N

AGa

b

c

d

h

••••

a

ARRAA

KOSEH

1

tqteg

E3f

h0

Journal of Hazardous Materials 272 (2014) 20–27

Contents lists available at ScienceDirect

Journal of Hazardous Materials

jo ur nal ho me p ag e: www.elsev ier .com/ locate / jhazmat

ano-based systems for oil spills control and cleanup

ntonio F. Avilaa,b,∗, Viviane C. Munhozb, Aline M. de Oliveirab, Mayara C.G. Santosc,lenda R.B.S. Lacerdac, Camila P. Gonc alvesd

Department of Mechanical Engineering, Universidade Federal de Minas Gerais, Belo Horizonte, MG, BrazilMechanical Engineering Graduate Program, Universidade Federal de Minas Gerais, Belo Horizonte, MG, BrazilChemistry Graduate Program, Universidade Federal de Minas Gerais, Belo Horizonte, MG, BrazilCivil Engineering Program, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil

i g h l i g h t s

New system for oil spill cleanup using the nanotechnology approach.It combines superhydrophobic nanomembranes and oleophilic graphite nanoplatelets.It is environment friendly with oil absorption capacity peak of 32 g/g.The Tea-bag shape was design for keep the system efficient and easy to be recycled.

r t i c l e i n f o

rticle history:eceived 22 November 2013eceived in revised form 19 February 2014ccepted 23 February 2014vailable online 12 March 2014

eywords:il spills clean-upub-micro membranes

a b s t r a c t

This paper reports the development of superhydrophobic nanocomposite systems which are alsooleophilic. As hydrophobicity is based on low energy surface and surface roughness, the electrospinningtechnique was selected as the manufacturing technique. N,N′ dimethylformamide (DMF) was employed asthe polystyrene (PS) solvent. The “Tea-bag” (T-B) nanocomposite system is based on exfoliated graphitesurrounded by PS superhydrophobic membranes. The T-B systems were tested regarding its adsorp-tion and absorption rates. To test these properties, it was employed three different water/oil emulsions,i.e., new and used motor oil, which have physical properties (viscosity and specific gravity) similar toheavy crude oil extracted in Brazil, and vacuum pump oil (which does not form oil/water emulsion).

xfoliated graphiteydrophobicity

It was observed that oil adsorption rate is dependent on oil surface tension, while the absorption rateis mainly dependent on membrane/exfoliated graphite surface area. Experimental data show that oilabsorption rates ranged between 2.5 g/g and 40 g/g, while the adsorption rate oscillated from 0.32 g/g/minto 0.80 g/g/min. Furthermore, T-B systems were tested as containment barriers and sorbent materials withgood results including its recyclability.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

As mentioned by Roulia et al. [1], it is important to emphasizehat oil must be removed from the water and/or soil surface asuickly as possible. They even mentioned that any delay or ineffec-

iveness can make the situation worse. As discussed to Korhonent al. [2], the oil spill cleaning methods can be divided into threeroups: separation and collection of oil from water surface; mixing

∗ Corresponding author at: Universidade Federal de Minas Gerais, Mechanicalngineering, 6627 Antonio Carlos Avenue, College of Engineering—Building, Room250, 31270-901 Belo Horizonte, MG, Brazil. Tel.: +55 3134095238;ax: +55 3134433783.

E-mail addresses: avila@demec.ufmg.br, aavila@netuno.lcc.ufmg.br (A.F. Avila).

ttp://dx.doi.org/10.1016/j.jhazmat.2014.02.038304-3894/© 2014 Elsevier B.V. All rights reserved.

water and oil by applying dispersant agents to assist the naturaldegradation; and in situ burning of the oil spill. The first optionis often the preferable one, as it allows proper disposal of the oil.According to Srinivasan and Viraraghava [3], when oil and watergenerates an emulsified solution, the best method for oil removefrom water is based on adsorption techniques. However, adsorptionand/or absorption techniques alone are not capable of an efficientcleanup. It must be associated to absorbent materials. During thefirst cleanup phase, there is an accumulation of oil on the surfacefollowed by a diffusive process where molecules of oil are able topass through the surface and enter the bulk material.

