8
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/223937853 Activated Carbon/Iron Oxide Magnetic Composites for the Adsorption of Contaminants in Water ARTICLE in CARBON · JANUARY 2002 Impact Factor: 6.2 · DOI: 10.1016/S0008-6223(02)00076-3 CITATIONS 211 READS 839 6 AUTHORS, INCLUDING: Rachel V.R.A. Rios 25 PUBLICATIONS 856 CITATIONS SEE PROFILE José Domingos Fabris Federal University of Minas Gerais 237 PUBLICATIONS 2,458 CITATIONS SEE PROFILE Vijayendra Kumar Garg University of Brasília 188 PUBLICATIONS 1,277 CITATIONS SEE PROFILE Karim Sapag Universidad Nacional de San Luis 87 PUBLICATIONS 1,214 CITATIONS SEE PROFILE Available from: Vijayendra Kumar Garg Retrieved on: 27 January 2016

Artigo Composto de Ferroe Carbono

Embed Size (px)

DESCRIPTION

lista

Citation preview

Page 1: Artigo Composto de Ferroe Carbono

Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/223937853

ActivatedCarbon/IronOxideMagneticCompositesfortheAdsorptionofContaminantsinWater

ARTICLEinCARBON·JANUARY2002

ImpactFactor:6.2·DOI:10.1016/S0008-6223(02)00076-3

CITATIONS

211

READS

839

6AUTHORS,INCLUDING:

RachelV.R.A.Rios

25PUBLICATIONS856CITATIONS

SEEPROFILE

JoséDomingosFabris

FederalUniversityofMinasGerais

237PUBLICATIONS2,458CITATIONS

SEEPROFILE

VijayendraKumarGarg

UniversityofBrasília

188PUBLICATIONS1,277CITATIONS

SEEPROFILE

KarimSapag

UniversidadNacionaldeSanLuis

87PUBLICATIONS1,214CITATIONS

SEEPROFILE

Availablefrom:VijayendraKumarGarg

Retrievedon:27January2016

Page 2: Artigo Composto de Ferroe Carbono

Carbon 40 (2002) 2177–2183

Activated carbon/ iron oxide magnetic composites for theadsorption of contaminants in watera a a c b´Luiz C.A. Oliveira , Rachel. V.R.A. Rios , Jose D. Fabris , V. Garg , Karim Sapag ,

a ,*Rochel M. Lagoa ´Departamento de Quımica, UFMG, Av. Antonio Carlos6627,CEP 31270901,Belo Horizonte, MG, Brazil

bLab. Cs. Superficie Medios Porosos, Univ. Nacional de San Luis, R. Chacabuco917 (5700),Argentinac ´ ´Universidade Federal de Brasılia, UNB, Brasılia, DF, Brazil

Received 28 November 2001; accepted 13 February 2002

Abstract

In this work the adsorption features of activated carbon and the magnetic properties of iron oxides were combined in acomposite to produce magnetic adsorbents. These magnetic particles can be used as adsorbent for a wide range ofcontaminants in water and can subsequently be removed from the medium by a simple magnetic procedure. Activatedcarbon/ iron oxide magnetic composites were prepared with weight ratios of 2:1, 1.5:1 and 1:1 and characterized by powder

¨XRD, TG, magnetization measurements, chemical analyses, TPR, N adsorption–desorption isotherms, Mossbauer2

spectroscopy and SEM. The results suggest that the main magnetic phase present is maghemite (g-Fe O ) with small2 3

amounts of magnetite (Fe O ). Magnetization enhancement can be produced by treatment with H at 6008C to reduce3 4 2

maghemite to magnetite. N adsorption measurements showed that the presence of iron oxides did not significantly affect the2

surface area or the pore structure of the activated carbon. The adsorption isotherms of volatile organic compounds such aschloroform, phenol, chlorobenzene and drimaren red dye from aqueous solution onto the composites also showed that thepresence of iron oxide did not affect the adsorption capacity of the activated carbon. 2002 Published by Elsevier Science Ltd.

