Aminoclay-templated nanoscale zero-valent iron (nZVI) synthesis for efficient harvesting of oleaginous microalga, Chlorella sp. KR-1

Synthesis of aminoclay-templated nanoscale zero-valent iron (nZVI) for e ﬃ cient harvesting of oleaginous microalgae was demonstrated. According to various aminoclay loadings (0, 0.25, 0.5, 1.0, 2.5, 5.0, and 7.5 aminoclay – nZVI ratios), the stability of nZVI was investigated as a function of sedimentation rate. Aminoclay-coated nZVI (aminoclay – nZVI composites) showed optimal dispersibility at the 1.0 ratio, resulting in the smallest aggregated size and uniform coating of aminoclay nanoparticles onto nZVI due to electrostatic attraction between nZVI and aminoclay nanoparticles. This silica-coated nZVI composite (ratio 1.0) exhibited a highly positively charged surface ( (cid:1) +40 mV) and a ferromagnetic property ( (cid:1) 30 emu g (cid:3) 1 ). On the basis of these characteristics, oleaginous Chlorella sp. KR-1 was harvested within 3 min at a > 20 g L (cid:3) 1 loading under a magnetic ﬁ eld. In a scaled-up (24 L) microalga harvesting process using magnetic rods, micro-algae were successfully collected by attachment to the magnetic rods or by precipitation. It is believed that this approach, thanks to the recyclability of aminoclay – nZVI composites, can be applied in a continuous harvesting mode.


Introduction
2][3] Thanks to its utilization of barren lands and nongrain feedstocks, microalgae-based biorenement is a promising alternative source of transportation fuels. 4,5However, challenges presented by the limitations of the relevant bio-renement processes have to be overcome before microalgae can be considered as a feedstock. 6,7][10] Biorenement's commercialization requires that all of these processes be developed for signicantly improved cost-efficiency.Indeed, many researchers have devoted themselves to the study and improvement of these processes.Conventional centrifugation, occulation by organic/ inorganic coagulants, precipitation by pH increment, ltration including membrane processes, and otation have all been investigated and reported on for harvesting purposes. 7Recently too, the use of magnetic nanoparticles (Fe 3 O 4 ), owing to the fast harvesting and easy recovery thus achievable, has become an issue. 11,12However, the use of magnetic nanoparticles for microalgae harvest is depending on specic pH range at which zeta potential of magnetic nanoparticles shows positively charged surface.As a result, the coating of cationic charged materials onto magnetic particles is demanding. 135][16] Among them, 3-aminopropyl-functionalized magnesium phyllosilicate (denoted as aminoclay) has been widely used in the fabrication of bio(nano)composites [17][18][19][20] and environmental clean-ups, [21][22][23][24] which are associated to specically via a process of selfassembly due to the protonation of amine groups and the delamination of aminoclay sheets in aqueous solution.8][29] In this harvesting application, the recovery and recycling of aminoclay represent major challenges.
One approach to overcoming this limitation is to coat aminoclay onto magnetic nanoparticles (nanoscale zero-valet iron (nZVI) or magnetite (Fe 3 O 4 )) for improved recovery efficiency.nZVI usually is synthesized by reduction of Fe 2+ or Fe 3+ ions with reducing agents (which are applied within the eld of environmental sciences both as reduction and oxidation agents. 30,31) Specically, nZVI is synthesized by solution chemistry in the presence of aminoclay (<50 nm average hydrodynamic diameter) as a template, 32 as an immediate consequence of which, organo-building blocks of aminoclay coat nZVI nanoparticles.In these pages the focus is on nZVI synthesis rather than Fe 3 O 4 , due its stronger magnetic property and larger size for facile aminoclay coating. 33n this context, the present study investigated simple synthesis of aminoclay-coated nZVI for efficient harvesting of an oleaginous microalga, Chlorella sp.KR-1.Harvesting in a batch mode was successfully completed within 3 min.Moreover, a scaled-up (24 L) process for the purpose of continuous harvesting was successfully achieved, which also reduced the aminoclay loading.Thus, this aminoclay-coated-nZVI harvesting approach might represent a promising candidate as a downstream process for microalgae feedstocks in the bio-renement context.
In a typical precipitation method, 34 aer preparation of 4 g aminoclay solution in 50 mL distilled water, 1.9 g FeCl 3 $6H 2 O and 0.7 g FeCl 2 $4H 2 O were dissolved with stirring.To above mixture, 30 mL of 1.0 M NaOH was added rapidly with vigorous stirring.And then centrifugation (6000 rpm for 10 min) was collected with the aminoclay-Fe 3 O 4 composite.
Synthesis of the aminoclay-nZVI composites 35 (i.e., aminoclay-coated nZVI) was performed as noted below.First, the desired amount of aminoclay was solubilized in 100 mL of double-distilled water preparatory to 10 min sonication in a 250 mL beaker, to which was introduced 50 mL of a ferric (Fe 3+ ) solution (35.8 mM, FeCl 3 $6H 2 O, Sigma-Aldrich, USA).Then, 50 mL of sodium borohydride solution (179 mM, Sigma-Aldrich, USA) was introduced into the beaker drop-wise and vigorously mixed for 25 min, thereby reducing the ferric ions to nZVI.The concentrations of solutions were determined based on eqn (1) and the conditions indicated in Table 1.The aminoclay-nZVI composite slurry was aged under the anoxic conditions for 1 day for full coating of nZVI and the resultant reduction of H 2 gas generation.Aer two-times washing with N 2 -gas-purged doubledistilled water, the aminoclay-nZVI composite slurry was ready for further microalgae harvesting.
Chlorella sp.KR-1 strain and growth conditions The oleaginous freshwater microalgal species employed in this study was Chlorella sp.KR-1, 27,28 as cultured in a nutrient media at pH 6.5 (media constituents: KNO

