical analysis

Each physiology and radical scavenging experiment was at least triplicate. Data were expressed as means ± standard error (SE). All mean comparisons were done using the Student’s t test for independent samples. In all cases, the confidence coefficient was set at 0.05.

Results and discussions

General approach for facile and scalable fabrication of FNPs

Our engineering aimed to produce FNPs with large-scale productivity. The general approach for one-step preparation of the water-soluble FNPs under improved O2/NaOH condition is depicted in Fig. 1, which fabrication up to 10 g was performed in laboratory and could also be directly used in industrial production if needed. Previously, no one considered that extremely concentrated NaOH (e.g., 1.0 g/mL) solution was strong corrosion and unnecessary as the hydroxylation reagent. The low-volume concentrate of NaOH led to form bulk quantities of sludge at the onset of the hydroxylation step, which O2 in air was hard to further attack the sludgy intermediates on the bottom of the reaction container and subsequently resulted in unacceptable water-insoluble by-product. Furthermore, a huge volume of the solvents with low dissolving capacity were utilized without regard to high pollution, barrier to bulk production and economizing through the elimination of the normally high volumes. Herein, our group selected the moderately concentrated NaOH (e.g., 0.1 g/mL) solution as the hydroxylation reagent and ODCB (the value of C60 was 27 mg/mL under room temperature (Beeby et al. 1994)) as the best solvent after tens of experiments. In comparison with the approaches in the literatures (Alves et al. 2006; Husebo et al. 2004; Li et al. 1993), the innovation of our engineering is significance that the feed concentration of the hydroxylation reagent cut down to one tenth and the volume of the solvent decrease at least by 90% at an equal productivity of FNPs. Significantly, the hydrophilic intermediates from the hydroxylation step (Wang et al. 2015a) were quickly coming close to the aqueous NaOH layer on the upper of the reaction system with the H2O-ODCB binary solvents and had plenty of opportunity exposure to O2 originated from air, attacked easily by O2, followed closely by conversion into the water-soluble product after exhaustive rehydroxylation and structural rearrangement, due to the availability of a large volume of mildly concentrated NaOH solution as the hydroxylation reagent and ODCB as solvent. Almost no sludge or precipitate was detected in this innovative process. After reaction, the raw product was separated from ODCB phase and filtered, and the final product of the water-soluble FNPs (abbreviated as Fol) was dried and then obtained after dialysis, to avoid the repeated precipitation with vast methanol, for a complete removal of TBAH and NaOH.

Fig. 1
figure 1

General approach for one-step fabrication of water-soluble C60(OH)22·8H2O with large-scale productivity by improved alkaline-oxidation approach

Full size image

Our one-step strategy for the fabrication of Fol by improved O2/NaOH approach is promising with considerable merits: (1) the reaction can be performed with a minimum of solvent under relatively mild alkaline condition; (2) the hydroxylation and oxidation processes are synergistic and easily manageable; (3) the after-treatment is labor-saving and simple; (4) almost no water-insoluble by-product is formed; (5) the product have almost no other groups (e.g., carbonyl, hereinafter proved by FTIR) except the desired -OH is anchored on the pristine fullerenes; and (6) the engineering leads to a weight increase of the product compared with the pristine fullerenes and increase of productivity.