Keywords: A. Activated carbon; C. Adsorption; D. Magnetic properties, Surface areas

1. Introduction particles for oil spill remediation [4]. However, all thesematerials have the drawback of a small surface area or a

The application of magnetic particle technology to solve small adsorption capacity, which limits their application.environmental problems has received considerable atten- Moreover, the preparation of these magnetic materialstion in recent years. Magnetic particles can be used to requires several steps, and special chemicals and pro-adsorb contaminants from aqueous or gaseous effluents cedures. In this work, high surface area and high ad-and, after adsorption, can be separated from the medium sorption capacity magnetic composites based on activatedby a simple magnetic process. Examples of this technology carbon/ iron oxide were prepared by a simple method andare the use of magnetite particles to accelerate the coagula- used for the removal of contaminants from aqueoustion of sewage [1], a magnetite-coated functionalized effluents.polymer such as a resin to remove radionuclides from milk Activated carbons offer an attractive and inexpensive[2], poly(oxy-2,6-dimethyl-1,4-phenylene) for the adsorp- option for the removal of organic and inorganic con-tion of organic dyes [3] and polymer-coated magnetic taminants from water [5,6]. Due to its high surface area

and porous structure it can efficiently adsorb gases andcompounds dispersed or dissolved in liquids [7–9]. The*Corresponding author. Tel.:155-31-34995-719; fax:155-adsorption of several organic contaminants in water, such31-34995-700.

E-mail address: [email protected] (R.M. Lago). as pesticides, phenols and chlorophenols, has recently been

0008-6223/02/$ – see front matter 2002 Published by Elsevier Science Ltd.PI I : S0008-6223( 02 )00076-3

Page 3: Artigo Composto de Ferroe Carbono

2178 L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183

reported [10–16]. Moreover, activated carbon can easily befunctionalized and used as an efficient adsorbent for heavymetal cationic contaminants [17].

2. Experimental

The composites were prepared from a suspension ofactivated carbon (Aldrich Darco G60, 100 mesh, AmericanNorit) in a 400 ml solution of FeCl (7.8 g, 28 mmol) and3

FeSO (3.9 g, 14 mmol) at 708C. NaOH solution (100 ml,4215 mol l ) was added dropwise to precipitate the iron

oxides. The amount of activated carbon was adjusted inorder to obtain activated carbon/ iron oxide weight ratiosof 1:1, 2:1 and 3:1. The obtained materials were dried inan oven at 1008C for 3 h and characterized by powder

˚ Fig. 1. Bulk sigma magnetization for different activated carbon/XRD (Ni filtered Cu Ka, l5 1.5418 A), TG analysis andFe oxide composites.magnetization measurements carried out in a portable

magnetometer with a fixed magnetic field of 0.3 T [18].21The samples were also analyzed chemically for Fe and

31Fe . Temperature programmed reduction (TPR) profiles, iron oxides or hydroxides onto a carbon surface. After57¨Mossbauer spectra (obtained with a Co/Rh source at preparation, a test with a 0.3 T magnet showed that the

liquid N temperature calibrated witha-Fe) and scanning whole material was magnetic and completely attracted to2

electron microscopy (SEM) analyses were also used for the magnet. The magnetization measurements, Fe content,characterization. The adsorption isotherms were obtained BET surface area and microporous volumes (t-plot meth-in batch equilibrium experiments with 50 mg of the od) obtained for the prepared composites are shown incomposites in 50 ml solutions of phenol (concentrations up Table 1.