Microalgae harvesting process
Microalgae solution of 1.5 g L À1 microalgae concentration was removed from the 6 L scaled bubble-column reactor.For the purposes of batch-scale microalgae harvesting, the aminoclay-nZVI composites were mixed in a 6 mL microalgal solution in a 15 mL test tube.Aer vigorous mixing for 1 min, multiple test tubes containing various dosages of aminoclay-nZVI composites were le on magnetic ferrite bricks (surface magnetic-eld strength: 1570 G).The supernatant was obtained for optical density (OD) measurement at the 680 nm wavelength for 3 and 30 min at the 2/3 height from the vial bottom using a UV-visible spectrophotometer (UV-1800, Shimadzu, Japan).The harvesting efficiency (%) was determined by equation eqn (2), where OD f and OD i are the nal and initial OD values of the culture media, respectively.All of the experiments were conducted in duplicate, and the values were averaged.For a 24 L-scale demonstration, appropriate amounts of the aminoclay-nZVI composites were directly added to a designed reactor containing 24 L of microalgal solution.To make homogenous mixtures, air-bubbling by means of a metal membrane sparger was maintained for several minutes, aer which 5 magnetic rods (surface magnetic-eld strength: 9200 G; diameter: 22 mm; length: 500 mm) were applied for magnetophoretic separation by microalgae occulated with aminoclay-nZVI composites.

Sample characterization and instruments
Bare nZVI and the aminoclay-nZVI composites were analyzed under transmission electron microscopy (TEM, JEM-2100F) on 300-mesh carbon-coated copper grids.Photographic images were all captured with an iPhone 5 digital camera.