Among the various materials used for oil adsorption and/orabsorption, biomaterials are the most common employed due tocost and availability. Srinivasan and Viraraghavan [3] tested twotypes of fungi, i.e. Mucor rouxii and Abidia coerulea, chitosan and

A.F. Avila et al. / Journal of Hazardous Materials 272 (2014) 20–27 21

atic re

wwas

Fb

Fig. 1. Electrospinning schem

alnut shells as sorbent materials. The major criticism of theirork is the type of oil used, i.e. standard mineral oil, canola oil

nd cutting oil, which are not commonly encountered into oilpills. Annunciado et al. [4] used of mixed leaves residues, sawdust,

ig. 2. PS nanomembranes SEM observations: (a) top view of electrospun mem-rane; (b) electrospun fiber transverse cut.

presentation. From Ref. [15].

sisal (Agave sisalana), coir (Cocos nucifera), sponge-gourd (Luffacylindrical) and silk-floss (Chorisia speciosa) fibers as sorbent mate-rials for oil removal from aquatic systems. They investigated twosorption conditions, i.e. static and dynamic, for a period of timeup to 24 h. Their results reviewed that under severe agitation theoil sorption is reduced. However, for superhydrophobic materials,e.g. silk-floss, regardless the kinetics of sorption, the oil sorption“capacity” seems to reach an asymptotic value, in their case 85 goil/g sorbent. This value is more than twice the ones reported byChoi and Cloud [5] for milkweed (Asclepias), and 12 times higherthan the results reported by Payne et al. [6] for wood-derived cellu-losic fiber treated with bleach. The reason for such large oil-sorptioncapacity can be attributed to the hydrophobic nature of silk-floss.Moreover, the water uptake seems to a severe limitation on non-

hydrophobic materials, in special those biomaterials, when thesematerials are employed as oil sorbents into dynamic conditions.To overcome Korhonen [2] proposed a highly porous nanocellulose

Table 1Oil physical properties.

ID Type of Oil Viscosity [cP] Specific gravity [g/cm3]

1 Used motor oil 411.03 0.8812 Vacuum pump oil 213.42 0.9033 New motor oil 376.34 0.871

Fig. 3. EG morphology.

22 A.F. Avila et al. / Journal of Hazardous Materials 272 (2014) 20–27

Table 2Nano-modified cotton fabric properties.

Weave Type Areal Density [g/cm2] Water CA [deg] Glycerol CA [deg] �d [mJ/m2] × 10−2 �p [mJ/m2] × 10−2 � [mJ/m2] × 10−2

aTwsiirTsumb

pnwatv1wop

is another indication of surface heterogeneity. Despite these prob-lems, the idea of using a low energy surface material for creating asuperhydrophobic oil sorbent material is interesting.

Satin 0.0258 148 ± 5 161 ± 1

Plain 0.0126 156 ± 3 149 ± 3

erogel coated by a thin film of TiO2, i.e. a hydrophobic surface.hey reported paraffin oil (kerosene) absorption of 30 times the dryeight of the aerogel. Although this result seems to be impressive,

ome comments must be made. First of all, the kerosene viscositys considerable lower than crude oil; second the aerogel structures designed in such a way that it can be easily destroyed by envi-onmental conditions, e.g. rain, wind and ocean waves. Finally, theiO2 coating using atomic layer deposition is another additionaltep that virtually makes this proposed solution for oil spill cleaningnpractical. An alternative to biomaterials is the use of man-madeaterials with high oil absorption rate and low uptake water capa-

ility.Wei et al. [7] studied five different configurations of non-woven

olypropylene nets for oil absorption. To be able to evaluate theseets performance, Wei et al. employed crude oil, 25% and 50%eathered oils. According to them, nets with large porosity are

ble to have high initial sorption rate for all oils. However, due tohis high porosity, the oil retention is poor. The highest sorptionalue for these non-woven polypropylene (PP) sorbents was around4 g/g (grams of oil per grams of sorbent material). The problem

ith these PP nets is the high volume of water absorbed during the

il cleaning up, which reduces the net floating proprieties. A muchractical solution was proposed by Tu et al. [8]. They proposed a

Fig. 4. Water and ethylene glycol droplets woven fabrics: (a) satin; (b) plain.