Characterization

The structure of Fol was characterized by a variety of spectroscopic techniques. Figure 2a was the positive-ion mode MALDI-TOF-MS for Fol in deionized water, obtained using dithranol as matrix, which had only a single prominent ion peak at m/z 720. The isotope profile was expanded views at around m/z 720 and was conformable with those calculated from natural isotopic abundance, which was attributed to isotopic variants of \( {C}_{60}^{+\cdot } \) (Jagerovic et al. 1996). It provided the convincing evidence that Fol was the derivative of fullerene C60. In the 1H-NMR spectrum of Fol shown in Fig. 2b, singlet resonance peak corresponding to the protons of -OH was observed at 4.8 ppm. The results of TOF-MS and 1H-NMR are consistent with what have been observed for other FNPs (Yao et al. 2010). The FTIR spectrum of Fol is depicted in Fig. 2c and showed characteristic features at 1060, 1370, 1587, and 3397 cm−1, which were assigned to νC-OH, δ S C-OH, νC=C, and νO-H adsorptions. These characteristics of signals were reported for ever as diagnostic criteria for various FNPs (Chiang et al. 1992b; Husebo et al. 2004; Kokubo et al. 2008; Kokubo et al. 2011; Li et al. 1993; Schneider et al. 1994; Vileno et al. 2006; Xing et al. 2004), which elucidated the existence of -OH and of the remaining π-bonded carbons of the C60 cage in Fol. It was noteworthy that Fol had no the absorption peak between 1587 and 1800 cm−1, which implied the nonexistence of the impure oxygenous groups such as carboxylic group O-C=O, carbonyl group C=O, and hemiacetal structure (Chiang et al. 1992b; Kokubo et al. 2008; Kokubo et al. 2011; Li et al. 1993; Schneider et al. 1994; Xing et al. 2004). Combined with the MS result, the FTIR result revealed that the molecular structure of Fol was much more stable than those of FNPs reported by Xing et al. (2004), which the impure groups led to a cage-opened structure and lower the stability of FNPs. As shown in Fig. 2d, UV-Vis absorption spectrum of aqueous solution of Fol exhibited a broad asymmetrical decrease in absorbance with the wavelength range of 200 to 900 nm. The spectrum, no visible absorption zones, was featureless, in which similar results were obtained for FNPs synthesized by the variable methods (Kokubo et al. 2008; Vileno et al. 2006).

Fig. 2
figure 2

MALDI-TOF-MS (a), 1H-NMR (b), FTIR (c), and UV-Vis spectra (d) of C60(OH)22·8H2O

Full size image

Thermal analysis

The TGA and DSC measurements were used to evaluate the decomposition behaviors of Fol and also to evaluate the -OH number attached to its carbon cage. The TG, differential thermogravimetry (DTG), and DSC thermograms under N2 atmosphere are shown in Fig. 3. It revealed two-step degradation at “lower” range of temperatures (ca. 45–208 °C) and complex multistep decomposition at “higher” range of temperatures (ca. 208–750 °C). At “lower” range of temperatures, the percentage weight loss of Fol to the extent of 6.35% occurred between 45 and 95 °C in the first step, which the initial weight loss could be negligible due to the removal of low boiling units or volatiles according to the literature (Cataldo et al. 2008b). The second-step decomposition of Fol was due to weight loss (11.47%) of second bound water from 95 to 208 °C. The water molecules entrapped tightly in fullerenols could not be dissociated by the usual heating up to 150 °C (Chiang et al. 1993; Kokubo et al. 2008). It was also supported by a small DSC endotherm at 180 °C and a strong DTG peak. From DTG curve, the onset thermal decomposition began at 95 °C and the bound water was completely removed up to 208 °C, which the maximum decomposition temperature (crest temperature) was obtained at 165 °C.

Fig. 3
figure 3

Derivatograms of C60(OH)22·8H2O obtained under N2 atmosphere at a 10 °C/min heating rate: І, TG curve; II, DTG curve; III, DSC curve

Full size image

At “higher” range of temperatures, the release mass of Fol was assigned to the process of dehydroxylation, which had been usually interpreted in terms of eliminating all of -OH as reported for other fullerenols such as C60(OH)28 (Singh and Goswami 2011), C60(OH)36·8H2O (Kokubo et al. 2008), and C60(OH)40·9H2O (Kokubo et al. 2008). Five small endothermic peaks at 295, 393, 460, 610, and 655 °C were shown in the DSC thermogram, which the dehydroxylated process of Fol experienced clearly to the distinct stepwise degradation. It was associated with these observations in the DTG curve. Fol was decomposed steady and continuously from 208 to 630 °C in the TG curve, and then, the weight loss exhibited a significant staged decrease up to 750 °C. TG trace showed a weight loss to the extent of 28.44% from 208 to 750 °C. In addition, it is noteworthy that a sharp and relatively weak exotherm with crest temperature at around 510 °C was observed by DSC. This makes an agreement with those found for fullerene-amine adduct (Janaki et al. 2000). It may be attributed to some structural rearrangement and/or relaxation process associated with elimination by cleaving the segmental appendant groups (Cataldo et al. 2008a; Janaki et al. 2000).