21to 500 mg l at pH 5), chloroform (concentrations up to It can be observed from Table 1 that the bulk sigma21500 mg l ), chlorobenzene (concentrations up to 25 mg magnetization increases for a higher Fe oxide content in

21l ), and drimaren red dye (concentrations up to 100 mg the composite. However, this increase is not proportional21l ). All the solutions were kept for 24 h at 2561 8C in a to the Fe oxide content (Fig. 1), indicating that the

temperature-controlled bath. The concentrations of volatile concentration of non-magnetic iron oxides increases for theorganic compounds (phenol, chloroform and chloro- composites with higher activated carbon content. Four ironbenzene) were measured by MIMS (membrane intro- oxides are commonly formed under the reaction conditionsduction mass spectrometry) [19–22]. The concentration of employed, i.e. Fe O (magnetite),g-Fe O (maghemite),3 4 2 3

red drimaren dye was measured by UV/vis spectrophotom- a-Fe O (hematite) anda-FeO(OH) (goethite) [23].2 3

etry. Among these oxides, only the first two, magnetite andmaghemite, are magnetic with magnetizations of 100 and

21 2160 J T kg , respectively [24]. Pure Fe oxide prepared3. Results and discussion without activated carbon showed a magnetization of 62 J

21 21T kg , which is similar to g-Fe O , maghemite.2 331 213.1. Characterization of the composites Chemical analyses of pure Fe oxide showed an Fe /Fe

ratio of approximately 7, which is much higher than the31 21The composites were prepared by the precipitation of original ratio used (Fe /Fe52), probably due to

Table 1Magnetization, BET surface area and microporosity measurements for the activated carbon/ iron oxide composites

Composite Bulk sigma Fe O BET V2 3 microporous3 21magnetization (wt%) surface area (cm g )

21 21 2 21(J T kg ) (m g )

Pure Fe oxide 62 |100 66 0.010Composite 1:1 21 47 – –Composite 2:1 9 33 626 0.172Composite 3:1 7 25 658 0.177Pure activated carbon 0 0 933 0.264

Page 4: Artigo Composto de Ferroe Carbono

L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183 2179

Fig. 3. Powder XRD for the 3:1 activated carbon/Fe oxideFig. 2. Nitrogen adsorption–desorption isotherms for pure iron composite and pure Fe oxide.oxide, activated carbon and the composites.

Thermogravimetric analysis of the pure Fe oxide21oxidation of Fe during preparation. These results suggest showed a weight loss due to water vapour of approximate-

31the formation of mainly Fe compounds in the composite. ly 3% at temperatures lower the 1008C. In the temperatureFig. 2 shows the N adsorption isotherms for pure iron2 range 120–1808C a weight gain of ca. 0.3% is observed,

oxide, pure activated carbon and for the prepared compos- which is likely related to the incorporation of oxygen dueites. The data shown in Table 1 suggest that the BETsurface area and the microporous volume are not sig-nificantly affected by the presence of iron oxide in thecomposites. For example, for the 2:1 composite, reductionsof the surface area from 933 (for pure activated carbon) to

2 21626 m g and for the microporous volume from 0.264 to3 210.172 cm g are observed. In both cases, the reductions

of approximately 33% correspond to the decrease expecteddue to the presence of iron oxide in the 2:1 composite(33% iron oxide and 67% activated carbon). As iron oxidehas a relatively small surface area and microporous volume

2 21 3 21(66 m g and 0.010 cm g , respectively) its presencein the composites should cause a decrease in the surfacearea and microporous volume compared to pure activatedcarbon.

XRD analyses of pure Fe oxide (Fig. 3a) suggest thepresence of a cubic iron oxide phase (d 5 2.50, 2.91, 2.07,

˚1,60, 3.20 A), which may be related to the presence ofmaghemite and also some magnetite. For the compositethese peaks appear broader, suggesting a smaller crystallitesize (ca. 25 nm for pure Fe oxide and 16 nm for thecomposite, obtained using Scherrer’s equation [25]). Weak

˚diffraction peaks are also observed atd 5 2.45 and 2.70 A,which may be related to the presence of small amounts ofgoethite and also some hematite.