5.73 ± 0.11 2.62 ± 0.22 8.35 ± 0.334.63 ± 0.10 3.36 ± 0.06 7.99 ± 0.16

surfaced modified sorbent material. The idea was the developmentof a superhydrophobic and superoleophilic polystyrene (PS) filmusing an airbrush method. Despite the methodology proposed by Tuet al. was very simple and easy to be implemented, there are somedraw-backs. One limitation is the PS surface morphology createdduring the airbrushed film deposition. Non-uniform air/solutionflow leads to uneven solvent (tethrahydrofuran—THF) evaporationrate and thus a wrinkled surface. This non-uniform surface allowsthe water droplets to partially sink into the film surface, whichmakes the boundary between the water and the substrate unclear.Furthermore, Tu also reported water droplet instabilities, which it

Fig. 5. Pattern 1 adsorption and absorption rates; (a) satin weave; (b) plain weave.

A.F. Avila et al. / Journal of Hazardous Materials 272 (2014) 20–27 23

with

fs(

2

1ANtwwiafa

pbdtsvpa

t3aapt

m

Fig. 6. T-B systems after 60 min of contact

This paper focuses on development of a novel composite systemor oil spills cleaning up in a “tea-bag” like format, where the outerurface is superhydrophobic and oleophilic while the inner materialexpanded graphite) is also oleophilic.

. Materials and experiments

The materials employed in this research were PS (Mw:90,000 g/mol) and DMF (HPCL grade) acquired from Sigma-ldrich, while expandable graphite (HC11-IQ) was provided byacional Grafite Incorporated. The graphite exfoliation followed

he methodology described by Dhakate et al. [9], i.e. the HC11-IQas expanded at 800–900 ◦C for 10–20 s in a kiln. The second stepas a double mixing process, as described in Ávila et al. [10,11],

.e. an acetone/HC11-IQ (5% concentration) solution was sonicatedt 20 kHz for 30 min followed by high shear mixing at 17,400 RPMor 30 min. Once the double mixing process was performed, thequeous solution was dried in a vacuum furnace for 24 h.

The PS solutions (20/80 ratio—PS/DMF in weight) were pre-ared following the methodology described by Ávila et al. [12] i.e.y sonication at 20 kHz for 90 min. The electrospinning techniqueescribed by Andrady [13] and Ko and Gandhi [14] was employed inhis research. A schematic representation of the device assembled,imilar to the one from Park [15], is described in Fig. 1. The high-oltage power supply employed was a Glassman EH30R model witheak voltage of 30 kV. The polymeric solution flow rate was guar-nteed by a syringe

Pump from Kd Scientific Series 100 model. The needle used inhis pump has an 18 G × 1.5” (diameter of 1.2 mm and length of8 mm). The flow rate employed was between 1.0 and 2.0 ml/h atpplied voltage of 15 kV and the distance between the needle tipnd the collecting plate was 100 mm. Once the nanomembrane was

laced into a Teflon sheet, this membrane can be peeled and theea-bag like (T-B) system can be assembled.

The water contact angle (CA) was measured using an adjustableicropipette (0.1–1.0 �l) as the source of distillated water and a

oil-spill: (a) satin weave; (b) plain weave.

24.0 megapixel camera (Nikon D3200). The public domain softwareImageJ for image-processing [16] was used for water CA measure-ments. At least ten images measurements were performed for eachsample prepared. The nanofibers morphology was investigatedusing a Quanta 200-FEG-FEI-2006 scanning electron microscope.The adsorption rate is defined by the amount of oil sorbed by theexternal skin into the initial 5 min, while the absorption rate isdefined by the amount of oil passed through the external skin andsorbed by internal graphite after 5 min of exposure.

To be able to evaluate the oil absorption rate of the proposedmaterials, three different oils were employed. Table 1 summa-rizes each oil specific gravity and viscosity at room temperature.Note that motor oils (new and used) have viscosity and spe-cific gravity close to heavy crude oil described by Annunciadoet al. [4]. As discussed by Lee and Baik [17], the vacuum pumpoil has good demulsibility to separate quickly from water andresist emulsion formation. This special characteristic could behelpful to evaluate the exfoliated graphite absorption capabili-ties.