Thermal analysis techniques have been successfully exploited to calculate the -OH number in fullerenols, which is denoted in Eq. (2) (Goswami et al. 2004; Singh and Goswami 2011).

$$ x=\frac{720}{M_1}\times \frac{y_1}{y_2} $$
(2)

where M 1 is the weight of each appendant -OH and equals to 17. y 1 is the percentage weight loss corresponded to the removal of all of -OH per fullerene cage. y 2 is the percentage weight loss plus char residue due to fullerene C60 only. Turning back to Fig. 3, the summary of the detailed and critical data reproduced in Table 1, sum of x have been calculated to be 22.4 for Fol based on the simultaneous TG-DTG-DSC analysis. Taking into account that molecule cannot be fractioned (Singh and Goswami 2011), thus Fol was described as the average molecular formula C60(OH)22. Moreover, to further validate the number of bound water (denoted as m) in Fol, the authors proposed the following formula (Eq. (3)) to calculate the value of m.

$$ m=\frac{720}{M_2}\times \frac{y_3}{y_2} $$
(3)
Table 1 Calculation of the number of -OH and bound water per fullerene for Fol on the basis of TGA and DSC results
Full size table

where M 2 (equal to 18) is the weight of each water molecule. y 3 is the percentage weight loss corresponded to the removal of all bound water per fullerene cage. Sum of m have been calculated to be 8.48 for Fol, which was further expressed as C60(OH)22·8H2O.

Size distribution and morphology of nanoassembly [C60(OH)22·8H2O]n

C60(OH)22·8H2O (50 μg/mL) in deionized water were successively measured 30 times by dynamic light scattering (DLS) technique with a NanoBrook Omni particle size and zeta potential analyzer, and one outlier was discarded. As shown on Fig. 4, C60(OH)22·8H2O formed aggregates and displayed monomodal nanosized distributions, the mean HD and polydispersity index were 138.1 ± 3.2 nm and 0.35, respectively. These results indicated that C60(OH)22·8H2O had a strong tendency to nanoassembly in water. Figure 5 shows the surface morphologies of the aggregates by SEM, which one drop of the aqueous solution (50 μg/mL) was deposited onto the silicon wafer substrate, followed by drying in air, finally coating with gold. The representative SEM images shown in Fig. 5a, b presented the nanoscale particle-aggregated structure of C60(OH)22·8H2O. Thus, Fol nanoparticles could be described as [C60(OH)22·8H2O]n. As shown in Fig. 5c, d, the microstructure of [C60(OH)22·8H2O]n were exhibited at higher magnification and were irregular near-spheres with smooth surface. The sizes of 60 granules of [C60(OH)22·8H2O]n (in Fig. 5c) were statistically analyzed with a formal randomization procedure. The average size was 122.3 ± 2.2 nm with distribution in the range of 85–150 nm. It was in agreement with the conclusion by DLS analysis.

Fig. 4
figure 4

Size distribution of 50 μg/mL [C60(OH)22·8H2O]n aggregation in deionized water obtained by DLS, showing an average HD of 138.1 ± 3.2 nm, successively measured 30 times and discarded one outlier

Full size image
Fig. 5
figure 5

Representative SEM images of [C60(OH)22·8H2O]n (a) × 50,000, (b) × 50,000, (c) × 100,000, (d) × 350,000. Those of c and d are the local zoom within the area of b

Full size image

ESR measurements of [C60(OH)22·8H2O]n

he ESR spectra of [C60(OH)22·8H2O]n in solid state and in aqueous solution were measured at 22 °C by a conventional X-band spectrometer and shown in Fig. 6a, b, respectively. The silimar spectral patterns were observed but the relative intensity and resolution of the peaks differed significantly. The powder sample yielded an intense and sharp ESR signal at around 3268.2 G and corresponded to g = 2.00001 and S = 1/2, which was in good agreement with the results of Nan +[C60Ox(OH)y]n- n = 2–3, x = 7–9, y = 12–15 at 1.5 K (g = 2, S = 1/2) reported by Husebo et al. (2004). This signal shown in the inset of Fig. 6a for the expanded region of magnetic field from 3215 to 3310 G was no hyperfine structure. The g value and paramagnetic S = 1/2 spin state revealed that an odd number of electrons existed on C60(OH)22·8H2O fullerenol core. It indicated that the newly synthesized [C60(OH)22·8H2O]n was a stable radical such as cyclopentadienyl radicals and could be used as a spin-trapping agent to eliminate free radicals (Husebo et al. 2004).