Characteristics of aminoclay-nZVI composites
The as-prepared aminoclay was composed of a cationic metal center (Mg 2+ ) with sandwiching of -(CH 2 ) 3 NH 2 organofunctional pendents via covalent bonding, as shown in its approximate unit structure in ESI, Fig. S2 †. 10 In a proton-rich aqueous solution the amine groups' high-nitrogen density was protonated to yield cationic charged clusters.For delaminated aminoclay nanoparticle (<50 nm)-coated nZVI surfaces, the growth of the nZVI nanoparticles was stable, which is to say, well-dispersible in aqueous solution, due to the electrostatic repulsion by aminoclay nanoparticles.
Sedimentation proles of the synthesized aminoclay-nZVI composites with different aminoclay loadings were obtained by spectrophotometry.Inferences on the sedimentation behavior were evaluated based on the evolutionary characteristics of the suspensions' optical absorbances over time.The normalized absorbances of the suspensions with different amounts of aminoclay are plotted in Fig. 1(a) as functions of time.nZVI nanoparticles, due to their large density and diameter, settle down; consequently, the general tendency of the normalized absorbance is to decrease with time.The sedimentation patterns could be interpreted by the equation [36][37][38] where I t is the absorbance of the solution at time t, I 0 is the initial absorbance, and s is the characteristic time related to the hydrodynamic radius of the particles.
As illustrated in Fig. 1(a), the sedimentation prole was signicantly affected by the aminoclay-Fe weight ratio.With the increase in the aminoclay loading upto 1.0, the stability was signicantly increased.However, at ratios higher than 2.5, sedimentation was accelerated, which fast sedimentation might be due to the signicant occulation phenomena.The characteristic time, s, as calculated from the sedimentation prole, is presented in Fig. 1(b).According to the sedimentation proles, the characteristic time varied from 4.30 to 539 min, and the aminoclay-nZVI composite showed the most stable colloidal behaviors at the 1.0 ratio.
It is tting that the aggregated size (diameter, nm) as measured by dynamic light scattering was the smallest, 130 nm, at the aminoclay-nZVI composite ratio of 1.0 (Fig. 1c).At that time, the slurry of the aminoclay-nZVI composites changed to a highly positively charged surface: from À12 mV in the absence of aminoclay to $+40 mV aer aminoclay coating (Fig. 1e).
The morphological alternations of the aminoclay-nZVI composites also were examined (Fig. 2).Bare nZVI normally synthesizes chain-like spheres due to magnetic interaction and the core-shell structure of thin-shell iron oxides (g-Fe 2 O 3 / Fe 3 O 4 ) in Fe(0) (Fig. 2a and b). 39,40In the presence of aminoclay however, the nZVI surface, in the present results, was uniformly coated with aminoclay (Fig. 2c-f); this result corresponded with the elemental mapping data (see ESI, Fig. S3 †), indicating that Fe is from nZVI, Mg and Si represent the presence of aminoclay, and O and Cl are from both nZVI and aminoclay.Accordingly, the silica-coated nZVI hybrids became disconnected as the aminoclay loading increased (see ESI, Fig. S4 †).At the higher aminoclay concentrations, linking of the nZVI nanoparticles was prevented by the steric effect, and signicant occulations occurred along with a signicant increase of coating thickness, leading to increased sedimentation.
Aer 1 week's aging (i.e., as stored in aminoclay solution; see ESI, Fig. S5 †), the microalgal morphology was changed to a more chestnut-like shape where nZVI was oxidized, but the aminoclay coating was retained.Notably, in the case of coprecipitation of Fe 3 O 4 synthesis in the presence of aminoclay (see ESI, Fig. S6 †), the aminoclay coating was not inconsistent, owing to the smaller Fe 3 O 4 size (<10 nm), which might be not suitable for aminoclay-particle coating.A larger size of Fe 3 O 4 for aminoclay-coating applications, then, is recommended.
For their magnetic properties (see ESI, Fig. S8 †), the nonlinear hysteresis loops with nonzero remnant magnetization (M r ) and coercivity (H c ) exhibited ferromagnetic properties like iron oxide core-shell structures. 42The M s values of nZVI and aminoclay-nZVI (ratio 1.0) showed $100 and $35 emu g À1 , respectively.The M s value was reduced about 3-fold by aminoclay coating.This was owed to the organofunctional pendents in aminoclay. 43

Batch-mode microalgae harvesting by aminoclay-nZVI composites
Based on the stability results of aminoclay-nZVI composites, the 1.0 ratio was selected as optimal for microalgae harvesting.Therefore, this highly positively charged composite was tested on a laboratory scale (10 mL vials) with oleaginous Chlorella sp.KR-1 according to loading amount under an external magnetic eld.At > 20 g L À1 of aminoclay-nZVI composite, the microalgae was precipitated within 3 min (Fig. 4a and b and ESI,  Fig. S9 and Movie S1 †).This is a very fast reaction, comparable to the previously reported results. 11,12For this loading range, the harvesting efficiencies attained were $100% (see ESI, Fig. S10 †).To investigate the effect of only-aminoclay (29.96 mg) and pure nZVI (84.79 mg), the concentrations of them were calculated based on 19.13 g L À1 of aminoclay-nZVI composite, microalgae harvesting experiments were performed.The harvesting efficiencies by only-aminoclay were $65% at 3 min and $70% at 30 min, while those of pure nZVI were $5% at 3 min and $40% at 30 min (Fig. 4c).It is imperative that aminoclay-coated nZVI shows a synergic harvesting efficiency from the viewpoint of both efficiency and time for equivalent amounts.The mechanism of efficient harvesting by aminoclay-nZVI composites is attributed mainly to the cationic property of aminoclay and induction of force from nZVI.The morphologies of the microalgae harvested by aminoclay-nZVI were shown to be intact, as conrmed by chlorophyll-emitting red uorescence and SEM images (see ESI, Fig. S11 †), meaning that cell effects by radical generation of nZVI was negligible, 44 Notably, the pH was changed from 6.2 to $9.0 in most cases (see ESI, Fig. S12 †); this was due to the aminoclay's protonated amine groups and nZVI's marginal production of H 2 where the aminoclay coating covered the nZVI surface by non-covalent bonding.
Feasibility study of scaled-up (24 L) microalgae harvesting process with aminoclay-nZVI composite (ratio 1.0)3]10 To that end, a rectangular reactor (50 cm Â 12 cm Â 40 cm) was designed (see the Fig. S13 † scheme).Flocculated microalgae biomass was collected on the surfaces of magnetic rods inserted into microalgal solution or precipitated for 15 min to facilitate harvesting (see ESI, Fig. S14 and Movie S2 †).Aer nishing harvesting, some aminoclay was released (<0.2 g L À1 ), nZVI oxidation proceeded slowly.However, it is anticipated that this process, due to its recyclability, could be applied on the tone scale in the continuous mode where integration system is consisted of parallel-based many 20 L tubes.The aminoclay-nZVI composite could easily be separated by acid digestion from a collected biomass, 11 and recovered by Fe 3+ reduction by polyphenols, 45 given that it is considered initial investment cost of magnetic nanoparticle.From now on, optimization of the container, magnetic rods, and mixing techniques is additionally required.Currently, we are planning lipid extraction from wet-microalgal biomasses using Fenton-like reaction with injection of hydrogen peroxide (H 2 O 2 ). 29