57 ¨The Fe Mossbauer spectrum for the pure Fe oxide(Fig. 4a) at liquid N temperature shows sextets with2

hyperfine fields of 47 and 48 T, suggesting the presence ofan oxidized magnetite or maghemite structure. On the

¨other hand, the Mossbauer spectra of the compositeindicate the presence of a mixture of maghemite /magnetiteand some goethite with hyperfine fields of 49, 46, 52 and ¨Fig. 4. Mossbauer spectra for the pure Fe oxide and the 3:138 T [23]. activated carbon/Fe oxide composite at liquid N temperature.2

Page 5: Artigo Composto de Ferroe Carbono

2180 L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183

21 31to the oxidation of Fe to Fe in the oxide [24]. Theactivated carbon/Fe oxide composite (Fig. 5b) showedweight loss due to water of ca. 10% up to 1008C and a65% weight loss from 350 to 5008C related to theoxidation of the carbon. A weight of 25% is found afteroxidation of the activated carbon, which is related to the Feoxide in the 3:1 composite.

The temperature programmed reduction (TPR) profile ofthe prepared Fe oxide showed a peak centered at 4508Cwith a shoulder at ca. 3808C and a broad peak from 500 to

¨800 8C (Fig. 6a). XRD and Mossbauer analyses of thesample with the reduction interrupted after the first peak at500 8C (Fig. 6b) showed the presence of mainly magnetite.This was corroborated by the increase in the bulk mag-

21 21netization from 62 to 95 J T kg . After reduction at900 8C, only pure Fe was detected by XRD, suggestingthat the following reduction processes are taking place:

Fig. 6. TPR analyses for (a) the prepared Fe oxide, (b) the3Fe O 1H 5 2Fe O 1H O T(5008C (1) prepared Fe oxide interrupted at 5008C and (c) the 3:1 activated2 3 2 3 4 2

carbon/Fe oxide composite.02Fe O 1 6H 5 6Fe 1 6H O T .8008C (2)3 4 2 2

¨XRD and Mossbauer analyses of the sample reduced up toSimilar results were observed for the 2:1 activated 5008C also showed the presence of mainly magnetite with

carbon/Fe oxide composite (Fig. 6c), which showed a the sample bulk magnetization increasing from 9 to 16 J21 21peak at 4508C and a broad peak centered at ca. 7508C. T kg . These results suggest that, under controlled

Fig. 7. Magnetization and Fe oxide content of the 3:1 activatedFig. 5. Thermogravimetric analyses, in air, of pure Fe oxide (a) carbon/Fe oxide composite exposed to pH 11, 9, 7, 5, 3 and 1 forand the 3:1 activated carbon/Fe oxide composite (b). 72 h.

Page 6: Artigo Composto de Ferroe Carbono

L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183 2181

conditions, non-magnetic iron oxides, such as hematite and pH 1 leads to complete dissolution of the Fe oxide with thegoethite, and the magnetic maghemite can be reduced by destruction of the composite.hydrogen to produce magnetite. This controlled reduction The morphologies of the 3:1 activated carbon/Fe oxideenhances the magnetization of the composite, which is composite, activated carbon and the prepared Fe oxideespecially important to improve the magnetic separation of were studied by SEM. The micrographs obtained for thesethe absorbent after application. materials are shown in Fig. 8a–c. Fig. 8a and b show

The resistance of the composite to the acidity of the general views of the composite. Small aggregates aremedium was also studied by exposing the 3:1 activated observed, which appear brighter (indicated by arrows),carbon/Fe oxide composite to pH 11, 9, 7, 5, 3 and 1 for supported on the darker surface of the activated carbon.72 h. After this period the composite was dried and its Fig. 8c and d show details of the prepared Fe oxidemagnetization and Fe oxide content were determined. The aggregate and the pure activated carbon.results are shown in Fig. 7.