Fig. 6
figure 6

ESR spectra for [C60(OH)22·8H2O]n in solid-state (a) and in deionized water (b) obtained on a conventional X-band spectrometer at 22 °C. The insets are the region of magnetic field 3215–3310 G

Full size image

In vitro scavenging activity of free radicals by [C60(OH)22·8H2O]n

To investigate whether or not [C60(OH)22·8H2O]n can scavenge free radicals in vitro, experiments were performed using the nitrogen-centered radicals ABTS+· and DPPH· and determined by spectrophotometry, which their structures were shown in the inset of Figs. 7b and 8, respectively. The effects of [C60(OH)22·8H2O]n on the UV-Vis spectra of ABTS+· radicals are depicted in Fig. 7a. A gradual reduction of the intensity of the absorbance peak at 734 nm was associated to the addition of [C60(OH)22·8H2O]n with concentrations from 12.5 to 225 μg/mL, which illustrated that contents of ABTS+· were reduced. In other words, [C60 (OH)22·8H2O]n can scavenge ABTS+· radicals in vitro in a dose-dependent manner. Figure 7b shows the correlations between the clearance and the contents of [C60 (OH)22·8H2O]n. The inhibition percentage of ABTS+· radicals was suddenly increased from 17.2 to 96.2% when [C60OH)22·8H2O]n was added with concentration from 12.5 to 150 μg/mL, and then reached to 96.8% until up to 225 μg/mL. The change of the inhibition percentage correlated to the concentrations was fitted by a three-parameter single exponential function of the nonlinear curve analysis with Origin 7.5 professional software. IC50 value of [C60(OH)22·8H2O]n, defined as the concentration of 50% inhibition percentage (Li et al. 2012b), was calculated to 45.3 ± 0.6 μg/mL. In addition, the change of the inhibition percentage of DPPH· radicals with concentration were plotted in Fig. 8, showed gradual increase from 2.5 to 24.6% after injection of [C60(OH)22·8H2O]n with concentration from 10 to 128 μg/mL. Maximal inhibition percentage of DPPH· (30.4%) was found at around 160 μg/mL, followed by slight decrease (26.5% at 223 μg/mL and 22.8% at 383 μg/mL). It has been reported that DPPH· and ABTS+· may be scavenged via donation of hydrogen atom (H·) to from a stable DPPH-H molecule (Li et al. 2012b) and via an electron (e) transfer reaction (Li et al. 2012b; Prior and Cao 1999), respectively. The scavenging effect of ABTS+· is significantly higher than that of DPPH· by [C60(OH)22·8H2O]n. It suggested that [C60(OH)22·8H2O]n more easily exerted radical scavenging action via donating electron (e) than donating H·. Based on the discussion above, it revealed that [C60(OH)22·8H2O]n will be an ideal candidate as a novel fullerene-based free radical scavenger in biological systems.

Fig. 7
figure 7

UV-Vis spectra (a) and the percentage inhibition (b) of ABTS+· treated by [C60(OH)22·8H2O]n with concentrations increased from 12.5 (I) to 225 μg/mL (X). Data are mean values ± SE for triplicate. The inset of a and b are the region of wavelength from 630 to 840 nm and the structural formula of ABTS+·, respectively

Full size image
Fig. 8
figure 8

The percentage inhibition of DPPH· treated by [C60(OH)22·8H2O]n with concentrations increased from 10 to 383 μg/mL. Data are mean values ± SE for triplicate. The inset is the structural formula of DPPH·

Full size image

Maize seeds germination promoted by [C60(OH)22·8H2O]n

Recently, various CNMs, including SWCNTs/MWCNTs, fullerenes and its derivates, have gained a lot of attentions on their potential applications in plant growth regulation (Khot et al. 2012; Mukherjee et al. 2016). In order to investigate the role in regulating plant growth of [C60(OH)22·8H2O]n, the germination of maize seeds soaked and cultured with [C60(OH)22·8H2O]n were studied. As shown in Table 2 and in Fig. 9a, there was little or no effect on the germination of seeds treatment with 5 μg/mL [C60(OH)22·8H2O]n solution compared with control. However, the germination rate, germination energy, and germination index of seeds were significantly enhanced by 50 μg/mL [C60(OH)22·8H2O]n. For example, the germination rate of seeds treatment with 50 μg/mL [C60(OH)22·8H2O]n solution at 2 days postexposure were 65% while 40% for control. The results revealed that the concentration of FNPs played an important role in seed germination, but the effectiveness varied in different plant species (Lahiani et al. 2015). As shown in Fig. 9b, the shoots and roots of seedlings cultured in 50 μg/mL [C60(OH)22·8H2O]n solution grew better than that of the control, and further studies found that the fresh weight was also obviously increased (data not shown). These findings were in an agreement with the previous reports, which FNPs enhanced bitter melon growth and Arabdopsis hypocotyls length (Gao et al. 2011; Kole et al. 2013). These results suggested that [C60(OH)22·8H2O]n as a plant growth regulator has great potential applications in agriculture.