As discussed by Toyoda et al. [18], to be able to cleanup any oilspill, the first step is to contain the oil. In many cases, polypropyl-ene nets are used as containing barriers. The major problem withthese nets is the amount of water uptake. To understand the T-Bsystems response as containment barriers and/or sorbent materi-als, two different set of T-B systems were considered. The two testgroups can be described as: (i) PS nanomembrane was directedelectrospun into two different cotton fabrics, i.e. the first one hasa satin weave configuration with area density of 0.0258 g/cm2,while the second one has a plain weave configuration with areadensity of 0.0126 g/cm2. These “nano-modified” fabrics were usedas the external surface/skin for the T-B system; (ii) a single layerPS membrane (0.0475 mm thickness) and multilayered membrane(0.190 mm) were employed as the T-B system external skin. For

the first set, grounded exfoliated graphite was used as the innersorbent material, while for the second group three different typesof exfoliated graphite were employed, i.e. grounded, spiral, and

2 ardou

mi

3

ptFpct

3

oficrp

tprtpgpaFc

nbbrgttfu

rahans0mcPTacTahaw

ibs

4 A.F. Avila et al. / Journal of Haz

odified magnetic graphite by the sol–gel reaction as describedn Wang et al. [19].

. Results and discussion

As discussed by Wang et al. [19], a sorbent is in general aorous material. Therefore, for the initial analysis a high magnifica-ion observation by scanning electron microscopy was performed.ig. 2A and B shows top and transversal cut of the nanomembranesrepared by electrospinning, while the EG is shown in Fig. 3. As itan be noticed, in both cases, the components are porous materials,hus with high probability to be oil sorbent materials.

.1. System with nano-modified cotton external surface

The main goal during this phase is the development of anil-spill containing barrier/net based on T-B nanocomposite con-guration. Following ASTM standard [20], the surface tension wasalculated using Owen–Wendt–Kaelble equation. Table 2 summa-ies the contact angle and the surface tension (�), including theirolar (�p) and dispersive (�d) components.

According to Cheng et al. [21], a superhydrophobic material ishe one with water contact angle of at least 150◦. Still, anotherarameter that can be used for defining a superhydrophobic mate-ial is its surface tension. As discussed by Carré [22], small surfaceensions associated to surface roughness are the two main com-onents to superhydrophobicity. The surface roughness can beuaranteed by the nanofibers random distribution and their surfaceorosity. Furthermore, these nanofibers randomly distributed cre-te locus for air pockets that are able to sustain the water droplets.ig. 4A and B shows these droplets placed on the two nano-modifiedotton fabrics (satin/blue and plain/red).

Based on the contact angle and the small surface tension, bothano-modified cotton fabrics can be assumed as superhydropho-ic materials. The T-B systems behavior as containment barrier orooms were evaluated by measuring its adsorption and absorptionates as a function of time. Distilled water (≈0.5 l) was placed in alass beaker and kept at constant temperature of 25 ◦C. Once theemperature reached the steady state, 100 ml of oil was added tohe water and stirred for a while. After the emulsion oil/water wasormed, the tea bags like composites were placed at the emulsion’spper surface.

As discussed by Toyoda et al. [18], a good containment bar-ier/net must be hydrophobic; it must have a high adsorption and

low/near zero absorption rates. Moreover, it has to be able toold the oil. From Fig. 5A and B some comments can be made. Thedsorption rate per minute (the amount of oil adhered to exter-al skin during the first 5 min divided by the initial weight) foratin wave configuration seems to be between 0.33 g/g/min and.61 g/g/min, while the plain weave one has its adsorption rate/perinute between 0.38 g/g/min and 0.80 g/g/min. These differences

an be attributed to two factors: (i) the surface area covered by theS nanomembrane; (ii) the oil diffusion through the cotton weaves.he satin wave has a 3 × 1 arrangement, i.e. one tow pass overnother one and jumps three ones, while the plain weave has a 1 × 1onfiguration, i.e. each tow pass over another without any jump.hus, the plain weave 1 × 1 configuration will provide larger surfacerea than the 3 × 1 satin weave. However, the satin weave patternas a much larger areal density through the thickness, which canct as a “filter” reducing the oil diffusion process through the cottoneaves.