There was no significant change in magnetization or Fe 3.2. Adsorption of organic compoundsoxide content in the pH range 5–11. At pH 3, a smalldecrease in magnetization occurred, probably due to partial The adsorption of volatile organic compounds, i.e.attack of the acid on the iron oxide, whereas treatment at phenol, chloroform and chlorobenzene, from aqueous

Fig. 8. SEM micrographs of (a) the 3:1 activated carbon/Fe oxide composite, (b) a detail of the composite, (c) pure Fe oxide and (d) pureactivated carbon.

Page 7: Artigo Composto de Ferroe Carbono

2182 L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183

The adsorption of drimaren red, a textile reactive dye,was also studied on pure activated carbon and on the 3:1and 2:1 activated carbon/Fe oxide composites. The results(Fig. 10) show similar adsorption isotherms up to the

21equilibrium concentration of 0.01 g l . At higher con-centrations the isotherms obtained for the compositespresent a deviation from the typical Langmuir-type iso-therm, which might be related to specific interactions ofthe reactive dye with the oxides. The adsorption capacityfor the drimaren red dye decreases in the order activatedcarbon.3:1 composite.2:1 composite. This result is alsoprobably related to the decrease in surface area caused bythe presence of the iron oxide in the composite.

4. ConclusionFig. 9. Adsorption isotherms at 2561 8C for phenol, chloroformand chlorobenzene over the 3:1 activated carbon/Fe oxide compo-

The magnetic composites reported in this work can besite.

prepared with a high adsorption capacity activated carbonby a very simple procedure using available and low cost

31 21solutions onto the 3:1 activated carbon/Fe oxide composite chemicals. Magnetization measurements, Fe /Fewas studied. The adsorption isotherms are shown in Fig. 9. ¨ratios, XRD and Mossbauer data suggest that the mainIt can be observed that the adsorption increases in the magnetic phase formed in the composites is maghemite,order phenol,chlorobenzene,chloroform with adsorption possibly with small amounts of magnetite, goethite andcapacities of approximately, 117, 305 and 710 mg/ hematite. Controlled TPR experiments showed that theg , respectively. Experiments with pure activatedcomposite Fe O oxides in the materials can be selectively reduced to2 3carbon showed higher adsorption capacities of ca. 162, 480 produce magnetite, Fe O , enhancing the magnetization of3 4and 910 mg/g for phenol, chlorobenzene and chloro-comp the composites.form, respectively. On the other hand, it is interesting to No significant decrease in the surface area or in thenote that the presence of the iron oxide does not sig- porosity of the activated carbon was caused by thenificantly affect the adsorption capacity of the activated presence of the Fe oxides in the composite. The compos-carbon on the composite. The lower adsorption capacity of ites showed high adsorption capacities for phenol, chloro-the composite is probably related to the small surface area form, chlorobenzene and drimaren red dye in aqueousof the iron oxide, which decreases the total surface area of solution and, more importantly, no reduction in adsorptionthe material to ca. 30% (Table 1) compared to pure was produced by the formation of the composite. More-activated carbon. over, the magnetic composites showed good resistance

over the pH range 5–11.

Acknowledgements

The authors are grateful to the CAPES/SCyT exchangeprogram and to Prof. Wagner N. Mussel for valuable

¨suggestions about the Mossbauer spectra.

References

[1] Booker NA, Keir D, Priestley A, Rithchie CD et al. Sewageclarification with magnetite particles. Water Sci Technol1991;23(7–9):1703–12.

[2] Sing KS. Technology profile. Ground Water Monitor1994;:60–76.

Fig. 10. Adsorption isotherms at 2561 8C for drimaren red dye [3] Safarik I, Safarikova M, Buricova V. Sorption of wateron the 3:1 and 2:1 activated carbon/Fe oxide composites and pure soluble organic dyes on magnetic poly(oxy-2,6-dimethyl-1,4-activated carbon. phenylene. Collect Czech Chem Commun 1995;60:1448–56.