Table 2 Effects of 0, 5 and 50 μg/ml [C60(OH)22·8H2O]n on maize germination
Full size table
Fig. 9
figure 9

Effects of 0, 5, and 50 μg/ml [C60(OH)22·8H2O]n on seed germination of maize. a Time dependence of the germination rate. b The photograph of the seedlings at 4 days after germination; left, 50 μg/ml; middle, 5 μg/ml; right, 0 μg/ml. Data are mean values ± SE for triplicate

Full size image

Maize seedlings tolerance to drought stress enhanced by [C60(OH)22·8H2O]n

Some exogenous antioxidant substances (e.g., vitamin C as free radical scavenger) mediated plant tolerance to stress (Shalata and Neumann 2001; Xue et al. 2009). Turning back to Figs. 7 and 8, [C60(OH)22·8H2O]n could significantly scavenge ABTS+· and DPPH· free radicals in vitro. Based on previous reports and our results, we speculated that [C60(OH)22·8H2O]n were involved in stimulating plant tolerance to stress.

In order to prove the hypothesis, the physiological indexes, including the protective enzyme (antioxidant enzymes SOD and CAT) activities and the contents of MDA and H2O2 were measured in maize cultured with 0, 5, and 50 μg/mL [C60(OH)22·8H2O]n solution under drought stress simulated by PEG, respectively. As shown in Figs. 10a, b, the SOD and CAT activities in seedlings had improved by [C60(OH)22·8H2O]n compared with control. The content of H2O2, a form of ROS generated as a result of oxidative stress, was decreased in maize seedlings cultured with [C60(OH)22·8H2O]n (Fig. 10c). As an indicator of stress-induced cell damage, the accumulation of MDA reflected usually the level of lipid peroxidation and also indirectly reflected the extent of membrane injury (Mittler 2002). It was noteworthy that MDA accumulation was significantly suppressed by 5 and 50 μg/mL of [C60(OH)22·8H2O]n solution for the entire length of the study (Fig. 10d, P < 0.05). The results showed that exogenous [C60(OH)22·8H2O]n could decrease the stress-induced damages in maize and enhance the plant tolerance to drought stress. Maybe [C60(OH)22·8H2O]n as an antioxidant remove the excessive accumulation of ROS in maize under drought stress and/or partially inhibit the extent of membrane injury like vitamin C, which need to be researched (Shalata and Neumann 2001). The possible reason might be that the new FNPs, to protect and increase the activities of antioxidant enzymes, played a role indirectly in antioxidant and promoted maize stress tolerance. Further studies of maize stress tolerance, enhanced by [C60(OH)22·8H2O]n through whatever mechanism (the former and/or the latter), is necessary.

Fig. 10
figure 10

Effects of 0, 5, and 50 μg/ml [C60(OH)22·8H2O]n on CAT (a) and SOD (b) activities, H2O2 (c) and MDA (d) contents in maize. Data are mean values ± SE for triplicate. Within each set of experiments, bars with different letters were significantly different at the 0.05 level

Full size image

Conclusions

In conclusion, a facile one-step technology to fabricate water-soluble FNPs in large quantity was developed by improved alkaline-oxidation approach using a large volume of moderately concentrated NaOH solution as the hydroxylation agent and ODCB as the solvent. It provided an economic way for industrial scale bulk production and less chemically impure groups of FNPs. The strategy is feasible in operation, convenient in practice, labor-saving, and inexpensive in cost. The resulted product, [C60(OH)22·8H2O]n, were characterized by various measurements including MALDI-TOF-MS, 1H-NMR, FTIR, UV-Vis, TGA, DSC, DLS, SEM, and ESR. ABTS+· and DPPH· scavenging assay in vitro confirm the high efficiency of [C60(OH)22·8H2O]n as a novel free radical scavenger. Additionally, [C60(OH)22·8H2O]n as an excellent candidate CNMs has the potential to serve as the plant defense stimulation agent, paving the avenue to new applications in agri-nanotechnology.