After 30 min the absorption rate reached an asymptotic behav-or. Samples with no steady-state condition have an absorption rateetween 0.016 g/g/min and 0.042 g/g/min. By observing the cottonatin weave configuration (Fig. 5A), it is possible to notice that for oil

s Materials 272 (2014) 20–27

type 1 (used motor oil—which is the closest one to crude oil consid-ering specific density and viscosity) the lower bound saturationlimit was reached in 30 min and at rate of 2.5 g/g. For cotton plainweave configuration (Fig. 5B), the smallest absorption rate wasreached, again in 30 min and for oil type 1, in a 4.2 g/g absorptionrate. It is important to mention that for both cases, the peak valueswere obtained using oil type 2 (vacuum pump oil). As described byLee and Baik [17], this type of oil has a special characteristic, its highresistance to water/oil emulsification. This is the probable cause ofhigher absorption rate.

As commented by Oebius [23], containment barriers/nets andbooms must be able to retain the water/oil emulsion at the uppersurface by floating. Once this upper surface emulsion is “captured”by the barrier, it is possible to manage the oil-spill. Hence, a smalloil absorption rate associated to the superhydrophobicity seemsto be the best option for containment barriers, as they can floatfor longer periods of time. As commented by Toyoda et al. [18],commercial containment barriers made of polypropylene have anabsorption rate of 15 g/g. The proposed containment barrier config-uration seems to be a much better option with an absorption rateof 2.5 g/g and 4.2 g/g for satin and plain weave, respectively. Theoverall T-B system performance was much better than the ones pro-posed by Aronu [24] and Aisien et al. [25]. The superhydrophobicmembrane blocked the water intake, which lead to a much betterfloating capability, while a much larger surface area could the rea-son for the increase on oil absorption. Those characteristics werenot available on granulated rubber from tires employed by Aronu[24] and Aisien [25].

Fig. 6A and B shows the two T-B systems samples after the60 min absorption time. One final observation on Fig. 6A andB reveals another important characteristic. The satin weave T-Bsystem was not able to hold the oil absorbed, as some leaks sur-rounding them were detected. An opposite result was spotted forthe plain weave configuration. It seems that plain weave textile pat-tern was able to block the oil molecules making the leaking virtuallyimpossible.

The plain weave T-B system seems to be indicated for both,heavy and light crude oil, while the T-B system with satin weavecotton fabric seems to be limited to heavy crude oil due to its pooroil retention capability. After analyzing all data, it is possible tomake some comments: (i) the PS nanofibers electrospun directlyover the cotton fabrics provide excellent water repellent capabil-ity; (ii) from the two weave fabric configuration studied, i.e. satinand plain weave, the plain weave pattern seems to be the mosteffective. This effectiveness can be translated into its adsorptionrate per unit of time and its capability of holding the absorbed oil;(iii) the fabric areal density has be limited to the barrier/net floatingcapability.

3.2. Systems with single or multilayered external nanomembrane

The idea behind this configuration is the development of a sor-bent material that can be placed over the oil/water emulsion. TheseT-B systems have the following characteristics: (i) water repellent;(ii) the oil adsorption guaranteed by the fibers nanoporous and thefibers’ random distribution; (iii) oil absorption capabilities basedon exfoliated graphite inside the T-B system.

To assure the T-B 2 configuration superhydrophobic capabilities,the water contact angles were measured. The PS nanomembraneswere electrospun over a Teflon® sheet, and the nano-membraneswere later on used to prepare the T-B systems. The results areshowed in Table 3.

The T-B systems were prepared by enclosing the exfoliatedgraphite with the PS nanomembranes. The borders were closed byapplying a pressure of 101.3 kPa at temperature of 60 ◦C for 60 s.The oil absorption rate was measured using the same methodology

A.F. Avila et al. / Journal of Hazardous Materials 272 (2014) 20–27 25

Table 3Single layer nano-memberane characteristics.

2] ×

.20

dap[aofmg

e(s5eiercaricph

tam(waaWire

external skin and exfoliated graphite seems to the best optionas sorbent material. Moreover, as shown in Fig. 10A and B, the

Water CA [deg] Glycerol CA [deg] �d [mJ/m160 ± 2 163 ± 4 5.44 ± 0

escribed at Section 3.1. Fig. 7 describes this oil absorption rates function of time for the T-B system 2. The T-B system is com-atible to rubber/butylstyrene blend proposed by Wu and Zhu26]. Furthermore, new system performance is between the uppernd lower bounds described by Wu and Zhu for the rubber blendbtained from disposable tire waste. In this case, the good per-ormance could be due to the large surface area, the resulting

aterial is highly spongy, created during the rubber/butylstyreneraft copolymerization.