Page 8: Artigo Composto de Ferroe Carbono

L.C.A. Oliveira et al. / Carbon 40 (2002) 2177–2183 2183

[4] Orbell JD, Godhino L, Bigger SW, Nguyen TM et al. Oil [15] Aksu Z, Yener J. A comparative adsorption/biosorptionspill remediation using magnetic particles. J Chem Educ study of mono-chlorinated phenols onto various sorbents.1997;74:1446. Waste Manage 2001;21(8):695–702.

[5] Pignon H, Brasquet C, La Cloirec P. Coupling ultrafiltration [16] Daifullah AAM, Girgis BS. Removal of some substitutedand adsorption onto activated carbon cloth: application to the phenols by activated carbon obtained from agriculturaltreatment of highly coloured wastewaters. Water Sci Technol waste. Water Res 1998;32(4):1169–77.2000;42(5/6):355–62. [17] Shim JW, Park SJ, Ryu SK. Effect of modification with

[6] Lua AC, Guo J. Adsorption of sulfur dioxide on activated HNO and NaOH on metals adsorption by pitch-based3

carbon from oil-palm waste. J Environ Eng ASCE activated carbon fiber. Carbon 2001;39(11):1635–42.2001;127(10):889–94. [18] Coey JMD, Cugat O, MaCauley J, Fabris JD. A portable soil

´[7] Culp GL, Culp RL. New concepts in water purification, New magnetometer. Revista de Fısica Aplicada e InstrumentalYork: Reinhold, 1974. 1992;7:25–30.

[8] Matson P, Mark HB. Activated carbon: surface chemistry [19] Augusti R, Diass AO, Rocha LL, Lago RM. Kinetics andand adsorption from solution, New York: Dekker, 1971. mechanism of benzene derivatives degradation with Fenton’s

[9] Ruthven DM. Principles of adsorption and adsorption pro- reagent in aqueous medium studied by MIMS. J Phys Chemcesses, New York: Wiley, 1984. A 1998;102:10723–7.

[10] Baup S, Jaffre C, Wolbert D. Adsorption of pesticides onto [20] Silva ACB, Augusti R, Dalmazio I, Windmoller D et al.granular activated carbon: determination of surface dif- MIMS pervaporation processes. Phys Chem Chem Physfusivities using simple batch experiments. Adsorption 1999;1:2501–4.2000;6(3):219–28. [21] Kotiaho T, Kostiainen R, Ketola RA, Ojala M et al.

[11] Garner IA, Watson-Craik IA, Kirkwood R. Dual solute Membrane inlet mass spectrometry: past and future. Advadsorption of 2,4,6-trichlorophenol andN-[2-(2,4,6-trichloro- Mass Spectrom 1999;14:501–27.phenoxy)propyl]amine onto activated carbon. J Chem Tech- [22] Lauritsen FR, Kotiaho T. Advances in membrane inlet massnol Biotechnol 2001;76(9):932–40. spectrometry. Rev Anal Chem 1996;15:237–65.

´ ´[12] Martın-Gullon I, Font R. Dynamic pesticide removal with [23] Perez OP, Umetsu Y, Sasaki H. Precipitation and densifica-activated carbon fibers. Water Res 2001;35(2):516–20. tion of magnetic iron compounds from aqueous solution at

[13] Jung M-W, Ahn K-H, Lee Y, Kim K-P, Rhee J-S. Adsorption room temperature. Hydrometallurgy 1998;50:223–9.characteristics of phenol and chlorophenols on granular [24] Oliveira LCA, Fabris JD, Mussel WN, Rios RVRA et al. Theactivated carbons (GAC). Microchem J 2001;70(2):123–31. effect of Mn substitution on the catalytic properties of

¨[14] Denizli A, Ozkan G, Uc¸ar M. Removal of chlorophenols ferrites. Stud Surf Sci Catal 2000;130:2165–8.from aquatic systems with dye-affinity microbeads. Sep Purif [25] Cullity BD, editor, Elements of X-ray diffraction, MenloTechnol 2001;24(1/2):255–62. Park: Addison-Wesley, 1978, pp. 123–31.