Conclusions
In summary, the characteristic time (s) and hydrodynamic diameter of aggregates of aminoclay-coated nZVI composites were evaluated as functions of sedimentation rate.The results showed that the optimal stability was achieved at the 1.0 aminoclay-nZVI ratio, which effected the smallest aggregate size ($100 nm) and a highly positively charged surface.Using the aminoclay-nZVI composite at the ratio (1.0) in which aminoclay uniformly coated the nZVI nanoparticles, oleaginous microalgae harvesting was successfully demonstrated.Within 3 min, $100% of Chlorella sp.KR-1 was harvested at a > 20 g L À1 loading of aminoclay-nZVI composite under an external magnetic eld.A scaled-up (24 L) harvesting process entailing microalgae collection on the surfaces of introduced magnetic rods or by precipitation was successfully conducted.It is believed that this novel strategy, due to the recyclability of aminoclay-nZVI composite and the reduction of aminoclay amount thereby enabled, can be effectively and efficiently applied to continuous harvesting.
) 3 $18H 2 O, 0.01 mM).It had been cultivated at 30 C in a Pyrex bubble-column reactor (working volume: 6 L) equipped with 12 uorescent lamps at the front and right/le sides (light intensity: 80 mmol m À2 s) in a constant-temperature room at 30 C for 7 days.It was supplied with 10% (v/v) CO 2 in air at a rate of 0.75 L min À1 .The oil content of Chlorella sp.KR-1 is 36.5-41.0%(weight of oil/dried biomass).
Powder X-ray diffraction (XRD) patterns were assessed by D/MAX-RB (Rigaku, 12 kW) aer drying (by removal of water using lter papers and tissues and by subsequent milling).The Fourier transform infrared (FT-IR) spectra (FTIR 4100) of KBr pellets composed of 10 wt% vacuum-dried targeting sample and 90 wt% KBr powder were recorded within the 4000 cm À1 -400 cm À1 range.Zetapotential measurements were obtained with a Zetasizer Nano-ZS particle analyzer (Zetasizer nano zs, Malvern, UK).Magnetic measurements of the powdered samples were carried out using a model 4HF vibrating-sample magnetometer (VSM, ADE Co. Ltd, USA) with a maximum magnetic eld of 20 kOe.A quantitative analysis of the aminoclay and nZVI was performed with silicon and total ionic iron by 5 mL of 10 mM HCl in aqueous solution aer 10 min standing by inductively coupled plasma atomic emission spectrometry (ICP-AES, Optima 7300 DV, each calibration curve in ESI, Fig.S1 †).The morphologies of the microalgae and harvested microalgae were examined under bright optical microscopy (Axio Imager A2, Carl Zeiss Microscopy GmbH, Jena, Germany) in the uorescence mode.Also, scanning electron microscopic (SEM) images were obtained following the manufacturer's bio-sample preparation protocol.The pH variation was monitored using a pH/ion meter (D-53, Horiba).