The T-B system 2 reached its saturation limit after 30 min ofxposure to oil/water emulsion. The peak of oil absorption rate≈28.5) was obtained by the spiral EG, which has a much largerurface area. Fig. 8A–C shows a typical T-B system pattern 2 after

min of exposure to oil/water emulsion. As discussed by Toyodat al. [18], adsorption process is followed by the absorption and, its most probable to occur during the first 5 min, which is the lin-ar part of absorption rate curve. The adsorption rate per minuteanged from 1.42 g/g/min to 2.49 g/g/min. These differences can beredited mainly to the surface tension between oil/water emulsionnd external skin surface area/roughness. These high adsorptionates hold the oil molecules into the external surface and by cap-llarity the surrounded oil molecules during the initial phase ofleanup. After a certain period of time, the absorption process takeslace. This T-B system configuration fulfills the two conditions, i.e.igh adsorption and absorption rates.

The next step is to investigate how the membrane’s permeabilityo oil will affect the T-B system overall performance. As perme-bility is also function of the membrane thickness, a single layerembrane (thickness of 0.0475 mm) and a four layer membrane

thickness of 0.190 mm) were employed. The total absorption rateas measured after 30 min exposure to an oil/water emulsion. In

ddition to the two exfoliated graphite configurations (groundednd spiral), the magnetic exfoliated graphite (MEG) as described inang [19] was also employed. Finally, to be able to have a compar-

son between the T-B system and others available in literature theesults reported by Aisien et al. [25], Wu and Zhu [26] and Lutfullint al. [27] were also included in Fig. 9A and B.

Fig. 7. T-B systems absorption rate for pattern 2.

10−2 �p [mJ/m2] × 10−2 � [mJ/m2] × 10−2

3.15 ± 0.02 8.59 ± 0.22

Based on observations of these figures, it is possible to drawsome comments: (i) the absorption rate is highly dependent ofthe external skin thickness and density; (ii) the sorbent elementsurface area, in this case the exfoliated graphite, seems to be thecritical issue on absorption capacity; (iii) denser external skins areable to hold the non-absorbed oil inside the T-B system due tonanomembranes surface tension and oleophilic properties; (iv) thecalcination step during the magnetic exfoliated graphite synthesisis the probable cause for MEG poor performance as sorbent mate-rial. This can be due to the drastic decrease on surface area; (v)the results are compatible to the ones reported in literature. Fur-thermore, the T-B system upper bound is close to the highest valueobtained by Lutfullin et al. [27], but not water intake is guaran-teed only on T-B system. The T-B system with folded/multi-layered

Fig. 8. T-B systems pattern 2 after 60 min of exposure to water/oil emulsion: (a) oiltype 1; (b) oil type 2; (c) oil type 3.

26 A.F. Avila et al. / Journal of Hazardous Materials 272 (2014) 20–27

Fig. 9. T-B systems after 30 min of exposure to water/oil emulsion: (a) single layer;(

fr

attotaFiotptdftb

Fig. 10. Photos of T-B systems after 30 min exposure to water/oil emulsion: (a)single layer; (b) multi-layered.

absorption capabilities. The PS nanomembranes provide the super-

b) multi-layered.

olded/multi-layered external skin not only has a high absorptionate but it is also capable of retain the oil inside its boundaries.

Finally, the analysis is not completed without the recyclabilitynalysis. According to Lin et al. [28], a good sorbent material haso be able to combine two important features, i.e. its high absorp-ion capability and its reusability. They even reported a decreasen absorption capacity of 40% after the first reuse. Moreover, afterhe tenth reuse the decrease on oil absorption seems to approachn asymptotic condition around the 40% of its original capability.ig. 11 shows the difference into absorption rate between the orig-nal and after the first reuse of the T-B systems. A sensible decreasen absorption rate (≈83% and 73% for spiral and grounded, respec-ively) was observed into the cases of EG pure conditions. Thishenomenon can be attributed to decrease on surface area due tohe compressive process to extract the absorbed oil. A completeifferent pattern was notice for the MEG samples (decrease rangedrom 2 top 29%). The areal density suffered a much smaller varia-

ion due to the compressive process for oil extraction. This coulde explained by the sol–gel process, which somehow clumps the

Fig. 11. Original × recycled T-B systems: oil absorption performance.

graphite platelets. Once this denser state is obtained, it a stable oilabsorption rate is expected to be obtained.

4. Conclusion

A novel system for oil-spill management and cleanup based onpolystyrene (PS) nanomembranes and exfoliated graphite was pro-posed. The tea bag shape was selected due to its adsorption and

hydrophobicity required to block the exfoliated graphite to uptakewater. Used motor oil mimicked crude oil with excellent results.

ardou

sfwbblswl

nbiraecattMdits

A

CMPM

R

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[

[

A.F. Avila et al. / Journal of Haz

Electrospun PS nanofibers directly into cotton fabrics surfaceeems to make this fabrics water repellent. From the two weaveabric configuration studied, i.e. satin and plain weave, the plaineave shape seems to be the most effective. This effectiveness can

e translated into its adsorption rate per unit of time and its capa-ility of holding the absorbed oil; The fabric areal density could

imit the barrier/net floating capability. The plain weave cotton T-Bystem seems to be indicated for both, heavy and light crude oil,hile the T-B system with satin weave cotton fabric seems to be

imited to heavy crude oil due to its poor oil retention capability.T-B systems based on external skins made exclusively of PS

anofibers and exfoliated graphite seems to be an efficient sor-ent material. The results indicate that: (i) the adsorption rate

s independent of the external skin thickness; (ii) the absorptionate, however, is highly dependent of the external skin thicknessnd density; (iii) the sorbent element surface area, in this case thexfoliated graphite, seems to be the critical issue on absorptionapacity; (iv) denser external skins are able to hold the non-bsorbed oil inside the T-B system due to nanomembranes surfaceension and oleophilic properties; (v) the calcination step duringhe magnetic exfoliated graphite synthesis is the probable cause for

EG poor performance as sorbent material. This can be due to therastic decrease on surface area. However, when the recyclability

s evaluated the MEG’s samples performed very well. The T-B sys-em with folded/multi-layer external skin and exfoliated graphiteeems to be the best option as sorbent material.

cknowledgments

This research was supported in part by the Brazilian Researchouncil (CNPq) under grants number 303447/2011-7 and theinas Gerais State Research Foundation (FAPEMIG) grant TEC-

PM00192-12. The authors are grateful to the UFMG’s Center oficroscopy and Microanalysis for the technical support.

eferences

[1] M. Roulia, K. Chassapis, C.H. Fotinopoulos, T.H. Savvidis, D. Katakis, Dispersionand sorption of oil spills by emulsifier-modified expanded perlite, Spill Sci.Technol. Bull. 8 (2003) 425–431.

[2] J.T. Korhonen, M. Kettunen, R.H.A. Ras, O. Ikkala, Hydrophobic nanocelluloseaerogels as floating, sustainable, reusable, and recyclable oil absorbents, Appl.Mater. Interfaces 3 (2011) 1813–1816.

[3] A. Srinivasan, T. Viraraghavan, Oil removal from water using biomaterials,Bioresour. Technol. 101 (2010) 6594–6600.

[4] T.R. Annunciado, T.H.D. Sydenstricker, S.C. Amico, Experimental investigationof various vegetable fibers as sorbent materials for oil spills, Mar. Pollut. Bull.50 (2005) 1340–1346.

[5] H.-M. Choi, R.M. Cloud, Natural sorbents for oil spill cleanup, Environ. Sci. Tech-nol. 26 (1992) 772–776.

[

s Materials 272 (2014) 20–27 27

[6] K.C. Payne, C.D. Jackson, C.E. Aizpurua, O.J. Rojas, M.A. Hubbe, Oil spills abate-ment: factors affecting oil uptake by cellulosic fibers, Environ. Sci. Technol. 46(2012) 7725–7730.

[7] Q.F. Wei, R.R. Mather, A.F. Fotheringham, R.D. Yang, Evaluation of nonwovenpolypropylene oil sorbent in marine oil-spill recovery, Mar. Pollut. Bull. 46(2003) 780–783.

[8] C.-W. Tu, C.-H. Tsai, C.-F. Wang, S.-W. Kuo, F.-C. Chang, Fabrication of super-hydrophobic and a superoleophilic polystyrene surfaces by a facile one-stepmethod, Macromol. Rapid Commun. 28 (2007) 2262–2266.

[9] S.R. Dhakate, N. Chauhan, S. Sharma, J. Tawale, S. Singh, P.D. Sahare, R.B. Mathur,An approach to produce single and double layer graphene from re-exfoliationof expanded graphite, Carbon 49 (2011) 1946–1954.

10] A.F. Ávila, A.M. de Oliveira, G.R.B.S. Lacerda, V.C. Munhoz, M.C.G. Santos, P.F.Santos, M. Triplett, A nano-modified superhydrophobic membrane, Mater. Res.16 (2013) 609–613.

11] A.F. Ávila, J. de Ávila Jr., D. Jardim, M. Tripplet, Moving from hydrophilic tosuper hydrophobic surfaces, in: M. Hyer (Ed.), Proceedings of the 26th AmericanSociety for Composites Technical Conference, Destech Publishers, Lancaster,2011, CDROM.

12] A.F. Ávila, M.I. Yoshida, M.G.R. Carvalho, E.C. Dias, J. de Ávila Jr., An investi-gation on post-fire behavior of hybrid nanocomposites under bending loads,Composites Part B 41 (2010) 380–387.

13] A.L. Andrady, Science and Technology of Polymer Nanofibers, John Wiley &Sons, New York, NY, 2008.

14] F.K. Ko, B. Gandhi, Producing nanofiber structures by electrospinning for tissueengineering, in: Textiles, P.J. Brown, K. Stevens (Eds.), Nanofibers and Nano-technology, Woodhead Publishing Company, New York, NY, 2007, p. 22.

15] J.-S. Park, Electrospinning and its Applications, Adv. Nat. Sci.: Nanosci. Nan-otechnol. 1 (2010) 043002-07.

16] T.J. Collins, ImageJ for Microscopy, BioTechniques 43 (2007) S25–S30.17] C. Lee, S. Baik, Vertically-aligned carbon nano-tube membrane filters with

superhydrophobicity and superoleophilicity, Carbon 48 (2010) 2192–2197.18] M. Toyoda, M. Inagaki, Heavy oil sorption using exfoliated graphite: new appli-

cation of exfoliated graphite to protect heavy oil pollution, Carbon 38 (2000)199–210.

19] G. Wang, O. Sun, Y. Zhang, J. Fan, L. Ma, Sorption and regeneration of magneticexfoliated graphite as a new sorbent for oil pollution, Desalination 263 (2010)183–188.

20] ASTM D7490-08, Standard test for measurement of the surface tension of solidcoatings, substrates and pigments using contact angle measurements, in: ASTMYearbook, ASTM, Boras, 2008, pp. 1–4.

21] Y.-T. Cheng, D.E. Rodak, Is the lotus leaf superhydrophobic, Appl. Phys. Lett. 86(2005) 144101–144103.

22] A. Carré, Polar interactions at liquid/polymer interfaces, J. Adhes. Sci. Technol.21 (2007) 961–981.

23] H.U. Oebius, Physical properties and processes that influence the cleanup of oilspills in the marine environment, Spill Sci. Technol. Bull. 5 (1999) 177–289.

24] U.E. Aronu, Absorption mats for oil decontamination: towards sustainable end-of-life tyre management, in: Master’s Thesis, University of Boras, Sweden, 2007,p. 86.

25] F.A. Aisien, F.K. Hymore, R.O. Ebewele, Potential application of recycled rubberin oil pollution control, Environ. Monit. Assess. 85 (2003) 175–190.

26] B. Wu, M.H. Zhou, Recycling of waste tyre rubber into oil absorbent, WasteManage. 29 (2009) 355–359.

27] M.A. Lutfullin, O.N. Shornikova, A.V. Vasiliev, K.V. Pokholok, V.A. Osadchaya,M.I. Saidaminov, N.E. Sorokina, V.V. Avdeev, Petroleum products and water

sorption by expanded graphite enhanced with magnetic iron phases, Carbon66 (2014) 417–425.

28] C. Lin, Y.-J. Hong, A.H. Hu, Using a composite material containing waste tirepowder and polypropylene fiber cut end to recover spilled oil, Waste Manage.30 (2010) 263–267.